CN112842371A - Image processing method, image processing device, electronic equipment and storage medium - Google Patents
Image processing method, image processing device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN112842371A CN112842371A CN202110126885.XA CN202110126885A CN112842371A CN 112842371 A CN112842371 A CN 112842371A CN 202110126885 A CN202110126885 A CN 202110126885A CN 112842371 A CN112842371 A CN 112842371A
- Authority
- CN
- China
- Prior art keywords
- portal vein
- vein
- node
- hepatic
- target
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Withdrawn
Links
- 238000012545 processing Methods 0.000 title claims abstract description 38
- 238000003672 processing method Methods 0.000 title claims abstract description 22
- 210000004204 blood vessel Anatomy 0.000 claims abstract description 237
- 238000001514 detection method Methods 0.000 claims abstract description 206
- 238000000034 method Methods 0.000 claims abstract description 70
- 210000003240 portal vein Anatomy 0.000 claims description 908
- 210000002989 hepatic vein Anatomy 0.000 claims description 500
- 210000004185 liver Anatomy 0.000 claims description 289
- 230000011218 segmentation Effects 0.000 claims description 149
- 238000010586 diagram Methods 0.000 claims description 140
- 230000002440 hepatic effect Effects 0.000 claims description 103
- 210000001631 vena cava inferior Anatomy 0.000 claims description 16
- 238000004590 computer program Methods 0.000 claims description 10
- 238000007689 inspection Methods 0.000 claims description 5
- 230000008569 process Effects 0.000 description 19
- 238000013528 artificial neural network Methods 0.000 description 8
- 238000002372 labelling Methods 0.000 description 8
- 230000006870 function Effects 0.000 description 6
- 210000003462 vein Anatomy 0.000 description 6
- 230000009466 transformation Effects 0.000 description 5
- 230000002792 vascular Effects 0.000 description 5
- 238000012752 Hepatectomy Methods 0.000 description 4
- 230000008901 benefit Effects 0.000 description 3
- 239000003086 colorant Substances 0.000 description 3
- 238000004891 communication Methods 0.000 description 3
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 238000012549 training Methods 0.000 description 3
- 238000002591 computed tomography Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 201000007270 liver cancer Diseases 0.000 description 2
- 208000014018 liver neoplasm Diseases 0.000 description 2
- 238000010845 search algorithm Methods 0.000 description 2
- 206010028980 Neoplasm Diseases 0.000 description 1
- 206010052428 Wound Diseases 0.000 description 1
- 208000027418 Wounds and injury Diseases 0.000 description 1
- 230000000740 bleeding effect Effects 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 230000008030 elimination Effects 0.000 description 1
- 238000003379 elimination reaction Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000009545 invasion Effects 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000009885 systemic effect Effects 0.000 description 1
Images
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B6/00—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment
- A61B6/50—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment specially adapted for specific body parts; specially adapted for specific clinical applications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/05—Detecting, measuring or recording for diagnosis by means of electric currents or magnetic fields; Measuring using microwaves or radio waves
- A61B5/055—Detecting, measuring or recording for diagnosis by means of electric currents or magnetic fields; Measuring using microwaves or radio waves involving electronic [EMR] or nuclear [NMR] magnetic resonance, e.g. magnetic resonance imaging
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
- A61B5/7267—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems involving training the classification device
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B6/00—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment
- A61B6/50—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment specially adapted for specific body parts; specially adapted for specific clinical applications
- A61B6/504—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment specially adapted for specific body parts; specially adapted for specific clinical applications for diagnosis of blood vessels, e.g. by angiography
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B6/00—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment
- A61B6/52—Devices using data or image processing specially adapted for radiation diagnosis
- A61B6/5211—Devices using data or image processing specially adapted for radiation diagnosis involving processing of medical diagnostic data
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Medical Informatics (AREA)
- Physics & Mathematics (AREA)
- Animal Behavior & Ethology (AREA)
- Molecular Biology (AREA)
- Pathology (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Veterinary Medicine (AREA)
- Public Health (AREA)
- Surgery (AREA)
- General Health & Medical Sciences (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- High Energy & Nuclear Physics (AREA)
- Artificial Intelligence (AREA)
- Radiology & Medical Imaging (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Optics & Photonics (AREA)
- Dentistry (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Physiology (AREA)
- Psychiatry (AREA)
- Signal Processing (AREA)
- Evolutionary Computation (AREA)
- Fuzzy Systems (AREA)
- Mathematical Physics (AREA)
- Vascular Medicine (AREA)
- Apparatus For Radiation Diagnosis (AREA)
Abstract
The present disclosure provides an image processing method, an apparatus, an electronic device, and a storage medium, the method including: acquiring at least one detection image of a target object, wherein the detection image comprises a first detection image of a first target area of the target object and a second detection image of a second target area of the target object; generating a first structural map of the vessel structure in the first target region based on the first detected image; determining a category of each central line in the first structural image, which characterizes the blood vessel branch, based on the first structural image and the second detection image; and carrying out hierarchical division processing on the image of the first target area based on the category of the central line and the first detection image to obtain a target image. The target image may be applied to an auxiliary medical means for a target region of a target object, for example, the target image may be used as preoperative data to guide a medical procedure; alternatively, a medical plan corresponding to the target object may be generated based on the target image.
Description
Technical Field
The present disclosure relates to the field of image detection technologies, and in particular, to an image processing method and apparatus, an electronic device, and a storage medium.
Background
Liver cancer is a common tumor disease, and precise hepatectomy is a treatment means for treating liver cancer. Accurate hepatectomy aims to ensure the completeness and the maximization of functional volume of a residual liver anatomical structure while pursuing the complete elimination of a target focus, controls surgical bleeding and systemic wound invasion to the maximum extent and finally enables a surgical patient to obtain the optimal rehabilitation effect.
Generally, precise hepatectomy depends on accurate segmentation results of hepatic veins, hierarchical division results of hepatic veins, and the like. For example, the segmentation result and the hierarchical segmentation result of the hepatic vein may be obtained by a manual labeling manner.
Disclosure of Invention
In view of the above, the present disclosure provides at least an image processing method, an image processing apparatus, an electronic device, and a storage medium.
In a first aspect, the present disclosure provides an image processing method, including:
acquiring at least one detection image of a target object, wherein the detection image comprises a first detection image of a first target area of the target object and a second detection image of a second target area of the target object;
generating a first structural map of a vessel structure in the first target region based on the first detected image;
determining a category of each centerline characterizing a vessel branch in the first configuration map based on the first configuration map and the second inspection image; and
and carrying out hierarchical division processing on the image of the first target area based on the category of the central line and the first detection image to obtain a target image.
With the above method, by generating a first structural map of the blood vessel structure in the first target region, based on the first structural map and the second detection image, the category of each centerline characterizing a blood vessel branch in the first structural map can be determined; the image of the first target area is subjected to hierarchical division processing based on the category of the central line and the first detection image to obtain a target image, so that automatic division of the first detection image is realized, and the efficiency of dividing the first detection image is improved; and by constructing the first structural diagram, manual marking is not needed, and marking errors caused by manual marking can be reduced.
In a possible embodiment, in a case where the first detection image includes a hepatic vein detection image, generating a first structural map of a vascular structure in the first target region based on the first detection image includes:
determining a connected component included in the first detection image based on a connection relationship between a hepatic vein and an inferior vena cava in the first detection image, wherein the connected component includes a plurality of hepatic vein blood vessel branches having a connection relationship;
determining a hepatic vein centerline for each hepatic vein vessel branch in the connected component;
determining a plurality of hepatic vein nodes based on intersection points between hepatic vein central lines of different hepatic vein branches; and
and generating a hepatic vein vessel tree diagram corresponding to the connected component based on the plurality of hepatic vein nodes and each hepatic vein central line.
By adopting the method, one or more connected components in the hepatic vein detection image can be determined, the hepatic vein central line of each hepatic vein blood vessel branch in each connected component is determined, the hepatic vein central line is used for representing the hepatic vein blood vessel branches, and the hepatic vein blood vessel tree-like graph is determined based on the hepatic vein central line and hepatic vein nodes, so that the hepatic vein blood vessel tree-like graph can accurately represent the structural characteristics of the hepatic veins.
In one possible embodiment, generating a hepatic vein vessel tree diagram corresponding to the connected component based on the plurality of hepatic vein nodes and the respective hepatic vein centerlines includes:
acquiring a minimum distance value of each hepatic vein node from the contour region of the inferior vena cava in the first detection image;
determining hepatic vein root nodes corresponding to the connected components from the plurality of hepatic vein nodes based on the minimum distance value corresponding to each hepatic vein node and the position information of each hepatic vein node in the first detection image; and
and generating a hepatic vein vessel tree diagram corresponding to the connected component based on the hepatic vein root node corresponding to the determined connected component, other hepatic vein nodes except the hepatic vein root node in the plurality of hepatic vein nodes and the central line of each hepatic vein.
Here, based on the minimum distance value corresponding to each hepatic vein node and the position information of each hepatic vein node in the hepatic vein detection image, the hepatic vein root node corresponding to the connected component can be determined more accurately from the plurality of hepatic vein nodes, so that after the hepatic vein vessel tree diagram corresponding to the connected component is generated based on the hepatic vein root node corresponding to the determined connected component, other hepatic vein nodes except the hepatic vein root node in the plurality of hepatic vein nodes, and each hepatic vein center line, the hepatic vein detection image can be more accurately hierarchically divided based on the generated hepatic vein vessel tree diagram.
In one possible embodiment, determining a category of each centerline characterizing a blood vessel branch in the first structural map based on the first structural map and the second detection image further includes:
dividing a plurality of liver segments included in the second detection image according to a first liver region dividing mode to generate a first liver segmentation image containing different liver region contour regions; and
and determining a first category of each hepatic vein central line in the hepatic vein vessel tree diagram based on the hepatic vein vessel tree diagram and the first liver segmentation image, wherein the first category is used for indicating a liver region to which a hepatic vein vessel branch corresponding to the hepatic vein central line belongs.
By adopting the method, the plurality of liver segments included in the second detection image are divided according to the first liver region division mode to generate the first liver segmentation image containing different liver region outline regions, and data support is provided for subsequently determining the first category of each hepatic vein central line in the hepatic vein blood vessel tree-shaped image.
In one possible embodiment, determining a first category of each hepatic vein centerline in the hepatic vein vessel tree based on the hepatic vein vessel tree and the first liver segmentation image includes:
taking the hepatic vein root node as a target hepatic vein node, and calculating a first proportion of the lengths of center lines respectively positioned in different hepatic region contour areas in at least one hepatic vein center line corresponding to the target hepatic vein node in the hepatic vein center line based on the hepatic vein vessel dendrogram and the first liver segmentation image, wherein the first proportion accounts for the total length of the hepatic vein center line corresponding to the target hepatic vein node;
under the condition that the maximum proportion in the first proportions corresponding to different hepatic region contour areas is larger than or equal to a target proportion threshold corresponding to the target hepatic vein node, determining a first category of at least one hepatic vein central line under the target hepatic vein node based on the hepatic region corresponding to the maximum proportion; and
and under the condition that the first ratios corresponding to different hepatic region contour areas are smaller than the determined target ratio threshold corresponding to the target hepatic vein node, taking each hepatic vein node in the hepatic vein nodes which are connected with the target hepatic vein node and are not traversed as a new target hepatic vein node, and returning to the step of calculating the first ratio until the first category of each hepatic vein central line in the hepatic vein blood vessel tree graph is determined.
By adopting the method, the first proportion of the lengths of the central lines respectively positioned in different hepatic region contour areas in at least one hepatic vein central line under each target hepatic vein node is determined, and the first category of each hepatic vein central line can be accurately determined based on the first proportion of the lengths of the central lines; for example, when the first ratio corresponding to the different hepatic region contour regions is smaller than the target ratio threshold corresponding to the determined target hepatic vein node, at least one hepatic vein centerline under the target hepatic vein node includes hepatic vein centerlines located in the different hepatic region contour regions, and at this time, a new target hepatic vein node needs to be determined.
In one possible embodiment, the target proportion threshold corresponding to the target hepatic vein node is determined according to the following steps:
under the condition that the minimum distance value corresponding to the target hepatic vein node is smaller than a set distance threshold value for dividing the hepatic vein node, determining a target proportion threshold value corresponding to the target hepatic vein node based on the minimum distance value corresponding to the target hepatic vein node; and
and under the condition that the minimum distance value corresponding to the target hepatic vein node is greater than or equal to a set distance threshold, determining a preset proportion threshold as the target proportion threshold corresponding to the target hepatic vein node.
Because the probability that at least one hepatic vein blood vessel branch under the target hepatic vein node is located in different hepatic region contour regions is high when the minimum distance value corresponding to the target hepatic vein node is small, different target proportion thresholds can be set for the target hepatic vein nodes with different minimum distance values, so that the target proportion threshold corresponding to the target hepatic vein node with a small minimum distance value is large, and the accuracy of determining the first category of each hepatic vein central line is improved.
In a possible embodiment, in a case where the first detection image comprises a portal vein detection image, generating a first structural map of a vascular structure in the first target region based on the first detection image comprises:
determining a portal vein center line of each portal vein blood vessel branch in the first detection image;
determining a plurality of portal vein nodes based on intersection points between the portal vein centerlines of different portal vein branches;
determining a portal vein root node from the plurality of portal vein nodes based on the location information of each portal vein node in the first detected image; and
and generating a portal vein dendriographic graph based on the portal vein root node, portal vein nodes except the portal vein root node in the portal vein nodes and all portal vein central lines.
By adopting the method, the portal vein central line of each portal vein branch in the first detection image can be determined, the portal vein central line is used for representing the portal vein branches, the portal vein root node is determined, and the portal vein arborescence map is determined based on the determined portal vein central line, the portal vein root node and other portal vein nodes, so that the portal vein arborescence map can accurately represent the structural characteristics of the portal vein.
In one possible embodiment, determining a category of each centerline characterizing a vessel branch in the first structural map based on the first structural map and the second detection image further includes:
dividing a plurality of liver segments included in the second detection image according to a second liver region dividing mode to generate a second liver segmentation image containing different liver region contour regions; and
and determining a second category of each portal vein central line in the portal vein tree graph based on the portal vein tree graph and the second liver segmentation image, wherein the second category is used for indicating a liver region to which the portal vein branch corresponding to the portal vein central line belongs.
By adopting the method, the liver segments in the second detection image are divided according to a second liver region dividing mode to generate a second liver segmentation image containing different liver region contour regions, and data support is provided for subsequently determining the second category of each portal vein central line in the portal vein blood vessel tree diagram.
In a possible embodiment, the second category includes: the main portal vein blood vessel branch, the left main branch blood vessel, the left internal blood vessel and the left external blood vessel corresponding to the left portal vein blood vessel branch, and the right main branch blood vessel, the right front blood vessel and the right back blood vessel corresponding to the right portal vein blood vessel branch; and
the determining a second category of each portal vein centerline in the portal vein tree based on the portal vein tree and the second liver segmentation image comprises:
dividing the portal vein dendrogram into a portal vein left blood vessel dendrogram, a portal vein right blood vessel dendrogram and a portal vein main central line corresponding to the portal vein main blood vessel branch based on the portal vein dendrogram and the second liver segmentation image;
determining a second category of each portal vein left centerline in the portal vein left vessel dendrogram based on the portal vein left vessel dendrogram and the second liver segmentation image; and
determining a second category of each portal vein right center line in the portal vein right vessel tree diagram based on the portal vein right vessel tree diagram and the second liver segmentation image.
