CN114820554A - Liver blood vessel extraction method, system and computer readable storage medium based on global automatic growth - Google Patents
Liver blood vessel extraction method, system and computer readable storage medium based on global automatic growth Download PDFInfo
- Publication number
- CN114820554A CN114820554A CN202210513891.5A CN202210513891A CN114820554A CN 114820554 A CN114820554 A CN 114820554A CN 202210513891 A CN202210513891 A CN 202210513891A CN 114820554 A CN114820554 A CN 114820554A
- Authority
- CN
- China
- Prior art keywords
- image
- tree
- liver
- growth
- point
- 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.)
- Pending
Links
- 210000004185 liver Anatomy 0.000 title claims abstract description 100
- 210000004204 blood vessel Anatomy 0.000 title claims abstract description 56
- 238000000605 extraction Methods 0.000 title claims abstract description 41
- 238000003860 storage Methods 0.000 title claims abstract description 8
- 230000002440 hepatic effect Effects 0.000 claims abstract description 60
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 44
- 238000000034 method Methods 0.000 claims abstract description 44
- 230000008569 process Effects 0.000 claims abstract description 39
- 230000002792 vascular Effects 0.000 claims abstract description 37
- 238000005259 measurement Methods 0.000 claims abstract description 36
- 238000013138 pruning Methods 0.000 claims abstract description 27
- 238000005070 sampling Methods 0.000 claims description 46
- 230000011218 segmentation Effects 0.000 claims description 18
- 238000001514 detection method Methods 0.000 claims description 12
- 238000012545 processing Methods 0.000 claims description 12
- 210000004351 coronary vessel Anatomy 0.000 claims description 10
- 238000003384 imaging method Methods 0.000 claims description 10
- 230000002159 abnormal effect Effects 0.000 claims description 6
- 238000009826 distribution Methods 0.000 claims description 6
- 238000011156 evaluation Methods 0.000 claims description 6
- 230000002452 interceptive effect Effects 0.000 claims description 6
- 238000002583 angiography Methods 0.000 claims description 5
- 238000004590 computer program Methods 0.000 claims description 3
- 210000005161 hepatic lobe Anatomy 0.000 claims description 3
- 238000004445 quantitative analysis Methods 0.000 abstract description 2
- 230000006870 function Effects 0.000 description 23
- 238000004458 analytical method Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 3
- 238000003745 diagnosis Methods 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 230000000873 masking effect Effects 0.000 description 2
- 238000005293 physical law Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 241000204795 Muraena helena Species 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000000903 blocking effect Effects 0.000 description 1
- 239000008280 blood Substances 0.000 description 1
- 210000004369 blood Anatomy 0.000 description 1
- 230000036770 blood supply Effects 0.000 description 1
- 238000003759 clinical diagnosis Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000005520 cutting process Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000002059 diagnostic imaging Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 208000019423 liver disease Diseases 0.000 description 1
- 230000003908 liver function Effects 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000004060 metabolic process Effects 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 210000000056 organ Anatomy 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000012549 training Methods 0.000 description 1
- 210000003462 vein Anatomy 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/194—Segmentation; Edge detection involving foreground-background segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10072—Tomographic images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10072—Tomographic images
- G06T2207/10088—Magnetic resonance imaging [MRI]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20112—Image segmentation details
- G06T2207/20132—Image cropping
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30056—Liver; Hepatic
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30101—Blood vessel; Artery; Vein; Vascular
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30204—Marker
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiology & Medical Imaging (AREA)
- Quality & Reliability (AREA)
- Image Processing (AREA)
Abstract
The invention provides a liver blood vessel extraction method, a system and a computer readable storage medium based on global automatic growth, wherein the liver blood vessel extraction method comprises the following steps: acquiring a medical image, and segmenting the medical image to obtain a liver region image; calibrating a starting point of a hepatic vascular tree in the hepatic region image, simulating the growth process of the hepatic vascular tree based on a cost measurement function and a fast random tree expansion algorithm, and evaluating the total growth energy of the hepatic vascular tree according to the cost measurement function in the growth process; and removing unreasonable connection