By adopting the method, according to the structural characteristics of the portal vein, the portal vein tree can be divided into a portal vein left blood vessel tree diagram, a portal vein right blood vessel tree diagram and a portal vein main central line corresponding to the portal vein main blood vessel branch; and determining a second category of each portal vein left center line in the portal vein left blood vessel tree diagram and a second category of each portal vein right center line in the portal vein right blood vessel tree diagram, so as to realize the hierarchical division of the portal vein detection image.
In one possible embodiment, the dividing the portal vein dendrogram into a portal vein left blood vessel dendrogram, a portal vein right blood vessel dendrogram and a portal vein main central line corresponding to the portal vein main blood vessel branch based on the portal vein dendrogram and the second liver segmentation image includes:
determining the portal vein central line from the portal vein root node to the outside of the liver region in the portal vein tree graph as the portal vein main central line corresponding to the portal vein main vessel branch;
determining the portal vein root node as a target portal vein node, and calculating a second proportion of the lengths of central lines respectively positioned in different hepatic region contour areas in at least one portal vein central line corresponding to the target portal vein node in the total length of the portal vein central line corresponding to the target portal vein node on the basis of the portal vein vessel dendrogram and the second liver segmentation image;
under the condition that the maximum proportion in the second proportions corresponding to different hepatic region contour areas is larger than or equal to a set first proportion threshold, determining an intermediate class of at least one portal vein central line under the target portal vein node based on the hepatic region corresponding to the maximum proportion, wherein the intermediate class comprises the portal vein left blood vessel branch and the portal vein right blood vessel branch;
under the condition that the second proportion corresponding to different hepatic region contour areas is smaller than a set first proportion threshold, taking each portal vein node in at least one non-traversed portal vein node connected with the target portal vein node as a target portal vein node respectively, and returning to the step of calculating the second proportion until the middle category of each portal vein center line in the portal vein blood vessel dendrogram is determined; and
and dividing the portal vein centerlines except the main portal vein centerline in the portal vein dendrogram into a portal vein left blood vessel dendrogram and a portal vein right blood vessel dendrogram based on the middle category of each portal vein centerline.
In one possible embodiment, the determining the second category of each portal vein left center line in the portal vein left blood vessel tree based on the portal vein left blood vessel tree and the second liver segmentation image includes:
determining a left main branch central line corresponding to a left main branch blood vessel in the portal vein left blood vessel tree diagram based on the portal vein left blood vessel tree diagram and the second liver segmentation image; and
and determining a second category of the left central lines of other portal veins except the left main branch central line in the portal left blood vessel tree diagram based on the left main branch central line corresponding to the left main branch blood vessel in the portal left blood vessel tree diagram, the portal left blood vessel tree diagram and the second liver segmentation image.
Here, the left main branch centerline corresponding to the left main branch vessel in the portal left vessel tree-like map may be determined first, the second category of the left centerline of the portal vein other than the left main branch centerline in the portal left vessel tree-like map is determined based on the left main branch centerline, the portal left vessel tree-like map, and the second liver segmentation image corresponding to the left main branch vessel in the portal left vessel tree-like map, and when the second category of each left centerline of the portal vein in the portal left vessel tree-like map is determined, the structural characteristics of the left vessel of the portal vein are combined, so that the accuracy of the second category of each left centerline of the portal vein in the portal left vessel tree-like map is determined to be higher.
In a possible embodiment, the determining, based on the portal left blood vessel tree map and the second liver segmentation image, a left main branch centerline corresponding to a left main branch blood vessel in the portal left blood vessel tree map includes:
determining a first intersection point between the portal vein left blood vessel dendrogram and the portal vein right blood vessel dendrogram in the portal vein dendrogram;
determining at least one candidate portal vein left node which is connected with the target portal vein left node and is not traversed in the portal vein left blood vessel tree diagram by taking the first intersection as the target portal vein left node;
determining whether each candidate portal vein left node is an undetermined portal vein left node or not based on the portal vein left blood vessel tree diagram and the second liver segmentation image, wherein at least one portal vein left center line connected with the undetermined portal vein left node comprises a portal vein center line belonging to a left inner liver region and a portal vein center line belonging to a left outer liver region;
under the condition that the number of the left nodes of the portal vein to be determined is multiple, determining a first length of a portal vein central line belonging to a left inner hepatic region and a second length of a portal vein central line belonging to a left outer hepatic region in at least one portal vein left central line connected with each left node of the portal vein to be determined, and determining the smaller length of the first length and the second length as a target length corresponding to the left node of the portal vein to be determined;
selecting an undetermined portal vein left node with the longest corresponding target length from a plurality of undetermined portal vein left nodes, taking the selected undetermined portal vein left node as an updated target portal vein left node, and returning to the step of determining at least one unexplored candidate portal vein left node connected with the target portal vein left node in the portal vein left blood vessel tree graph until the undetermined portal vein left node does not exist in the at least one candidate portal vein left node connected with the target portal vein left node or until the unexplored candidate portal vein left node connected with the target portal vein left node does not exist; and
and determining a left main branch central line corresponding to the left main branch vessel of the portal vein left vessel tree diagram based on each target portal vein left node.
In one possible embodiment, determining whether each candidate portal vein left node is a pending portal vein left node based on the portal vein left blood vessel tree map and the second liver segmentation image includes:
for each candidate portal vein left node, determining a first left inner length proportion of the length of a central line belonging to a left inner blood vessel in at least one portal vein left central line corresponding to the candidate portal vein left node in the total length of the portal vein left central line corresponding to the candidate portal vein left node, and a first left outer length proportion of the length of a central line belonging to a left outer blood vessel in the total length of the portal vein left central line corresponding to the candidate portal vein left node; and
and under the condition that the smaller length proportion of the first left inner length proportion and the first left outer length proportion is larger than a set second proportion threshold, determining the candidate portal vein left node as the pending portal vein left node.
Here, considering that the second detected image may have an error, so that the obtained second liver segmentation image may also have an error, a second proportion threshold is set, and when the minimum length proportion of at least one portal vein left center line of the candidate portal vein left node belongs to the length proportion of the left internal blood vessel and the length proportion of the left external blood vessel is less than or equal to the set second proportion threshold, it is determined that the candidate portal vein left node includes blood vessel branches in different liver regions due to the error of the second liver segmentation image, and at this time, the candidate portal vein left node may be screened out.
In one possible embodiment, the determining, based on the left main branch centerline corresponding to the left main branch vessel in the portal left vessel tree diagram, and the second liver segmentation image, a second category of portal left vessel left centerlines in the portal left vessel tree diagram except for the left main branch centerline includes:
determining at least one left branch vessel tree connected with a left main branch vessel in the portal left vessel tree diagram based on a left main branch center line corresponding to the left main branch vessel and the portal left vessel tree diagram;
for each left branch vessel tree, determining a second left inner length proportion of the left central line length of the portal vein belonging to the left inner vessel in the left branch vessel tree to the total length of the left branch vessel tree and a second left outer length proportion of the left central line length of the portal vein belonging to the left outer vessel to the total length of the left branch vessel tree based on the second liver segmentation image; and
and determining a second category corresponding to the maximum length ratio in the second left inner length ratio and the second left outer length ratio as a second category of the left central line of each portal vein in the left branch vessel tree.
In one possible embodiment, determining the second category of each portal vein centerline in the portal vein right vessel tree based on the portal vein right vessel tree and the second liver segmentation image includes:
determining a first intersection point between the portal vein left blood vessel dendrogram and the portal vein right blood vessel dendrogram in the portal vein dendrogram;
determining the first intersection point as a target portal vein right node, and calculating a right anterior length proportion of the length of the center line in the right anterior hepatic region outline area to the total length of the portal vein right center line corresponding to the target portal vein right node and a right posterior length proportion of the length of the center line in the right posterior hepatic region outline area to the total length of the portal vein right center line corresponding to the target portal vein right node in at least one portal vein right center line corresponding to the target portal vein right node based on the portal vein right vessel dendrogram and the second liver segmentation image;
under the condition that the maximum length proportion of the front right length proportion and the rear right length proportion is larger than or equal to a set third proportion threshold, determining a second category of at least one portal vein right center line under the target portal vein right node based on the liver region category corresponding to the maximum length proportion;
and under the condition that the maximum length proportion in the right front length proportion and the right rear length proportion is smaller than a set third proportion threshold, taking each portal vein node in at least one non-traversed portal vein node which is connected with the target portal vein right node in the portal vein right vessel dendrogram as the target portal vein right node, and returning to the step of calculating the right front length proportion and the right rear length proportion until the second category of each portal vein right center line in the portal vein right vessel dendrogram is determined.
In a possible embodiment, determining a second category of each portal vein right center line in the portal vein right vessel tree map based on the portal vein right vessel tree map and the second liver segmentation image further includes:
determining a second intersection point of a right front blood vessel and a right back blood vessel in the portal vein right blood vessel tree diagram from each target portal vein right node;
and determining a right main branch central line corresponding to the right main branch vessel in the portal vein right vessel tree diagram based on the second intersection point and the first intersection point.
In one possible embodiment, the step of performing hierarchical division processing on the image of the first target area based on the category of the center line and the first detection image to obtain a target image includes:
for each pixel point to be classified in the first target region of the first detection image, determining a target pixel point with the minimum distance to the pixel point to be classified on the first structural graph; determining the category of the center line to which the target pixel point belongs as the category corresponding to the pixel point to be classified;
and carrying out hierarchical division processing on the image of the first target area based on the category corresponding to each pixel point to be classified respectively to obtain a target image.
By adopting the method, aiming at each pixel point to be classified in the first target area of the first detection image, the target pixel point with the minimum distance to the pixel point to be classified on the first structural diagram is determined, the category corresponding to the pixel point to be classified is determined based on the category of the central line to which the target pixel point belongs, and the target image subjected to hierarchical division processing is generated based on the categories respectively corresponding to the pixel points to be classified, so that automatic hierarchical division of the first detection image is realized, and compared with the manual labeling process, the efficiency of the division is improved.
The following descriptions of the effects of the apparatus, the electronic device, and the like refer to the description of the above method, and are not repeated here.
In a second aspect, the present disclosure provides an image processing apparatus comprising:
the device comprises an acquisition module, a detection module and a display module, wherein the acquisition module is used for acquiring at least one detection image of a target object, and the detection image comprises a first detection image of a first target area of the target object and a second detection image of a second target area of the target object;
a generating module for generating a first structural map of a vessel structure in the first target region based on the first detection image;
a determination module for determining a category of each centerline characterizing a vessel branch in the first map based on the first map and the second detected image;
and the dividing module is used for carrying out hierarchical dividing processing on the image of the first target area based on the category of the central line and the first detection image to obtain a target image.
In a third aspect, the present disclosure provides an electronic device comprising: a processor, a memory and a bus, the memory storing machine-readable instructions executable by the processor, the processor and the memory communicating via the bus when the electronic device is running, the machine-readable instructions when executed by the processor performing the steps of the image processing method according to the first aspect or any of the embodiments.
In a fourth aspect, the present disclosure provides a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, performs the steps of the image processing method according to the first aspect or any one of the embodiments.
In order to make the aforementioned objects, features and advantages of the present disclosure more comprehensible, preferred embodiments accompanied with figures are described in detail below.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present disclosure, the drawings required for use in the embodiments will be briefly described below, and the drawings herein incorporated in and forming a part of the specification illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the technical solutions of the present disclosure. It is appreciated that the following drawings depict only certain embodiments of the disclosure and are therefore not to be considered limiting of its scope, for those skilled in the art will be able to derive additional related drawings therefrom without the benefit of the inventive faculty.
Fig. 1 shows a schematic flow chart of an image processing method provided by an embodiment of the present disclosure;
fig. 2 is a schematic diagram illustrating a first liver segmentation image in an image processing method provided by an embodiment of the present disclosure;
fig. 3 is a schematic diagram illustrating a hepatic vein tree diagram in an image processing method provided by an embodiment of the disclosure;
fig. 4 is a schematic flowchart illustrating a process of generating a target image after hierarchical division processing corresponding to a hepatic vein detection image in an image processing method provided by an embodiment of the present disclosure;
fig. 5 is a schematic flowchart illustrating a process of generating a target image after hierarchical division processing corresponding to a portal vein detection image in an image processing method provided by an embodiment of the present disclosure;
fig. 6 is a schematic diagram illustrating an architecture of an image processing apparatus provided in an embodiment of the present disclosure;
fig. 7 shows a schematic structural diagram of an electronic device 700 provided in an embodiment of the present disclosure.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present disclosure more clear, the technical solutions of the embodiments of the present disclosure will be described clearly and completely with reference to the drawings in the embodiments of the present disclosure, and it is obvious that the described embodiments are only a part of the embodiments of the present disclosure, not all of the embodiments. The components of the embodiments of the present disclosure, generally described and illustrated in the figures herein, can be arranged and designed in a wide variety of different configurations. Thus, the following detailed description of the embodiments of the present disclosure, presented in the figures, is not intended to limit the scope of the claimed disclosure, but is merely representative of selected embodiments of the disclosure. All other embodiments, which can be derived by a person skilled in the art from the embodiments of the disclosure without making creative efforts, shall fall within the protection scope of the disclosure.
Generally, precise hepatectomy depends on accurate segmentation results of hepatic veins, hierarchical division results of hepatic veins, and the like. For example, the segmentation result and the hierarchical segmentation result of the hepatic vein may be obtained by a manual labeling manner. However, when the result of the hierarchical division of the hepatic vein is obtained by the manual labeling, the efficiency is low, and the manual labeling is performed according to the experience of the operator, so that a labeling error is likely to occur. Therefore, in order to solve the above problem, an embodiment of the present disclosure provides an image processing method.
For the purpose of understanding the embodiments of the present disclosure, an image processing method disclosed in the embodiments of the present disclosure will be described in detail first.
Referring to fig. 1, a schematic flow chart of an image processing method provided in the embodiment of the present disclosure is shown, the method includes S101-S104, where:
s101, at least one detection image of the target object is obtained, and the detection image comprises a first detection image of a first target area of the target object and a second detection image of a second target area of the target object.
S102, generating a first structural diagram of the blood vessel structure in the first target area based on the first detection image.
S103, based on the first structural image and the second detection image, a category of each central line representing a blood vessel branch in the first structural image is determined.
And S104, carrying out hierarchical division processing on the image of the first target area based on the category of the central line and the first detection image to obtain a target image.
With the above method, by generating a first structural map of the blood vessel structure in the first target region, based on the first structural map and the second detection image, the category of each centerline characterizing a blood vessel branch in the first structural map can be determined; the image of the first target area is subjected to hierarchical division processing based on the category of the central line and the first detection image to obtain a target image, so that automatic division of the first detection image is realized, and the efficiency of dividing the first detection image is improved; and by constructing the first structural diagram, manual marking is not needed, and marking errors caused by manual marking can be reduced.
In S101, the first target region of the target object may be a vein contour region, and in the liver, the vein contour region includes a portal vein contour region and a hepatic vein contour region, that is, the first detection image may be a vein detection image, and the first detection image may include: a hepatic vein detection image representing hepatic vein contour information, and/or a portal vein detection image representing portal vein contour information.
The second target region can be the whole liver region, the second detection image comprises profile information of each liver segment, namely the second detection image can be a liver segment detection image, wherein each liver segment in the liver is an existing liver division result, namely the liver is divided into 8 independent segments according to functions, and therefore the liver segments comprise a first liver segment, a second liver segment, a third liver segment, a fourth liver segment, a fifth liver segment, a sixth liver segment, a VII liver segment and a VIII liver segment. Among them, the hepatic vein detection image, the portal vein detection image, and the second detection image may be three-dimensional images, for example, the three-dimensional images may be Computed Tomography (CT) images, magnetic resonance images, or the like.
Illustratively, a three-dimensional liver image of a target object can be acquired, and the three-dimensional liver image is input into a trained first neural network to obtain a three-dimensional liver segment detection image; and inputting the three-dimensional liver image into the trained second neural network to obtain a three-dimensional vein detection image, wherein the three-dimensional vein detection image comprises a three-dimensional portal vein detection image and a three-dimensional hepatic vein detection image. The manner of obtaining the multiple detection images may be multiple, and this is merely an exemplary illustration. The first neural network is a trained first segmentation network used for detecting a liver segment contour region corresponding to the liver, and the second neural network is a trained second segmentation network used for detecting a contour region of a hepatic vein blood vessel branch and/or a portal vein blood vessel branch in the liver.
For example, the first neural network may be trained in multiple rounds using a first training sample labeled with a liver segment contour region until the trained first neural network satisfies a preset condition. And performing multiple rounds of training on the second neural network by using a second training sample marked with the hepatic vein vessel branch and/or the portal vein vessel branch until the trained second neural network meets the preset condition. Wherein the preset condition may be that the accuracy is greater than a set accuracy threshold; alternatively, the preset condition may be that the loss value is smaller than a set loss threshold value, or the like. The accuracy threshold, or loss threshold, may be set as desired.
In specific implementation, the hepatic vein detection image can be subjected to hierarchical division to obtain a target image corresponding to the hepatic vein of the target object; and carrying out hierarchy division on the portal vein detection image to obtain a target image corresponding to the portal vein of the target object.
The following may be specifically described with reference to S102 to S104, which first performs hierarchical division on the hepatic vein detection image to obtain a target image corresponding to the hepatic vein of the target object.
Here, the first structural diagram is a dendrogram characterizing the structure of venous blood vessels. When the first inspection image comprises a hepatic vein inspection image, the first structural map comprises a hepatic vein vessel dendrogram characterizing a hepatic vein vessel structure.
Wherein, in a case where the first detection image includes a hepatic vein detection image, generating a first structural map of a vascular structure in the first target region based on the first detection image includes:
a1, determining a connected component included in the first detected image based on a connection relationship between the hepatic vein and the inferior vena cava in the first detected image. Wherein, the connected component comprises a plurality of hepatic vein branches with connection relation.
And A2, determining the hepatic vein central line of each hepatic vein blood vessel branch in the connected component.
A3, determining a plurality of hepatic vein nodes based on the intersection points between the hepatic vein centerlines of different hepatic vein vessel branches.
And A4, generating a hepatic vein blood vessel tree diagram corresponding to the connected components based on the plurality of hepatic vein nodes and the hepatic vein central lines.
By adopting the method, the connected components in the hepatic vein detection image can be determined, the hepatic vein central line of each hepatic vein blood vessel branch in each connected component is determined, the hepatic vein central line is used for representing the hepatic vein blood vessel branches, and the hepatic vein blood vessel tree-like graph is determined based on the hepatic vein central line and hepatic vein nodes, so that the hepatic vein blood vessel tree-like graph can accurately represent the structural characteristics of the hepatic veins.
In step a1, each connected component has at least one connection point with the inferior vena cava, and for example, after the contour region belonging to the inferior vena cava in the first detection image (hepatic vein detection image) is removed, the hepatic vein vessel branches having a connection relationship in the first detection image constitute one connected component.
In step a2, when the connected component is plural, the hepatic vein center line of each hepatic vein blood vessel branch in the connected component is extracted for each connected component. There are various methods for determining the hepatic vein center line of the hepatic vein branch, and a detailed description thereof will not be provided herein. For example, the method for extracting the center line may be a distance transformation algorithm, a shortest path algorithm, or the like.
In step a3, intersections between hepatic vein centerlines of different hepatic vein vessel branches may be determined, and the determined intersections between the different hepatic vein centerlines may be determined as a plurality of hepatic vein nodes.
In step a4, generating a hepatic vein vessel tree diagram corresponding to connected components based on the plurality of hepatic vein nodes and the respective hepatic vein centerlines, which may include:
and B1, acquiring the minimum distance value of each hepatic vein node from the contour region of the inferior vena cava in the first detection image.
And B2, determining a hepatic vein root node corresponding to the connected component from the plurality of hepatic vein nodes based on the minimum distance value corresponding to each hepatic vein node and the position information of each hepatic vein node in the first detection image.
And B3, generating a hepatic vein vessel tree diagram corresponding to the connected component based on the hepatic vein root node corresponding to the determined connected component, other hepatic vein nodes except the hepatic vein root node in the plurality of hepatic vein nodes and the central line of each hepatic vein.
In step B1, for example, a minimum distance value of each hepatic vein node from the contour region of the inferior vena cava in the hepatic vein detection image may be calculated by a distance transformation algorithm.
In step B2, a first distance threshold may be set, and at least one candidate hepatic vein node having a minimum distance value smaller than the set first distance threshold is selected from the plurality of hepatic vein nodes; wherein the first distance threshold may be an empirical value determined from a plurality of hepatic vein detection samples. And determining the hepatic vein root node of the connected component from at least one candidate hepatic vein node based on the position information of each candidate hepatic vein node in the hepatic vein detection image. For example, the hepatic vein candidate node whose position information is located at the top in the hepatic vein detection image (i.e., the hepatic vein node close to the head of the target object corresponding to the detection image) may be selected as the hepatic vein root node of the connected component.
In step B3, a hepatic vein vessel tree diagram corresponding to the connected component may be generated according to the hepatic vein root node of the determined connected component, other hepatic vein nodes except the hepatic vein root node among the plurality of hepatic vein nodes, and each hepatic vein centerline.
Here, based on the minimum distance value corresponding to each hepatic vein node and the position information of each hepatic vein node in the hepatic vein detection image, the hepatic vein root node corresponding to the connected component can be determined more accurately from the plurality of hepatic vein nodes, so that after the hepatic vein vessel tree diagram corresponding to the connected component is generated based on the hepatic vein root node corresponding to the determined connected component, other hepatic vein nodes except the hepatic vein root node in the plurality of hepatic vein nodes, and each hepatic vein center line, the hepatic vein detection image can be more accurately hierarchically divided based on the generated hepatic vein vessel tree diagram.
In an alternative embodiment, determining a category of each centerline characterizing a vessel branch in the first map based on the first map and the second detected image comprises:
step one, dividing a plurality of liver segments included in the second detection image according to a first liver region dividing mode to generate a first liver segmentation image containing different liver region contour regions.
And secondly, determining a first category of each hepatic vein central line in the hepatic vein tree graph based on the hepatic vein tree graph and the first liver segmentation image, wherein the first category is used for indicating a liver region to which the hepatic vein branch corresponding to the hepatic vein central line belongs.
Here, the plurality of liver segments included in the second detection image may be divided in accordance with a set first liver region division manner, and a first liver segmentation image including different liver region contour regions may be generated.
Referring to fig. 2, a contour region composed of a second liver segment ii and a third liver segment iii included in the second detection image a may be determined as a first region 21; determining a contour region formed by the first hepatic segment and the fourth hepatic segment as a second region 22; determining a contour region formed by the hepatic segment V and the hepatic segment VIII as a third region 23; the contour region formed by the VI liver segment and the VII liver segment is defined as the fourth region 24. Fig. 2 is a schematic diagram of a contour of a liver segment, and a contour region of the liver segment may be determined according to an actual situation.
Then, a first minimum distance value from each pixel point in the second region to the first region and a second minimum distance value from each pixel point in the second region to the third region may be calculated through a distance transformation algorithm, the pixel points with the same first minimum distance value and second minimum distance value are connected to obtain a first bisection plane 25 corresponding to the second region, and the first bisection plane corresponding to the second region divides the second region into two regions, namely, a second first region and a second region.
And calculating a third minimum distance value from each pixel point in the third region to the second region and a fourth minimum distance value from each pixel point in the third region to the fourth region through a distance transformation algorithm, connecting the pixel points with the same third minimum distance value and fourth minimum distance value to obtain a second bisection plane 26 corresponding to the third region, and dividing the third region into two regions by the second bisection plane corresponding to the third region, namely a third region and a third second region.
And combining a second region adjacent to the first region in the second region with the first region to obtain a left liver region outline region. Combining a second region of the second region vector with a third region vector and a third region of the third region vector adjacent to the second region to obtain a contour region of the middle liver region; and combining a third region adjacent to the fourth region in the third region to obtain a right liver region outline region. And obtaining a first liver segmentation image b comprising a left liver region outline region, a middle liver region outline region and a right liver region outline region.
Here, the process of generating the first liver segmentation image may be performed at any step before the first category of each hepatic vein centerline in the hepatic vein vessel tree is determined. For example, the process of generating the first liver segmentation image may be performed before or after the hepatic vein vessel tree map is generated.
And determining a first category of each hepatic vein central line in the hepatic vein blood vessel tree diagram based on the hepatic vein blood vessel tree diagram and the first liver segmentation image, wherein the first category is used for indicating a hepatic region to which the hepatic vein blood vessel branch corresponding to the hepatic vein central line belongs. Wherein the first category includes left hepatic vein vessels, middle hepatic vein vessels, and right hepatic vein vessels. For example, if the category of the hepatic vein centerline is left hepatic vein blood vessel, it indicates that the blood vessel branch corresponding to the centerline belongs to the left hepatic region.
By adopting the method, the plurality of liver segments included in the second detection image are divided according to the first liver region division mode to generate the first liver segmentation image containing different liver region outline regions, and data support is provided for subsequently determining the first category of each hepatic vein central line in the hepatic vein blood vessel tree-shaped image.
In an alternative embodiment, determining a first category of each hepatic vein centerline in the hepatic vein vessel tree based on the hepatic vein vessel tree and the first liver segmentation image may include:
the method comprises the steps of taking a hepatic vein root node as a target hepatic vein node, and calculating a first proportion of the length of central lines respectively located in different hepatic region contour areas in at least one hepatic vein central line corresponding to the target hepatic vein node in the total length of the hepatic vein central line corresponding to the target hepatic vein node based on a hepatic vein blood vessel dendrogram and a first liver segmentation image.
And secondly, under the condition that the maximum proportion in the first proportions corresponding to different hepatic region contour areas is larger than or equal to a target proportion threshold corresponding to the determined target hepatic vein node, determining a first category of at least one hepatic vein central line under the target hepatic vein node based on the hepatic region corresponding to the maximum proportion.
And thirdly, under the condition that the first proportion corresponding to different hepatic region contour regions is smaller than the target proportion threshold corresponding to the determined target hepatic vein node, taking each hepatic vein node in at least one hepatic vein node which is connected with the target hepatic vein node and is not traversed as a new target hepatic vein node, and returning to the step of calculating the first proportion until the first class of each hepatic vein centerline in the hepatic vein blood vessel tree graph is determined.
By adopting the method, the first proportion of the lengths of the central lines respectively positioned in different hepatic region contour areas in at least one hepatic vein central line under each target hepatic vein node is determined, and the first category of each hepatic vein central line can be accurately determined based on the first proportion of the lengths of the central lines; for example, when the first ratio corresponding to the different hepatic region contour regions is smaller than the target ratio threshold corresponding to the determined target hepatic vein node, at least one hepatic vein centerline under the target hepatic vein node includes hepatic vein centerlines located in the different hepatic region contour regions, and at this time, a new target hepatic vein node needs to be determined.
Illustratively, a search algorithm of breadth-first traversal may be used to sequentially traverse the hepatic vein vessel tree map, starting from the hepatic vein root node in the hepatic vein vessel tree map, to determine the first category of each hepatic vein centerline.
During specific implementation, firstly, a hepatic vein root node is determined as a target hepatic vein node, based on a hepatic vein blood vessel dendrogram and a first liver segmentation image, the total length of at least one hepatic vein central line corresponding to the target hepatic vein node is determined, in the at least one hepatic vein central line corresponding to the target hepatic vein node, the length of a central line in a left hepatic region outline region accounts for a first proportion of the total length corresponding to the target hepatic vein node, the length of a central line in a middle hepatic region outline region accounts for a first proportion of the total length corresponding to the target hepatic vein node, and the length of a central line in a right hepatic region outline region accounts for a first proportion of the total length corresponding to the target hepatic vein node.
The hepatic vein central line corresponding to the target hepatic vein node is the hepatic vein central line corresponding to one side, which is far away from the hepatic vein root node and has a connection relation with the target hepatic vein node, in the hepatic vein blood vessel tree diagram.
And when the maximum ratio among the first ratio of the length of the central line in the contour area of the left hepatic region, the first ratio of the length of the central line in the contour area of the middle hepatic region and the first ratio of the length of the central line in the contour area of the right hepatic region is greater than or equal to the determined target ratio threshold, determining the first category of each hepatic vein central line under the target hepatic vein node based on the hepatic region corresponding to the maximum ratio.
For example, if the first ratio of the length of the central line located in the contour region of the middle hepatic region is the maximum ratio, and the maximum ratio is greater than or equal to the target ratio threshold corresponding to the determined target hepatic vein node, the first category of each hepatic vein central line under the target hepatic vein node is determined to be the left hepatic vein.
And when the maximum proportion among the first proportion of the length of the central line in the contour area of the left hepatic region, the first proportion of the length of the central line in the contour area of the middle hepatic region and the first proportion of the length of the central line in the contour area of the right hepatic region is smaller than a determined target proportion threshold, respectively taking each hepatic vein node in at least one hepatic vein node which is connected with the target hepatic vein node and is not traversed as a new target hepatic vein node, determining at least one hepatic vein central line corresponding to the new target hepatic vein node, and respectively locating the first proportions of the lengths of the central lines in different hepatic region contour areas until the first category of each hepatic vein central line in the hepatic vein tree graph is determined.
Referring to fig. 3, a schematic diagram of a hepatic vein vessel tree diagram in an image processing method is shown, where the diagram includes a hepatic vein root node 31, hepatic vein nodes 32 except the hepatic vein root node 31 in a plurality of hepatic vein nodes, and hepatic vein centerlines 33.
With reference to fig. 3, a process of determining a first category of each hepatic vein centerline in the hepatic vein tree based on the hepatic vein tree and the first liver segmentation image will be briefly described. When the hepatic vein root node is used as a target hepatic vein node, a plurality of hepatic vein center lines below the hepatic vein root node are at least one hepatic vein center line corresponding to the hepatic vein root node, and then, in combination with a first liver segmentation image, a first proportion of the length of center lines respectively located in different hepatic region contour areas in the at least one hepatic vein center line below the target hepatic vein node to the total length of the hepatic vein center line corresponding to the target hepatic vein node is calculated.
And under the condition that the first proportion corresponding to different hepatic region contour areas is smaller than the target proportion threshold corresponding to the determined target hepatic vein node, taking each hepatic vein node in at least one hepatic vein node which is connected with the target hepatic vein node and is not traversed as a new target hepatic vein node, namely taking the hepatic vein node 321 and the hepatic vein node 322 in fig. 3 as new target hepatic vein nodes respectively.
Furthermore, the hepatic vein node 321 is used as a new target hepatic vein node, and based on the hepatic vein vessel dendrogram and the first liver segmentation image, a first proportion of the central line length of at least one hepatic vein central line corresponding to the hepatic vein node 321, which is respectively located in different hepatic region contour areas, to the total length of the hepatic vein central line corresponding to the target hepatic vein node (i.e., the hepatic vein node 321) is calculated.
And taking the hepatic vein node 322 as a new target hepatic vein node, and calculating a first proportion of the length of central lines respectively located in different hepatic region contour areas in at least one hepatic vein central line corresponding to the hepatic vein node 322 to the total length of the hepatic vein central line corresponding to the target hepatic vein node (namely the hepatic vein node 322) based on the hepatic vein vessel dendrogram and the first liver segmentation image. And determining a first category of hepatic vein central lines between the hepatic vein root node 31 and the hepatic vein node 321 by combining the first liver segmentation image; and determining a first category of hepatic vein centerlines between hepatic vein root node 31 and hepatic vein node 322. Therefore, the first category of each hepatic vein central line in the hepatic vein vessel tree diagram can be obtained through the process.
Generally, a first middle intersection point between the hepatic vein central line of the left hepatic region and the hepatic vein central line of the middle hepatic region, and a second intersection point between the hepatic vein central line of the middle hepatic region and the hepatic vein central line of the right hepatic region are closer to the inferior vena cava, that is, the minimum distance value between the first intersection point and the inferior vena cava is smaller, and the minimum distance value between the second intersection point and the inferior vena cava is smaller.
Therefore, if the minimum distance value corresponding to the target hepatic vein node is smaller, the probability that the target hepatic vein node belongs to the first intermediate intersection or the second intermediate intersection is higher, and at this time, the target proportion threshold corresponding to the target hepatic vein node may be set to be larger, so that the connected component may be further divided; if the minimum distance value corresponding to the target hepatic vein node is larger, the probability that the target hepatic vein node belongs to the first intermediate intersection or the second intermediate intersection is lower, and at this time, the target proportion threshold corresponding to the target hepatic vein node can be set to be smaller, so that the first category of the at least one hepatic vein central line corresponding to the target hepatic vein node can be directly determined.
Therefore, the size of the target proportion threshold is related to the minimum distance value corresponding to the target hepatic vein node, wherein the minimum distance value corresponding to the target hepatic vein node may be: and determining the minimum distance between the target hepatic vein node and the inferior vena cava through a distance transformation algorithm.
Further, a target proportion threshold corresponding to the target hepatic vein node may be determined according to the following steps:
in the first situation, when the minimum distance value corresponding to the target hepatic vein node is smaller than the set threshold value for dividing the distance of the hepatic vein node, the target proportion threshold value corresponding to the target hepatic vein node is determined based on the minimum distance value corresponding to the target hepatic vein node.
And secondly, under the condition that the minimum distance value corresponding to the target hepatic vein node is greater than or equal to the set distance threshold, determining the preset proportion threshold as the target proportion threshold corresponding to the target hepatic vein node.
Here, a minimum distance value corresponding to the target hepatic vein node may be determined first, and if the minimum distance value is smaller than a set distance threshold for dividing the hepatic vein node, a target proportional threshold corresponding to the target hepatic vein node may be determined based on the minimum distance value corresponding to the target hepatic vein node and a mapping relationship between the set distance and the threshold. For example, a relationship function between the distance and the threshold may be generated based on a plurality of hepatic vein detection samples; and determining a target proportion threshold corresponding to the target hepatic vein node based on the minimum distance value corresponding to the target hepatic vein node and the generated relation function.
Alternatively, the corresponding relationship between the distance and the threshold may also be set based on experience, for example, if the set distance threshold is 4cm, the corresponding relationship between the distance and the threshold may be: a threshold value corresponding between 4cm and 3cm (excluding 4cm and including 3cm) is 50%, a threshold value corresponding between 3cm and 2cm (excluding 3cm and including 2cm) is 60%, a threshold value corresponding between 2cm and 1cm (excluding 2cm and including 1cm) is 70%, and the like. And when the minimum distance value corresponding to the target hepatic vein node A is determined to be 1.5cm, determining that the target proportion threshold value corresponding to the target hepatic vein node A is 70%. The corresponding relationship between the distance and the threshold may be set according to actual needs, and is only an exemplary illustration here.
And when the minimum distance value corresponding to the target hepatic vein node is greater than or equal to the set distance threshold, determining the preset proportion threshold as the target proportion threshold corresponding to the target hepatic vein node. For example, the preset proportional threshold may be 30%.
Because the probability that at least one hepatic vein blood vessel branch under the target hepatic vein node is located in different hepatic region contour regions is high when the minimum distance value corresponding to the target hepatic vein node is small, different target proportion thresholds can be set for the target hepatic vein nodes with different minimum distance values, so that the target proportion threshold corresponding to the target hepatic vein node with a small minimum distance value is large, and the accuracy of determining the first category of each hepatic vein central line is improved.
After the first category of each hepatic vein centerline in the hepatic vein tree diagram with the connected component obtained, the target image after the hierarchical division processing may be generated based on the first category of each hepatic vein centerline in the hepatic vein tree diagram and a hepatic vein detection image (first detection image). The specific process can be as follows:
firstly, aiming at each first pixel point to be classified in a first target region of a first detection image (hepatic vein detection image), determining a target pixel point with the minimum distance from the first pixel point to be classified on a hepatic vein arborescence graph of connected components; and determining the category corresponding to the central line to which the target pixel point belongs as a first category corresponding to the first pixel point to be classified.
And secondly, carrying out hierarchical division processing on the image of the first target area based on the category corresponding to each first pixel point to be classified to obtain a target image.
Here, when the first detection image includes a hepatic vein detection image, the first target region is a hepatic vein segmentation region, and the pixel points to be classified may be first pixel points to be classified, that is, the minimum distance value between each first pixel point to be classified in the hepatic vein segmentation region of the first detection image and the hepatic vein vessel dendrogram may be calculated, and the target pixel point on the hepatic vein vessel dendrogram corresponding to the minimum distance value of the first pixel point to be classified is determined, that is, the target pixel point corresponding to each first pixel point to be classified is obtained. And determining the category corresponding to the center line to which the target pixel point belongs as a first category corresponding to the first pixel point to be classified. The first pixel point to be classified can be any pixel point located in the hepatic vein segmentation area in the hepatic vein detection image.
And then, based on the first category corresponding to each first pixel point to be classified, the image of the first target area is subjected to hierarchical division processing to generate a target image. The image of the first target region may be a contour image corresponding to the hepatic vein segmentation region in the first detection image.
For example, different colors may be used to label different first categories, for example, the color corresponding to the category of the left hepatic vein blood vessel may be red, the color corresponding to the category of the middle hepatic vein blood vessel may be green, and the color corresponding to the category of the right hepatic vein blood vessel may be blue, so that the pixel value corresponding to the first pixel value to be classified may be adjusted to the color pixel value corresponding to the category, and the generated target image includes the left hepatic vein blood vessel with the color of red, the middle hepatic vein blood vessel with the color of green, and the right hepatic vein blood vessel with the color of blue, thereby implementing hierarchical division of the hepatic vein detection image.
Referring to fig. 4, the image includes a hepatic vein detection image, the hepatic vein detection image includes inferior vena cava and hepatic vein, and a hepatic vein blood vessel tree diagram corresponding to each connected component on the hepatic vein detection image (first detection image) may be determined first; and the image also comprises a hepatic segment detection image (a second detection image), a plurality of hepatic segments included in the second detection image are divided according to a set first hepatic region division mode to generate a first hepatic segmentation image containing different hepatic region contour regions, a first category of each hepatic vein central line in the hepatic vein blood vessel tree image is determined based on the hepatic vein blood vessel tree image and the first hepatic segmentation image, and finally, a target image after hierarchical division processing is generated based on the first category of each hepatic vein central line in the hepatic vein blood vessel tree image.
The following may further describe, in conjunction with S102-S104, a process of performing hierarchical division on the portal vein detection image to obtain a target image corresponding to the portal vein of the target object.
Here, when the vein detection image includes a portal vein detection image, the first structural diagram includes a portal vein dendrogram representing a portal vein blood vessel structure.
Wherein generating a first map of structures of blood vessels in a first target region based on the first detected image comprises:
c1, determining the portal vein central line of each portal vein branch in the first detection image.
C2, determining a plurality of portal vein nodes based on the intersection points between the portal vein centerlines of the different portal vein vessel branches.
C3, determining a portal vein root node from the portal vein nodes based on the position information of each portal vein node in the first detection image.
C4, generating a portal vein vessel tree diagram based on the portal vein root node, portal vein nodes except the portal vein root node among the portal vein nodes, and the portal vein center lines.
By adopting the method, the portal vein central line of each portal vein branch in the first detection image can be determined, the portal vein central line is used for representing the portal vein branches, the portal vein root node is determined, and the portal vein arborescence map is determined based on the determined portal vein central line, the portal vein root node and other portal vein nodes, so that the portal vein arborescence map can accurately represent the structural characteristics of the portal vein.
Here, the portal vein center line of each portal vein blood vessel branch in the portal vein detection image may be extracted first. The portal vein centerline determination process may refer to the hepatic vein centerline determination process, which is not described herein. The intersection points between the central lines of different portal veins can be determined as a plurality of portal vein nodes. And determining a portal vein root node according to the position information of each portal vein node in the portal vein detection image. For example, a candidate portal vein node outside the liver contour region may be selected from a plurality of portal vein nodes; and based on the position information of the candidate portal vein node, selecting the candidate portal vein node which is positioned at the lowest position (namely, close to the direction of the target object foot corresponding to the detection image) in the portal vein detection image as the determined portal vein root node.
Further, a portal vein vessel dendrogram may be generated based on the determined portal vein root node, portal vein nodes other than the portal vein root node among the plurality of portal vein nodes, and the respective portal vein centerlines.
In an alternative embodiment, determining a category of each centerline characterizing a vessel branch in the first map based on the first map and the second detected image further comprises:
step one, dividing a plurality of liver segments included in a second detection image according to a second liver region dividing mode to generate a second liver segmentation image containing different liver region outline regions.
And secondly, determining a second category of each portal vein central line in the portal vein tree graph based on the portal vein tree graph and the second liver segmentation image, wherein the second category is used for indicating a portal vein liver area to which the portal vein branch corresponding to the portal vein central line belongs.
Here, the liver segments included in the second detection image may be divided according to a second liver region division manner, for example, liver segments i to iv may be divided into left liver regions, and liver segments v to viii may be divided into right liver regions; determining the region formed by the second liver segment and the third liver segment in the left liver region as a left outer liver region, and determining the region formed by the first liver segment and the fourth liver segment in the left liver region as a left inner liver region; determining the region formed by the liver segment V and the liver segment VIII in the right liver region as the right anterior liver region; the region composed of the VI liver segment and the VII liver segment in the right liver region is determined as the right posterior liver region. After liver segmentation, a second liver segmentation image was generated that contained the left liver region (left inner liver region and left outer liver region), and the right liver region (right front liver region and right rear liver region).
The step of generating a second liver segmentation image may be performed at any step prior to determining the second category for each portal vein centerline in the portal vein vessel dendrogram. For example, a second liver segmentation image may be determined prior to generating the portal vein dendrogram; alternatively, the second liver segmentation image may be determined after the portal vein dendrogram is generated.
And further, a second category of each portal vein central line in the portal vein tree graph can be determined based on the portal vein tree graph and the second liver segmentation image, wherein the second category is used for indicating a liver region to which the portal vein branch belongs.
By adopting the method, the liver segments in the second detection image are divided according to a second liver region dividing mode to generate a second liver segmentation image containing different liver region contour regions, and data support is provided for subsequently determining the second category of each portal vein central line in the portal vein blood vessel tree diagram.
In an alternative embodiment, the second category may include: the main portal vein blood vessel branch, the left main branch blood vessel, the left internal blood vessel and the left external blood vessel corresponding to the left portal vein blood vessel branch, and the right main branch blood vessel, the right front blood vessel and the right back blood vessel corresponding to the right portal vein blood vessel branch. Wherein, the portal vein main vessel branch is a vessel branch from the portal vein root node in the portal vein arborescence to the outside of the liver region; the portal vein main vessel branch is a vessel branch that does not belong to either the left or right liver region. The portal left vessel branch is a vessel branch located in the left liver region, and the portal right vessel branch is a vessel branch located in the right liver region.
In an alternative embodiment, determining the second category of each portal vein centerline in the portal vein vessel tree based on the portal vein vessel tree and the second liver segmentation image includes:
and D1, dividing the portal vein tree map into a portal vein left blood vessel tree map, a portal vein right blood vessel tree map and a portal vein main central line corresponding to the portal vein main blood vessel branch based on the portal vein tree map and the second liver segmentation image.
And D2, determining a second category of the left central line of each portal vein in the portal vein left blood vessel tree diagram based on the portal vein left blood vessel tree diagram and the second liver segmentation image.
And D3, determining a second category of the right central line of each portal vein in the portal vein right vessel tree diagram based on the portal vein right vessel tree diagram and the second liver segmentation image.
By adopting the method, according to the structural characteristics of the portal vein, the portal vein tree can be divided into a portal vein left blood vessel tree diagram, a portal vein right blood vessel tree diagram and a portal vein main central line corresponding to the portal vein main blood vessel branch; and determining a second category of each portal vein left center line in the portal vein left blood vessel tree diagram and a second category of each portal vein right center line in the portal vein right blood vessel tree diagram, so as to realize the hierarchical division of the portal vein detection image.
When determining the second category of each portal vein centerline in the portal vein tree map, the portal vein tree map may be divided into a portal vein left vessel tree map, a portal vein right vessel tree map, and a portal vein main centerline corresponding to a portal vein main vessel branch based on the portal vein tree map and the second liver segmentation image.
Further, the portal vein left blood vessel tree map and the portal vein right blood vessel tree map can be divided respectively, and the second category of each portal vein left central line in the portal vein left blood vessel tree map and the second category of each portal vein right central line in the portal vein right blood vessel tree map are determined.
The steps D1, D2, and D3 will be described in detail below.
For step D1:
based on the portal vein tree diagram and the second liver segmentation image, the portal vein tree diagram is divided into a portal vein left blood vessel tree diagram, a portal vein right blood vessel tree diagram and a portal vein main central line corresponding to the portal vein main blood vessel branch, which may include:
firstly, determining a portal vein central line from a portal vein root node in the portal vein tree diagram to the outside of the liver region as a portal vein main central line corresponding to the portal vein main vessel branch.
And secondly, determining the portal vein root node as a target portal vein node, and calculating a second proportion of the length of the central line in the contour region of different liver regions in at least one portal vein central line corresponding to the target portal vein node in the total length of the portal vein central line corresponding to the target portal vein node based on the portal vein blood vessel dendrogram and the second liver segmentation image.
Thirdly, under the condition that the maximum proportion in the second proportions corresponding to different hepatic region contour regions is larger than or equal to the set first proportion threshold, determining the middle category of at least one portal vein central line under the target portal vein node based on the hepatic region corresponding to the maximum proportion; wherein the middle category includes portal left vessel branches and portal right vessel branches.
And fourthly, under the condition that the second proportion corresponding to different hepatic region contour areas is smaller than the set first proportion threshold, respectively taking each portal vein node in at least one non-traversed portal vein node connected with the target portal vein node as the target portal vein node, and returning to the step of calculating the second proportion until the middle category of each portal vein center line in the portal vein blood vessel tree graph is determined.
And fifthly, dividing the portal vein central lines except the main portal vein central line in the portal vein tree diagram into a portal vein left blood vessel tree diagram and a portal vein right blood vessel tree diagram based on the middle category of each portal vein central line.
Here, a portal vein center line from a portal vein root node to outside the liver region in the portal vein tree map may be determined based on the portal vein tree map and the second liver segmentation image, and the determined portal vein center line may be determined as a portal vein main center line corresponding to the portal vein main blood vessel branch.
Illustratively, a search algorithm of breadth-first traversal can be further utilized to sequentially traverse each portal vein node in the portal vein vessel tree graph from the portal vein root node to determine the middle category of each portal vein centerline.
Firstly, taking a portal vein root node as a target portal vein node, determining the total length of at least one portal vein central line corresponding to the target portal vein node based on a portal vein blood vessel dendrogram and a second liver segmentation image, and calculating a second proportion of the length of a central line positioned in a left liver region outline area to the total length of the portal vein central line corresponding to the target portal vein node and a second proportion of the length of a central line positioned in a right liver region outline area to the total length of the portal vein central line corresponding to the target portal vein node in the at least one portal vein central line corresponding to the target portal vein node. And at least one portal vein central line corresponding to the target portal vein node is a portal vein central line which has a connection relation with the target portal vein node and is far away from one side of the portal vein root node.
And determining the maximum proportion in the second proportions corresponding to different hepatic region contour regions, and if the maximum proportion is greater than or equal to the set first proportion threshold, determining the middle category of at least one portal vein central line under the target portal vein node based on the hepatic region corresponding to the maximum proportion. For example, if the second proportion corresponding to the left hepatic region contour region is 70%, the second proportion corresponding to the right hepatic region contour region is 30%, and the first proportion threshold is 65%, the middle category of the at least one portal vein center line under the target portal vein node may be determined to be the portal vein left blood vessel branch based on the hepatic region (left hepatic region) corresponding to the maximum proportion (the second proportion corresponding to the left hepatic region contour region).
If the second proportions corresponding to different hepatic region contour regions are all smaller than the set first proportion threshold, for example, if the second proportion corresponding to the left hepatic region contour region is 60%, the second proportion corresponding to the right hepatic region contour region is 40%, and the first proportion threshold is 65%, each portal vein node in at least one non-traversed portal vein node connected to the target portal vein node may be respectively used as the target portal vein node, and the step of calculating the second proportions may be returned until the middle category of each portal vein centerline in the portal vein tree diagram is determined. The setting of the first proportional threshold value can be determined according to actual conditions.
Finally, the portal vein centerlines other than the main portal vein centerline in the portal vein dendrogram can be divided into a portal vein left blood vessel dendrogram and a portal vein right blood vessel dendrogram based on the middle category of each portal vein centerline.
For step D2:
determining a second category of each portal vein left centerline in the portal vein left vessel dendrogram based on the portal vein left vessel dendrogram and the second liver segmentation image may include:
firstly, determining a left main branch central line corresponding to a left main branch blood vessel in a portal vein left blood vessel tree diagram based on the portal vein left blood vessel tree diagram and a second liver segmentation image.
And secondly, determining a second category of the left central lines of other portal veins except the left main branch line in the portal vein left blood vessel tree diagram based on the left main branch line corresponding to the left main branch blood vessel in the portal vein left blood vessel tree diagram, the portal vein left blood vessel tree diagram and a second liver segmentation image.
By combining the structural characteristics of the left blood vessel of the portal vein, the left main branch central line corresponding to the left main branch blood vessel in the tree diagram of the left blood vessel of the portal vein can be determined, and then the left branch blood vessel tree connected with the left main branch blood vessel is determined based on the tree diagram of the left blood vessel of the portal vein; and then determining the second category of the left central line of each portal vein in each left branch vessel tree, namely obtaining the second category of the left central line of each portal vein in the portal vein left vessel tree diagram.
Here, the left main branch centerline corresponding to the left main branch vessel in the portal left vessel tree-like map may be determined first, the second category of the left centerline of the portal vein other than the left main branch centerline in the portal left vessel tree-like map is determined based on the left main branch centerline, the portal left vessel tree-like map, and the second liver segmentation image corresponding to the left main branch vessel in the portal left vessel tree-like map, and when the second category of each left centerline of the portal vein in the portal left vessel tree-like map is determined, the structural characteristics of the left vessel of the portal vein are combined, so that the accuracy of the second category of each left centerline of the portal vein in the portal left vessel tree-like map is determined to be higher.
In an alternative embodiment, determining a left main branch centerline corresponding to a left main branch vessel in the portal vein left vessel tree map based on the portal vein left vessel tree map and the second liver segmentation image includes:
step one, determining a first intersection point between a portal vein left blood vessel tree diagram and a portal vein right blood vessel tree diagram in the portal vein tree diagram.
And step two, taking the first intersection as a target portal vein left node, and determining at least one candidate portal vein left node which is connected with the target portal vein left node and is not traversed in the portal vein left blood vessel tree diagram.
Determining whether each candidate portal vein left node is an undetermined portal vein left node or not based on the portal vein left blood vessel tree diagram and the second liver segmentation image; the left central line of at least one portal vein connected with the left node of the portal vein to be determined comprises a portal vein central line belonging to a left inner liver area and a portal vein central line belonging to a left outer liver area.
And fourthly, under the condition that the number of the left nodes of the portal vein to be determined is multiple, determining a first length of the portal vein central line belonging to the left inner hepatic region and a second length of the portal vein central line belonging to the left outer hepatic region in at least one portal vein left central line connected with each left node of the portal vein to be determined, and determining the smaller length of the first length and the second length as a target length corresponding to the left node of the portal vein to be determined.
And fifthly, selecting the undetermined portal vein left node with the longest corresponding target length from the undetermined portal vein left nodes, taking the selected undetermined portal vein left node as an updated target portal vein left node, returning to the step of determining at least one candidate portal vein left node which is connected with the updated target portal vein left node and is not traversed in the portal vein left blood vessel tree graph until the undetermined portal vein left node does not exist in the at least one candidate portal vein left node which is connected with the target portal vein left node or until the candidate portal vein left node which is connected with the target portal vein left node and is not traversed does not exist.
And step six, determining a left main branch central line corresponding to a left main branch blood vessel of the portal vein left blood vessel tree diagram based on each target portal vein left node.
In the first step and the second step, a first intersection point between the portal vein left blood vessel tree map and the portal vein right blood vessel tree map in the portal vein tree map is determined. And firstly, the first intersection point is used as a target portal vein left node, and at least one candidate portal vein left node which is connected with the target portal vein left node and is not traversed in the portal vein left blood vessel tree diagram is determined.
In step three, at least one portal vein left center line below the left node of the portal vein to be determined comprises a portal vein center line of a left inner liver area and a portal vein center line of a left outer liver area.
As an optional embodiment, determining whether each candidate portal vein left node is a pending portal vein left node based on the portal vein left blood vessel tree map and the second liver segmentation image includes:
the method comprises the steps of determining, for each candidate portal vein left node, a first left inner length proportion of the length of a central line belonging to a left inner blood vessel in at least one portal vein left central line corresponding to the candidate portal vein left node in the total length of the portal vein left central line corresponding to the candidate portal vein left node, and a first left outer length proportion of the length of a central line belonging to a left outer blood vessel in the total length of the portal vein left central line corresponding to the candidate portal vein left node.
And secondly, determining the candidate portal vein left node as the pending portal vein left node under the condition that the smaller length proportion of the first left inner length proportion and the first left outer length proportion is larger than a set second proportion threshold.
For each candidate portal vein left node, a first left inner length proportion and a first left outer length proportion corresponding to the candidate portal vein left node are calculated, and a smaller length proportion in the first left inner length proportion and the second length proportion is determined, for example, if the first left inner length proportion is 30%, the first left outer length proportion is 70%, the smaller length proportion is the first left inner length proportion.
Considering that the second detected image may have an error, so that the obtained second liver segmentation image may also have an error, a second proportion threshold is set here, the minimum length proportion is compared with the set second proportion threshold, and if the minimum length proportion is greater than the set second proportion threshold, the candidate portal vein left node is determined to be the pending portal vein left node; and if the minimum length proportion is less than or equal to the set second proportion threshold, determining that the candidate portal vein left node does not belong to the pending portal vein left node.
For example, if the second proportion threshold is 30%, the length proportion of the center line belonging to the left internal blood vessel in at least one portal vein left blood vessel branch corresponding to the candidate portal vein left node a is 60%, and the length proportion of the center line belonging to the left external blood vessel is 40%, it is known that the length proportion of the center line belonging to the left external blood vessel is greater than the set second proportion threshold, and it is determined that the candidate portal vein left node a belongs to the pending portal vein left node. If the length proportion of the center line belonging to the left internal blood vessel in the at least one portal vein left blood vessel branch corresponding to the candidate portal vein left node B is 80%, and the length proportion of the center line belonging to the left external blood vessel is 20%, it is known that the length proportion of the center line belonging to the left external blood vessel is smaller than a set second proportion threshold, it may be that the length proportion of the center line belonging to the left external blood vessel in the determined at least one portal vein left blood vessel branch is 20% due to an error in the second liver segmentation image, so that it is determined that the candidate portal vein left node a does not belong to the undetermined portal vein left node.
Here, considering that the second detected image may have an error, so that the obtained second liver segmentation image may also have an error, a second proportion threshold is set, and when the minimum length proportion of at least one portal vein left center line of the candidate portal vein left node belongs to the length proportion of the left internal blood vessel and the length proportion of the left external blood vessel is less than or equal to the set second proportion threshold, it is determined that the candidate portal vein left node includes blood vessel branches in different liver regions due to the error of the second liver segmentation image, and at this time, the candidate portal vein left node may be screened out.
In the fourth step and the fifth step, if the number of the undetermined portal vein left nodes is one, the undetermined portal vein left node is used as a target portal vein left node, and the process in the second step is executed again.
And if the number of the undetermined portal vein left nodes is multiple, selecting one undetermined portal vein left node from the multiple undetermined portal vein left nodes, taking the selected undetermined portal vein left node as an updated target portal vein left node, and re-executing the process of determining at least one candidate portal vein left node which is connected with the target portal vein left node and is not traversed in the portal vein left blood vessel tree diagram in the step two. Specifically, when the number of the left nodes of the portal vein to be determined is multiple, for each portal vein node to be determined, determining a first length of a portal vein center line belonging to a left inner hepatic region and a second length of a portal vein center line belonging to a left outer hepatic region in at least one portal vein left center line under the portal vein node to be determined, and determining the smaller length of the first length and the second length as a target length corresponding to the left node of the portal vein to be determined; and selecting the undetermined portal vein left node corresponding to the maximum target length from the undetermined portal vein left nodes.
And selecting the undetermined portal vein nodes with longer lengths of the left internal blood vessel and the left external blood vessel according to the first length and the second length corresponding to each undetermined portal vein node. For example, if the plurality of pending portal vein left nodes include: if the portal vein left center line under the portal vein left node A to be determined is the portal vein center line belonging to the left inner liver area, the first length of the portal vein center line belonging to the left outer liver area is 20, and the second length of the portal vein center line belonging to the left outer liver area is 10; in at least one portal vein left center line under the left node B of the portal vein to be determined, the first length of the portal vein center line belonging to the left inner liver area is 15, and the second length of the portal vein center line belonging to the left outer liver area is 5; and if the second length in each undetermined portal vein left node is shorter, comparing the second lengths, and selecting the undetermined portal vein left node with the longer second length from the undetermined portal vein left node A and the undetermined portal vein left node B, namely selecting the undetermined portal vein left node A. Taking the selected left node A of the portal vein to be determined as a target portal vein left node, and re-executing the process in the step two; until at least one candidate portal vein left node connected with the target portal vein left node does not have the pending portal vein left node, or until a candidate portal vein left node which is connected with the target portal vein left node and is not traversed does not exist.
In step six, the left nodes of the target portal veins can be connected in sequence, and the left main branch central line corresponding to the left main branch vessel of the portal vein left vessel tree diagram is determined.
In an alternative embodiment, determining a second category of portal vein left center lines other than the left main branch line in the portal vein left blood vessel tree based on the left main branch line corresponding to the left main branch vessel in the portal vein left blood vessel tree, and the second liver segmentation image may include:
determining at least one left branch blood vessel tree connected with a left main branch center line based on a left main branch center line corresponding to a left main branch blood vessel in a portal left blood vessel tree diagram and the portal left blood vessel tree diagram.
Secondly, determining a second left inner length proportion of the left central line length of the portal vein belonging to the left inner blood vessel in the left branch blood vessel tree to the total length of the left branch blood vessel tree and a second left outer length proportion of the left central line length of the portal vein belonging to the left outer blood vessel to the total length of the left branch blood vessel tree based on a second liver segmentation image aiming at each left branch blood vessel tree; and determining a second category corresponding to the maximum length ratio in the second left inner length ratio and the second left outer length ratio as a second category of the left central line of each portal vein in the left branch vessel tree.
Here, the at least one left branch vessel tree to which the left main branch centerline is connected may be determined based on the left main branch centerline corresponding to the left main vessel in the portal left vessel tree and the portal left vessel tree.
And determining the total length of the left branch vessel tree aiming at each left branch vessel tree, determining a second left inner length proportion of the left central line length of the portal vein belonging to the left inner vessel in the left branch vessel tree in the total length of the left branch vessel tree and a second left outer length proportion of the left central line length of the portal vein belonging to the left outer vessel in the total length of the left branch vessel tree based on a second liver segmentation image, and determining the category with the larger length proportion in the second left inner length proportion and the second left outer length proportion as a second category of the left central line of each portal vein in the left branch vessel tree. For example, in the left branch vessel tree a, the second left inner length proportion is 80%, and the second left outer length proportion is 20%, then the second category of the left center line of each portal vein in the left branch vessel tree a is determined to be the left inner vessel.
For step D3:
determining a second category of each portal vein right center line in the portal vein right vessel dendrogram based on the portal vein right vessel dendrogram and the second liver segmentation image, wherein the second category comprises:
e1, determining the first intersection point between the portal vein left blood vessel tree diagram and the portal vein right blood vessel tree diagram in the portal vein tree diagram.
E2, determining the first intersection point as a target portal vein right node, and calculating a right anterior length proportion of the center line length of the right anterior hepatic region outline region in at least one portal vein right center line corresponding to the target portal vein right node in the total length of the portal vein right center line corresponding to the target portal vein right node and a right posterior length proportion of the center line length of the right posterior hepatic region outline region in the total length of the portal vein right center line corresponding to the target portal vein right node based on the portal vein right vessel dendrogram and the second liver segmentation image.
And E3, determining a second category of the right center line of at least one portal vein below the right node of the target portal vein based on the liver region category corresponding to the maximum length proportion when the maximum length proportion in the right front length proportion and the right rear length proportion is greater than or equal to the set third proportion threshold value.
E4, under the condition that the maximum length proportion in the right front length proportion and the right rear length proportion is smaller than a set third proportion threshold value, taking each portal vein node in at least one non-traversed portal vein node which is connected with the target portal vein right node in the portal vein right vessel dendrogram as the target portal vein right node respectively, and returning to the step of calculating the right front length proportion and the right rear length proportion until the second category of each portal vein right center line in the portal vein right vessel dendrogram is determined.
Here, a first intersection point between the portal vein left blood vessel tree and the portal vein right blood vessel tree in the portal vein tree may be determined, and the first intersection point may be a root node of the portal vein right blood vessel. Illustratively, a breadth-first traversal algorithm may be used to sequentially traverse the portal vein right nodes in the portal vein right vessel tree map starting from the first intersection point until the second category of each portal vein right centerline in the portal vein right vessel tree map is determined.
In specific implementation, the first intersection point may be used as a target portal vein right node, and the total length of at least one portal vein right center line corresponding to the target portal vein right node is determined; calculating the length of a central line in a right anterior hepatic region outline area and the length of a central line in a right posterior hepatic region outline area in at least one portal vein right central line corresponding to the target portal vein right node based on the portal vein right blood vessel dendrogram and the second liver segmentation image; finally, dividing the length of the central line in the outline area of the right anterior hepatic region by the total length of at least one portal vein right central line corresponding to the target portal vein right node, and determining the right anterior length proportion; and dividing the length of the central line in the contour area of the right posterior hepatic region by the total length of at least one right central line of the portal vein corresponding to the right node of the target portal vein, and determining the length proportion of the right posterior region.
And if the maximum length proportion in the front right length proportion and the rear right length proportion is larger than or equal to the set third proportion threshold, determining a second category of at least one portal vein right central line below the target portal vein right node from the liver region corresponding to the maximum length proportion. For example, if the right anterior length proportion corresponding to the target portal vein right node a is 80%, the right posterior length proportion is 20%, and the third proportion threshold is 75%, it is determined that the second category of the at least one portal vein right center line under the target portal vein right node a is the right anterior blood vessel.
And if the maximum length proportion of the front right length proportion and the rear right length proportion is smaller than the set third proportion threshold, taking each portal vein node in at least one non-traversed portal vein node connected with the target portal vein right node in the portal vein right vessel tree-shaped graph as the target portal vein right node, and re-executing the step E2 until the second category of each portal vein right center line in the portal vein right vessel tree-shaped graph is determined.
In an alternative embodiment, the determining the second category of the right center line of each portal vein in the portal vein right vessel tree based on the portal vein right vessel tree and the second liver segmentation image further includes:
first, from each target portal vein right node, determining a second intersection point of a right front blood vessel and a right back blood vessel in the portal vein right blood vessel tree diagram.
And secondly, determining a right main branch central line corresponding to the right main branch blood vessel in the portal vein right blood vessel tree diagram based on the second intersection point and the first intersection point.
In the process of determining the second category of each portal vein right center line in the portal vein right vessel tree graph, a second intersection point of a right front vessel and a right back vessel in the portal vein right vessel tree graph can be determined according to each target portal vein right node determined in the traversal process. In the plurality of portal vein right center lines under the second intersection point, at least one portal vein right center line belongs to the right anterior blood vessel and at least one portal vein right center line belongs to the right posterior blood vessel, and the right anterior blood vessel and the right posterior blood vessel are not intersected. And further determining a right main branch centerline of the portal vein right vessel dendrogram based on the first intersection point and the second intersection point. For example, the right main branch centerline is determined by connecting the right centerline of the portal vein between the first intersection and the second intersection.
After the second category of each portal vein right center line in the portal vein right vessel tree diagram is determined, a target image after hierarchical division processing can be generated based on the second category of each portal vein center line in the portal vein tree diagram and the portal vein detection image.
In an optional embodiment, the step of performing hierarchical division processing on the image of the first target area based on the category of the center line and the first detection image to obtain the target image includes:
the method comprises the steps that firstly, a target pixel point with the minimum distance to a pixel point to be classified is determined on a first structural graph aiming at each pixel point to be classified in a first target area of a first detection image; and determining the category of the center line to which the target pixel point belongs as the category corresponding to the pixel point to be classified.
And secondly, carrying out hierarchical division processing on the image of the first target area based on the category corresponding to each pixel point to be classified to obtain a target image.
Here, when the portal vein detection image is included in the first detection image, the first target region is a portal vein segmentation region, the pixel points to be classified can be second pixel points to be classified, the minimum distance value between each second pixel point to be classified and the portal vein arborescent graph included in the portal vein segmentation region of the first detection image can be calculated, the target pixel point on the portal vein arborescent graph corresponding to the minimum distance value of the second pixel point to be classified is determined, and the target pixel point corresponding to each second pixel point to be classified is obtained. And determining the second category of the center line to which the target pixel point belongs as the second category corresponding to the second pixel point to be classified. The second pixel point to be classified can be any pixel point located in the portal vein segmentation area in the portal vein detection image.
And then, based on the second category corresponding to each second pixel point to be classified, performing hierarchical division processing on the image of the first target area to obtain a target image, wherein the image of the first target area is a contour image corresponding to the portal vein segmentation area in the first detection image. Illustratively, different colors can be used for labeling different second categories, and different colors are used for distinguishing different categories of pixel points on the image of the portal vein detection image, so that the hierarchical division of the portal vein detection image is realized.
By adopting the method, aiming at each pixel point to be classified in the first target area of the first detection image, the target pixel point with the minimum distance to the pixel point to be classified on the first structural diagram is determined, the category corresponding to the pixel point to be classified is determined based on the category of the central line to which the target pixel point belongs, and the target image subjected to hierarchical division processing is generated based on the categories respectively corresponding to the pixel points to be classified, so that automatic hierarchical division of the first detection image is realized, and compared with the manual labeling process, the efficiency of the division is improved.
As shown in fig. 5, the graph includes a portal vein detection image (a first detection image), so that a portal vein blood vessel tree diagram corresponding to the portal vein detection image can be determined first; and dividing a plurality of liver segments included in the second detection image according to a set second liver region dividing mode to generate a second liver segmentation image including different liver region outline regions, wherein the first image is a liver segmentation image including a left liver region outline region and a right liver region outline region, the second image is a liver segmentation image including a left inner liver region outline region and a left outer liver region outline region, and the third image is a liver segmentation image including a right front liver region outline region and a right rear liver region outline region. And finally, generating a target image after hierarchical division processing based on the second category of each portal vein central line in the portal vein tree map.
It will be understood by those skilled in the art that in the method of the present invention, the order of writing the steps does not imply a strict order of execution and any limitations on the implementation, and the specific order of execution of the steps should be determined by their function and possible inherent logic.
Based on the same concept, an embodiment of the present disclosure further provides an image processing apparatus, as shown in fig. 6, which is an architecture schematic diagram of the image processing apparatus provided in the embodiment of the present disclosure, and includes an obtaining module 601, a generating module 602, a determining module 603, and a dividing module 604, specifically:
an obtaining module 601, configured to obtain at least one detection image of a target object, where the detection image includes a first detection image of a first target region of the target object and a second detection image of a second target region of the target object;
a generating module 602 for generating a first structural map of a vessel structure in the first target region based on the first detection image;
a determining module 603 configured to determine a category of each centerline characterizing a vessel branch in the first structural image based on the first structural image and the second detection image;
a dividing module 604, configured to perform hierarchical division processing on the image of the first target area based on the category of the central line and the first detection image to obtain a target image.
In a possible implementation, in case the first detection image comprises a hepatic vein detection image, the generating module 602, when generating the first structural map of the vascular structure in the first target region based on the first detection image, is configured to:
determining a connected component included in the first detection image based on a connection relationship between a hepatic vein and an inferior vena cava in the first detection image, wherein the connected component includes a plurality of hepatic vein blood vessel branches having a connection relationship;
determining a hepatic vein centerline for each hepatic vein vessel branch in the connected component;
determining a plurality of hepatic vein nodes based on intersection points between hepatic vein central lines of different hepatic vein branches; and
and generating a hepatic vein vessel tree diagram corresponding to the connected component based on the plurality of hepatic vein nodes and each hepatic vein central line.
In a possible implementation manner, the generating module 602, when generating the hepatic vein vessel tree map corresponding to the connected component based on a plurality of hepatic vein nodes and respective hepatic vein centerlines, is configured to:
acquiring a minimum distance value of each hepatic vein node from the contour region of the inferior vena cava in the first detection image;
determining hepatic vein root nodes corresponding to the connected components from the plurality of hepatic vein nodes based on the minimum distance value corresponding to each hepatic vein node and the position information of each hepatic vein node in the first detection image; and
and generating a hepatic vein vessel tree diagram corresponding to the connected component based on the hepatic vein root node corresponding to the determined connected component, other hepatic vein nodes except the hepatic vein root node in the plurality of hepatic vein nodes and the central line of each hepatic vein.
In a possible implementation, the determining module 603, when determining the category of each centerline that characterizes a blood vessel branch in the first structural map based on the first structural map and the second detection image, is further configured to:
dividing a plurality of liver segments included in the second detection image according to a first liver region dividing mode to generate a first liver segmentation image containing different liver region contour regions; and
and determining a first category of each hepatic vein central line in the hepatic vein vessel tree diagram based on the hepatic vein vessel tree diagram and the first liver segmentation image, wherein the first category is used for indicating a liver region to which the hepatic vein central line belongs.
In a possible embodiment, the determining module 603, when determining the first category of each hepatic vein centerline in the hepatic vein vessel tree based on the hepatic vein vessel tree and the first liver segmentation image, is configured to:
taking the hepatic vein root node as a target hepatic vein node, and calculating a first proportion of the lengths of center lines respectively positioned in different hepatic region contour areas in at least one hepatic vein center line corresponding to the target hepatic vein node in the hepatic vein center line based on the hepatic vein vessel dendrogram and the first liver segmentation image, wherein the first proportion accounts for the total length of the hepatic vein center line corresponding to the target hepatic vein node;
under the condition that the maximum proportion in the first proportions corresponding to different hepatic region contour areas is larger than or equal to a target proportion threshold corresponding to the target hepatic vein node, determining a first category of at least one hepatic vein central line under the target hepatic vein node based on the hepatic region corresponding to the maximum proportion; and
and under the condition that the first ratios corresponding to different hepatic region contour areas are smaller than the determined target ratio threshold corresponding to the target hepatic vein node, taking each hepatic vein node in the hepatic vein nodes which are connected with the target hepatic vein node and are not traversed as a new target hepatic vein node, and returning to the step of calculating the first ratio until the first category of each hepatic vein central line in the hepatic vein blood vessel tree graph is determined.
In a possible implementation, the determining module 603 is configured to determine the target proportion threshold corresponding to the target hepatic vein node according to the following steps:
determining a target proportion threshold corresponding to the target hepatic vein node based on the minimum distance value corresponding to the target hepatic vein node under the condition that the minimum distance value corresponding to the target hepatic vein node is smaller than a set threshold for dividing distances of the hepatic vein node; and
and under the condition that the minimum distance value corresponding to the target hepatic vein node is greater than or equal to a set distance threshold, determining a preset proportion threshold as the target proportion threshold corresponding to the target hepatic vein node.
In a possible implementation, in case the first detection image comprises a portal vein detection image, the generating module 602, when generating the first structural map of the vascular structure in the first target region based on the first detection image, is configured to:
determining a portal vein center line of each portal vein blood vessel branch in the first detection image;
determining a plurality of portal vein nodes based on intersection points between the portal vein centerlines of different portal vein branches;
determining a portal vein root node from the plurality of portal vein nodes based on the location information of each portal vein node in the first detected image; and
and generating a portal vein dendriographic graph based on the portal vein root node, portal vein nodes except the portal vein root node in the portal vein nodes and all portal vein central lines.
In one possible implementation, the determining module 603, when determining the category characterizing each centerline of a vessel branch in the first structural map based on the first structural map and the second detection image, is configured to:
dividing a plurality of liver segments included in the second detection image according to a second liver region dividing mode to generate a second liver segmentation image containing different liver region contour regions; and
and determining a second category of each portal vein central line in the portal vein tree graph based on the portal vein tree graph and the second liver segmentation image, wherein the second category is used for indicating a portal vein liver region to which the portal vein branch corresponding to the portal vein central line belongs.
In a possible embodiment, the second category includes: the main portal vein blood vessel branch, the left main branch blood vessel, the left internal blood vessel and the left external blood vessel corresponding to the left portal vein blood vessel branch, and the right main branch blood vessel, the right front blood vessel and the right back blood vessel corresponding to the right portal vein blood vessel branch;
the determining module 603, when determining the second category of each portal vein centerline in the portal vein vessel tree based on the portal vein vessel tree and the second liver segmentation image, is configured to:
dividing the portal vein dendrogram into a portal vein left blood vessel dendrogram, a portal vein right blood vessel dendrogram and a portal vein main central line corresponding to the portal vein main blood vessel branch based on the portal vein dendrogram and the second liver segmentation image;
determining a second category of each portal vein left centerline in the portal vein left vessel dendrogram based on the portal vein left vessel dendrogram and the second liver segmentation image; and
determining a second category of each portal vein right center line in the portal vein right vessel tree diagram based on the portal vein right vessel tree diagram and the second liver segmentation image.
In one possible embodiment, the determining module 603, when dividing the portal vein dendrogram into a portal vein left blood vessel dendrogram, a portal vein right blood vessel dendrogram and a portal vein main central line corresponding to a portal vein main blood vessel branch based on the portal vein dendrogram and the second liver segmentation image, is configured to:
determining the portal vein central line from the portal vein root node to the outside of the liver region in the portal vein tree graph as the portal vein main central line corresponding to the portal vein main vessel branch;
determining the portal vein root node as a target portal vein node, and calculating a second proportion of the lengths of central lines respectively positioned in different hepatic region contour areas in at least one portal vein central line corresponding to the target portal vein node in the total length of the portal vein central line corresponding to the target portal vein node on the basis of the portal vein vessel dendrogram and the second liver segmentation image;
under the condition that the maximum proportion in the second proportions corresponding to different hepatic region contour areas is larger than or equal to a set first proportion threshold, determining the middle category of at least one portal vein central line under the target portal vein node based on the hepatic region corresponding to the maximum proportion; wherein the middle category includes portal left vessel branches and portal right vessel branches;
under the condition that the second proportion corresponding to different hepatic region contour areas is smaller than a set first proportion threshold, taking each portal vein node in at least one non-traversed portal vein node connected with the target portal vein node as a target portal vein node respectively, and returning to the step of calculating the second proportion until the middle category of each portal vein center line in the portal vein blood vessel dendrogram is determined; and
and dividing the portal vein centerlines except the main portal vein centerline in the portal vein dendrogram into a portal vein left blood vessel dendrogram and a portal vein right blood vessel dendrogram based on the middle category of each portal vein centerline.
In a possible embodiment, the determining module 603, when determining the second category of each portal vein left center line in the portal vein left blood vessel tree based on the portal vein left blood vessel tree and the second liver segmentation image, is configured to:
determining a left main branch central line corresponding to a left main branch blood vessel in the portal vein left blood vessel tree diagram based on the portal vein left blood vessel tree diagram and the second liver segmentation image; and
and determining a second category of the left central lines of other portal veins except the left main branch central line in the portal left blood vessel tree diagram based on the left main branch central line corresponding to the left main branch blood vessel in the portal left blood vessel tree diagram, the portal left blood vessel tree diagram and the second liver segmentation image.
In one possible embodiment, the determining module 603, when determining a left main branch centerline corresponding to a left main branch vessel in the portal left vessel tree based on the portal left vessel tree and the second liver segmentation image, is configured to:
determining a first intersection point between the portal vein left blood vessel dendrogram and the portal vein right blood vessel dendrogram in the portal vein dendrogram;
determining at least one candidate portal vein left node which is connected with the target portal vein left node and is not traversed in the portal vein left blood vessel tree diagram by taking the first intersection as the target portal vein left node;
determining whether each candidate portal vein left node is a pending portal vein left node or not based on the portal vein left blood vessel dendrogram and the second liver segmentation image; the left central line of at least one portal vein connected with the left node of the portal vein to be determined comprises a portal vein central line belonging to a left inner liver area and a portal vein central line belonging to a left outer liver area;
under the condition that the number of the left nodes of the portal vein to be determined is multiple, determining a first length of a portal vein central line belonging to a left inner hepatic region and a second length of a portal vein central line belonging to a left outer hepatic region in at least one portal vein left central line connected with each left node of the portal vein to be determined, and determining the first length and the second length as target lengths corresponding to the left nodes of the portal vein to be determined;
selecting an undetermined portal vein left node with the longest corresponding target length from a plurality of undetermined portal vein left nodes, taking the selected undetermined portal vein left node as an updated target portal vein left node, and returning to the step of determining at least one unexplored candidate portal vein left node connected with the target portal vein left node in the portal vein left blood vessel tree graph until the undetermined portal vein left node does not exist in the at least one candidate portal vein left node connected with the target portal vein left node or until the unexplored candidate portal vein left node connected with the target portal vein left node does not exist; and
and determining a left main branch central line corresponding to the left main branch vessel of the portal vein left vessel tree diagram based on each target portal vein left node.
In a possible implementation, the determining module 603, when determining whether each candidate portal vein left node is a pending portal vein left node based on the portal vein left blood vessel tree map and the second liver segmentation image, is configured to:
for each candidate portal vein left node, determining a first left inner length proportion of the length of a central line belonging to a left inner blood vessel in at least one portal vein left central line corresponding to the candidate portal vein left node in the total length of the portal vein left central line corresponding to the candidate portal vein left node, and a first left outer length proportion of the length of a central line belonging to a left outer blood vessel in the total length of the portal vein left central line corresponding to the candidate portal vein left node; and
and under the condition that the smaller length proportion of the first left inner length proportion and the first left outer length proportion is larger than a set second proportion threshold, determining the candidate portal vein left node as the pending portal vein left node.
In one possible embodiment, the determining module 603, when determining a second category of portal vein left center lines other than the left main branch line in the portal vein left blood vessel tree based on the left main branch line corresponding to the left main branch vessel in the portal vein left blood vessel tree, and the second liver segmentation image, is configured to:
determining at least one left branch vessel tree connected with a left main branch vessel in the portal left vessel tree diagram based on a left main branch center line corresponding to the left main branch vessel and the portal left vessel tree diagram;
for each left branch vessel tree, determining a second left inner length proportion of the left central line length of the portal vein belonging to the left inner vessel in the left branch vessel tree to the total length of the left branch vessel tree and a second left outer length proportion of the left central line length of the portal vein belonging to the left outer vessel to the total length of the left branch vessel tree based on the second liver segmentation image; and
and determining a second category corresponding to the maximum length ratio in the second left inner length ratio and the second left outer length ratio as a second category of the left central line of each portal vein in the left branch vessel tree.
In a possible implementation, the determining module 603, when determining the second category of each portal vein right center line in the portal vein right vessel tree based on the portal vein right vessel tree and the second liver segmentation image, is configured to:
determining a first intersection point between the portal vein left blood vessel dendrogram and the portal vein right blood vessel dendrogram in the portal vein dendrogram;
determining the first intersection point as a target portal vein right node, and calculating a right anterior length proportion of the length of the center line in the right anterior hepatic region outline area to the total length of the portal vein right center line corresponding to the target portal vein right node and a right posterior length proportion of the length of the center line in the right posterior hepatic region outline area to the total length of the portal vein right center line corresponding to the target portal vein right node in at least one portal vein right center line corresponding to the target portal vein right node based on the portal vein right vessel dendrogram and the second liver segmentation image;
under the condition that the maximum length proportion of the front right length proportion and the rear right length proportion is larger than or equal to a set third proportion threshold, determining a second category of at least one portal vein right center line under the target portal vein right node based on the liver region category corresponding to the maximum length proportion;
and under the condition that the maximum length proportion in the right front length proportion and the right rear length proportion is smaller than a set third proportion threshold, taking each portal vein node in at least one non-traversed portal vein node which is connected with the target portal vein right node in the portal vein right vessel dendrogram as the target portal vein right node, and returning to the step of calculating the right front length proportion and the right rear length proportion until the second category of each portal vein right center line in the portal vein right vessel dendrogram is determined.
In a possible implementation, the determining module 603, when determining the second category of each portal vein right center line in the portal vein right vessel tree based on the portal vein right vessel tree and the second liver segmentation image, is further configured to:
determining a second intersection point of a right front blood vessel and a right back blood vessel in the portal vein right blood vessel tree diagram from each target portal vein right node;
and determining a right main branch central line corresponding to the right main branch vessel in the portal vein right vessel tree diagram based on the second intersection point and the first intersection point.
In a possible implementation manner, the dividing module 604, when performing hierarchical division processing on the image of the first target area based on the category of the central line and the first detection image to obtain a target image, is configured to:
for each pixel point to be classified in the first target region of the first detection image, determining a target pixel point with the minimum distance to the pixel point to be classified on the first structural graph; determining the category of the center line to which the target pixel point belongs as the category corresponding to the pixel point to be classified;
and carrying out hierarchical division processing on the image of the first target area based on the category corresponding to each pixel point to be classified respectively to obtain a target image.
In some embodiments, the functions of the apparatus provided in the embodiments of the present disclosure or the included templates may be used to execute the method described in the above method embodiments, and specific implementation thereof may refer to the description of the above method embodiments, and for brevity, no further description is provided here.
Based on the same technical concept, the embodiment of the disclosure also provides an electronic device. Referring to fig. 7, a schematic structural diagram of an electronic device provided in the embodiment of the present disclosure includes a processor 701, a memory 702, and a bus 703. The memory 702 is used for storing execution instructions and includes a memory 7021 and an external memory 7022; the memory 7021 is also referred to as an internal memory, and is used to temporarily store operation data in the processor 701 and data exchanged with an external memory 7022 such as a hard disk, the processor 701 exchanges data with the external memory 7022 through the memory 7021, and when the electronic device 700 is operated, the processor 701 and the memory 702 communicate with each other through the bus 703, so that the processor 701 executes the following instructions:
acquiring at least one detection image of a target object, wherein the detection image comprises a first detection image of a first target area of the target object and a second detection image of a second target area of the target object;
generating a first structural map of a vessel structure in the first target region based on the first detected image;
determining a category of each centerline characterizing a vessel branch in the first configuration map based on the first configuration map and the second inspection image;
and carrying out hierarchical division processing on the image of the first target area based on the category of the central line and the first detection image to obtain a target image.
Furthermore, the embodiments of the present disclosure also provide a computer-readable storage medium, on which a computer program is stored, and the computer program is executed by a processor to perform the steps of the image processing method described in the above method embodiments.
The embodiments of the present disclosure also provide a computer program product, where the computer program product carries a program code, and instructions included in the program code may be used to execute the steps of the image processing method in the foregoing method embodiments, which may be referred to specifically in the foregoing method embodiments, and are not described herein again.
The computer program product may be implemented by hardware, software or a combination thereof. In an alternative embodiment, the computer program product is embodied in a computer storage medium, and in another alternative embodiment, the computer program product is embodied in a Software product, such as a Software Development Kit (SDK), or the like.
It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the system and the apparatus described above may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again. In the several embodiments provided in the present disclosure, it should be understood that the disclosed system, apparatus, and method may be implemented in other ways. The above-described embodiments of the apparatus are merely illustrative, and for example, the division of the units is only one logical division, and there may be other divisions when actually implemented, and for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection of devices or units through some communication interfaces, and may be in an electrical, mechanical or other form.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, functional units in the embodiments of the present disclosure may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit.
The functions, if implemented in the form of software functional units and sold or used as a stand-alone product, may be stored in a non-volatile computer-readable storage medium executable by a processor. Based on such understanding, the technical solution of the present disclosure may be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present disclosure. And the aforementioned storage medium includes: various media capable of storing program codes, such as a usb disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.
The above are only specific embodiments of the present disclosure, but the scope of the present disclosure is not limited thereto, and any person skilled in the art can easily conceive of changes or substitutions within the technical scope of the present disclosure, and shall be covered by the scope of the present disclosure. Therefore, the protection scope of the present disclosure shall be subject to the protection scope of the claims.
Claims (20)
1. An image processing method, comprising:
acquiring at least one detection image of a target object, wherein the detection image comprises a first detection image of a first target area of the target object and a second detection image of a second target area of the target object;
generating a first structural map of a vessel structure in the first target region based on the first detected image;
determining a category of each centerline characterizing a vessel branch in the first configuration map based on the first configuration map and the second inspection image; and
and carrying out hierarchical division processing on the image of the first target area based on the category of the central line and the first detection image to obtain a target image.
2. The method of claim 1, wherein generating a first map of structures of vessels in the first target region based on the first detection image in a case where the first detection image comprises a hepatic vein detection image comprises:
determining a connected component included in the first detection image based on a connection relationship between a hepatic vein and an inferior vena cava in the first detection image, wherein the connected component includes a plurality of hepatic vein blood vessel branches having a connection relationship;
determining a hepatic vein centerline for each hepatic vein vessel branch in the connected component;
determining a plurality of hepatic vein nodes based on intersection points between hepatic vein central lines of different hepatic vein branches; and
and generating a hepatic vein vessel tree diagram corresponding to the connected component based on the plurality of hepatic vein nodes and each hepatic vein central line.
3. The method of claim 2, wherein generating a hepatic vein vessel dendrogram corresponding to the connected component based on the plurality of hepatic vein nodes and the respective hepatic vein centerlines comprises:
acquiring a minimum distance value of each hepatic vein node from the contour region of the inferior vena cava in the first detection image;
determining hepatic vein root nodes corresponding to the connected components from the plurality of hepatic vein nodes based on the minimum distance value corresponding to each hepatic vein node and the position information of each hepatic vein node in the first detection image; and
and generating a hepatic vein vessel tree diagram corresponding to the connected component based on the hepatic vein root node corresponding to the determined connected component, other hepatic vein nodes except the hepatic vein root node in the plurality of hepatic vein nodes and the respective hepatic vein central lines.
4. The method of claim 3, wherein determining a category for each centerline characterizing a vessel branch in the first map based on the first map and the second detected image comprises:
dividing a plurality of liver segments included in the second detection image according to a first liver region dividing mode to generate a first liver segmentation image containing different liver region contour regions; and
and determining a first category of each hepatic vein central line in the hepatic vein vessel tree diagram based on the hepatic vein vessel tree diagram and the first liver segmentation image, wherein the first category is used for indicating a liver region to which a hepatic vein vessel branch corresponding to the hepatic vein central line belongs.
5. The method of claim 4, wherein determining a first category of each hepatic vein centerline in the hepatic vein vessel tree based on the hepatic vein vessel tree and the first liver segmentation image comprises:
taking the hepatic vein root node as a target hepatic vein node, and calculating a first proportion of the lengths of center lines respectively positioned in different hepatic region contour areas in at least one hepatic vein center line corresponding to the target hepatic vein node in the hepatic vein center line based on the hepatic vein vessel dendrogram and the first liver segmentation image, wherein the first proportion accounts for the total length of the hepatic vein center line corresponding to the target hepatic vein node;
under the condition that the maximum proportion in the first proportions corresponding to different hepatic region contour areas is larger than or equal to a target proportion threshold corresponding to the target hepatic vein node, determining a first category of at least one hepatic vein central line under the target hepatic vein node based on the hepatic region corresponding to the maximum proportion; and
and under the condition that the first ratios corresponding to different hepatic region contour areas are smaller than the determined target ratio threshold corresponding to the target hepatic vein node, taking each hepatic vein node in the hepatic vein nodes which are connected with the target hepatic vein node and are not traversed as a new target hepatic vein node, and returning to the step of calculating the first ratio until the first category of each hepatic vein central line in the hepatic vein blood vessel tree graph is determined.
6. The method of claim 5, wherein the target proportion threshold corresponding to the target hepatic vein node is determined according to the following steps:
under the condition that the minimum distance value corresponding to the target hepatic vein node is smaller than a set distance threshold value for dividing the hepatic vein node, determining a target proportion threshold value corresponding to the target hepatic vein node based on the minimum distance value corresponding to the target hepatic vein node; and
and under the condition that the minimum distance value corresponding to the target hepatic vein node is greater than or equal to a set distance threshold, determining a preset proportion threshold as the target proportion threshold corresponding to the target hepatic vein node.
7. The method of claim 1, wherein generating a first map of structures of vessels in the first target region based on the first detection image in a case where the first detection image comprises a portal vein detection image comprises:
determining a portal vein center line of each portal vein blood vessel branch in the first detection image;
determining a plurality of portal vein nodes based on intersection points between the portal vein centerlines of different portal vein branches;
determining a portal vein root node from the plurality of portal vein nodes based on the location information of each portal vein node in the first detected image; and
and generating a portal vein dendriographic graph based on the portal vein root node, portal vein nodes except the portal vein root node in the portal vein nodes and all portal vein central lines.
8. The method of claim 7, wherein determining a category for each centerline characterizing a vessel branch in the first map based on the first map and the second detected image comprises:
dividing a plurality of liver segments included in the second detection image according to a second liver region dividing mode to generate a second liver segmentation image containing different liver region contour regions; and
and determining a second category of each portal vein central line in the portal vein tree graph based on the portal vein tree graph and the second liver segmentation image, wherein the second category is used for indicating a liver region to which the portal vein branch corresponding to the portal vein central line belongs.
9. The method of claim 8, wherein the second category comprises: the main portal vein blood vessel branch, the left main branch blood vessel, the left internal blood vessel and the left external blood vessel corresponding to the left portal vein blood vessel branch, and the right main branch blood vessel, the right front blood vessel and the right back blood vessel corresponding to the right portal vein blood vessel branch; and
the determining a second category of each portal vein centerline in the portal vein tree based on the portal vein tree and the second liver segmentation image comprises:
dividing the portal vein dendrogram into a portal vein left blood vessel dendrogram, a portal vein right blood vessel dendrogram and a portal vein main central line corresponding to the portal vein main blood vessel branch based on the portal vein dendrogram and the second liver segmentation image;
determining a second category of each portal vein left centerline in the portal vein left vessel dendrogram based on the portal vein left vessel dendrogram and the second liver segmentation image; and
determining a second category of each portal vein right center line in the portal vein right vessel tree diagram based on the portal vein right vessel tree diagram and the second liver segmentation image.
10. The method of claim 9, wherein the dividing the portal blood vessel dendrogram into a portal left blood vessel dendrogram, a portal right blood vessel dendrogram, and a portal main centerline corresponding to the portal main blood vessel branch based on the portal blood vessel dendrogram and the second liver segmentation image comprises:
determining the portal vein central line from the portal vein root node to the outside of the liver region in the portal vein tree graph as the portal vein main central line corresponding to the portal vein main vessel branch;
determining the portal vein root node as a target portal vein node, and calculating a second proportion of the lengths of central lines respectively positioned in different hepatic region contour areas in at least one portal vein central line corresponding to the target portal vein node in the total length of the portal vein central line corresponding to the target portal vein node on the basis of the portal vein vessel dendrogram and the second liver segmentation image;
under the condition that the maximum proportion in the second proportions corresponding to different hepatic region contour areas is larger than or equal to a set first proportion threshold, determining an intermediate class of at least one portal vein central line under the target portal vein node based on the hepatic region corresponding to the maximum proportion, wherein the intermediate class comprises the portal vein left blood vessel branch and the portal vein right blood vessel branch;
under the condition that the second proportion corresponding to different hepatic region contour areas is smaller than a set first proportion threshold, taking each portal vein node in at least one non-traversed portal vein node connected with the target portal vein node as a target portal vein node respectively, and returning to the step of calculating the second proportion until the middle category of each portal vein center line in the portal vein blood vessel dendrogram is determined; and
and dividing the portal vein centerlines except the main portal vein centerline in the portal vein dendrogram into a portal vein left blood vessel dendrogram and a portal vein right blood vessel dendrogram based on the middle category of each portal vein centerline.
11. The method of claim 9, wherein determining a second category for each portal left centerline in the portal left vessel dendrogram based on the portal left vessel dendrogram and the second liver segmentation image comprises:
determining a left main branch central line corresponding to a left main branch blood vessel in the portal vein left blood vessel tree diagram based on the portal vein left blood vessel tree diagram and the second liver segmentation image; and
and determining a second category of the left central lines of other portal veins except the left main branch central line in the portal left blood vessel tree diagram based on the left main branch central line corresponding to the left main branch blood vessel in the portal left blood vessel tree diagram, the portal left blood vessel tree diagram and the second liver segmentation image.
12. The method according to claim 11, wherein the determining a left main branch centerline corresponding to a left main branch vessel in the portal left vessel tree based on the portal left vessel tree and the second liver segmentation image comprises:
determining a first intersection point between the portal vein left blood vessel dendrogram and the portal vein right blood vessel dendrogram in the portal vein dendrogram;
determining at least one candidate portal vein left node which is connected with the target portal vein left node and is not traversed in the portal vein left blood vessel tree diagram by taking the first intersection as the target portal vein left node;
determining whether each candidate portal vein left node is an undetermined portal vein left node or not based on the portal vein left blood vessel tree diagram and the second liver segmentation image, wherein at least one portal vein left center line connected with the undetermined portal vein left node comprises a portal vein center line belonging to a left inner liver region and a portal vein center line belonging to a left outer liver region;
under the condition that the number of the left nodes of the portal vein to be determined is multiple, determining a first length of a portal vein central line belonging to a left inner hepatic region and a second length of a portal vein central line belonging to a left outer hepatic region in at least one portal vein left central line connected with each left node of the portal vein to be determined, and determining the smaller length of the first length and the second length as a target length corresponding to the left node of the portal vein to be determined;
selecting an undetermined portal vein left node with the longest corresponding target length from a plurality of undetermined portal vein left nodes, taking the selected undetermined portal vein left node as an updated target portal vein left node, and returning to the step of determining at least one unexplored candidate portal vein left node connected with the target portal vein left node in the portal vein left blood vessel tree graph until the undetermined portal vein left node does not exist in the at least one candidate portal vein left node connected with the target portal vein left node or until the unexplored candidate portal vein left node connected with the target portal vein left node does not exist; and
and determining a left main branch central line corresponding to the left main branch vessel of the portal vein left vessel tree diagram based on each target portal vein left node.
13. The method of claim 12, wherein determining whether each candidate portal left node is a pending portal left node based on the portal left vessel dendrogram and the second liver segmentation image comprises:
for each candidate portal vein left node, determining a first left inner length proportion of the length of a central line belonging to a left inner blood vessel in at least one portal vein left central line corresponding to the candidate portal vein left node in the total length of the portal vein left central line corresponding to the candidate portal vein left node, and a first left outer length proportion of the length of a central line belonging to a left outer blood vessel in the total length of the portal vein left central line corresponding to the candidate portal vein left node; and
and under the condition that the smaller length proportion of the first left inner length proportion and the first left outer length proportion is larger than a set second proportion threshold, determining the candidate portal vein left node as the pending portal vein left node.
14. The method according to claim 11, wherein the determining a second category of portal vein left center lines other than the left main branch line in the portal vein left blood vessel tree based on the left main branch line corresponding to the left main branch vessel in the portal vein left blood vessel tree, and the second liver segmentation image comprises:
determining at least one left branch vessel tree connected with a left main branch vessel in the portal left vessel tree diagram based on a left main branch center line corresponding to the left main branch vessel and the portal left vessel tree diagram;
for each left branch vessel tree, determining a second left inner length proportion of the left central line length of the portal vein belonging to the left inner vessel in the left branch vessel tree to the total length of the left branch vessel tree and a second left outer length proportion of the left central line length of the portal vein belonging to the left outer vessel to the total length of the left branch vessel tree based on the second liver segmentation image; and
and determining a second category corresponding to the maximum length ratio in the second left inner length ratio and the second left outer length ratio as a second category of the left central line of each portal vein in the left branch vessel tree.
15. The method of claim 9, wherein determining a second category for each portal vein right centerline in the portal vein right vessel dendrogram based on the portal vein right vessel dendrogram and the second liver segmentation image comprises:
determining a first intersection point between the portal vein left blood vessel dendrogram and the portal vein right blood vessel dendrogram in the portal vein dendrogram;
determining the first intersection point as a target portal vein right node, and calculating a right anterior length proportion of the length of the center line in the right anterior hepatic region outline area to the total length of the portal vein right center line corresponding to the target portal vein right node and a right posterior length proportion of the length of the center line in the right posterior hepatic region outline area to the total length of the portal vein right center line corresponding to the target portal vein right node in at least one portal vein right center line corresponding to the target portal vein right node based on the portal vein right vessel dendrogram and the second liver segmentation image;
under the condition that the maximum length proportion of the front right length proportion and the rear right length proportion is larger than or equal to a set third proportion threshold, determining a second category of at least one portal vein right center line under the target portal vein right node based on the liver region category corresponding to the maximum length proportion;
and under the condition that the maximum length proportion in the right front length proportion and the right rear length proportion is smaller than a set third proportion threshold, taking each portal vein node in at least one non-traversed portal vein node which is connected with the target portal vein right node in the portal vein right vessel dendrogram as the target portal vein right node, and returning to the step of calculating the right front length proportion and the right rear length proportion until the second category of each portal vein right center line in the portal vein right vessel dendrogram is determined.
16. The method of claim 15, wherein determining a second category for each portal vein right centerline in the portal vein right vessel dendrogram based on the portal vein right vessel dendrogram and the second liver segmentation image further comprises:
determining a second intersection point of a right front blood vessel and a right back blood vessel in the portal vein right blood vessel tree diagram from each target portal vein right node;
and determining a right main branch central line corresponding to the right main branch vessel in the portal vein right vessel tree diagram based on the second intersection point and the first intersection point.
17. The method according to any one of claims 1 to 16, wherein the step of performing hierarchical division processing on the image of the first target region based on the category of the center line and the first detection image to obtain a target image comprises:
for each pixel point to be classified in the first target region of the first detection image, determining a target pixel point with the minimum distance to the pixel point to be classified on the first structural graph;
determining the category of the center line to which the target pixel point belongs as the category corresponding to the pixel point to be classified;
and carrying out hierarchical division processing on the image of the first target area based on the category corresponding to each pixel point to be classified respectively to obtain a target image.
18. An image processing apparatus characterized by comprising:
the device comprises an acquisition module, a detection module and a display module, wherein the acquisition module is used for acquiring at least one detection image of a target object, and the detection image comprises a first detection image of a first target area of the target object and a second detection image of a second target area of the target object;
a generating module for generating a first structural map of a vessel structure in the first target region based on the first detection image;
a determination module for determining a category of each centerline characterizing a vessel branch in the first map based on the first map and the second detected image;
and the dividing module is used for carrying out hierarchical dividing processing on the image of the first target area based on the category of the central line and the first detection image to obtain a target image.
19. An electronic device, comprising: a processor, a memory and a bus, the memory storing machine-readable instructions executable by the processor, the processor and the memory communicating over the bus when the electronic device is operating, the machine-readable instructions when executed by the processor performing the steps of the image processing method according to any one of claims 1 to 17.
20. A computer-readable storage medium, characterized in that a computer program is stored on the computer-readable storage medium, which computer program, when being executed by a processor, performs the steps of the image processing method according to any one of claims 1 to 17.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110126885.XA CN112842371A (en) | 2021-01-29 | 2021-01-29 | Image processing method, image processing device, electronic equipment and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110126885.XA CN112842371A (en) | 2021-01-29 | 2021-01-29 | Image processing method, image processing device, electronic equipment and storage medium |
Publications (1)
Publication Number | Publication Date |
---|---|
CN112842371A true CN112842371A (en) | 2021-05-28 |
Family
ID=75986911
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110126885.XA Withdrawn CN112842371A (en) | 2021-01-29 | 2021-01-29 | Image processing method, image processing device, electronic equipment and storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN112842371A (en) |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103810752A (en) * | 2014-02-18 | 2014-05-21 | 海信集团有限公司 | Liver segmenting method based on medical image and liver segmenting system thereof |
WO2015059706A2 (en) * | 2013-10-24 | 2015-04-30 | Cathworks Ltd. | Vascular characteristic determination with correspondence modeling of a vascular tree |
WO2017028519A1 (en) * | 2015-08-18 | 2017-02-23 | 青岛海信医疗设备股份有限公司 | Hepatic vascular classification method |
JP2019103749A (en) * | 2017-12-14 | 2019-06-27 | キヤノンメディカルシステムズ株式会社 | Medical image processing apparatus, medical image diagnosis apparatus, and medical image processing program |
CN111145206A (en) * | 2019-12-27 | 2020-05-12 | 联想(北京)有限公司 | Liver image segmentation quality evaluation method and device and computer equipment |
-
2021
- 2021-01-29 CN CN202110126885.XA patent/CN112842371A/en not_active Withdrawn
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2015059706A2 (en) * | 2013-10-24 | 2015-04-30 | Cathworks Ltd. | Vascular characteristic determination with correspondence modeling of a vascular tree |
CN103810752A (en) * | 2014-02-18 | 2014-05-21 | 海信集团有限公司 | Liver segmenting method based on medical image and liver segmenting system thereof |
WO2017028519A1 (en) * | 2015-08-18 | 2017-02-23 | 青岛海信医疗设备股份有限公司 | Hepatic vascular classification method |
JP2019103749A (en) * | 2017-12-14 | 2019-06-27 | キヤノンメディカルシステムズ株式会社 | Medical image processing apparatus, medical image diagnosis apparatus, and medical image processing program |
CN111145206A (en) * | 2019-12-27 | 2020-05-12 | 联想(北京)有限公司 | Liver image segmentation quality evaluation method and device and computer equipment |
Non-Patent Citations (1)
Title |
---|
李占军;周智洋;章作铨;邱建平;谢学斌;邹学农;: "多层螺旋CT血管成像评价肝移植受者移植前后血管结构", 中国组织工程研究与临床康复, no. 44, pages 52 - 56 * |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111145206B (en) | Liver image segmentation quality assessment method and device and computer equipment | |
JP5478171B2 (en) | Method and apparatus for classification of coronary image data | |
CN111528871B (en) | Methods and systems for whole body bone removal and vessel visualization in medical images | |
EP3473177A1 (en) | Baseline impedance maps for tissue proximity indications | |
US9679389B2 (en) | Method and system for blood vessel segmentation and classification | |
CN112884826B (en) | Method and device for extracting center line of blood vessel | |
Gala et al. | Active learning of neuron morphology for accurate automated tracing of neurites | |
US11461895B2 (en) | Automatic identification and processing of anatomical structures in an anatomical map | |
CN105139030A (en) | Method for sorting hepatic vessels | |
CN107730542A (en) | Cone beam computed tomography image corresponds to and method for registering | |
CN117373070B (en) | Method and device for labeling blood vessel segments, electronic equipment and storage medium | |
CN113192069A (en) | Semantic segmentation method and device for tree structure in three-dimensional tomography image | |
CN111260704A (en) | Vascular structure 3D/2D rigid registration method and device based on heuristic tree search | |
KR101625955B1 (en) | Method of classifying artery and vein of organ | |
CN114155193B (en) | Blood vessel segmentation method and device based on feature enhancement | |
CN114299057A (en) | Method for extracting blood vessel center line and storage medium | |
Metzen et al. | Matching of anatomical tree structures for registration of medical images | |
CN107392891A (en) | Vessel tree extraction method, apparatus, equipment and storage medium | |
CN112842371A (en) | Image processing method, image processing device, electronic equipment and storage medium | |
Xie et al. | Automatic neuron tracing in volumetric microscopy images with anisotropic path searching | |
Zheng et al. | Precise segmentation of the left atrium in C-arm CT volumes with applications to atrial fibrillation ablation | |
CN114119645B (en) | Method, system, device and medium for determining image segmentation quality | |
Homann et al. | Vasculature segmentation of CT liver images using graph cuts and graph-based analysis | |
CN111242915B (en) | Vascular 3D/2D registration method and device based on Monte Carlo tree search | |
CN112446893A (en) | Contour segmentation method and device for liver image |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
WW01 | Invention patent application withdrawn after publication |
Application publication date: 20210528 |
|
WW01 | Invention patent application withdrawn after publication |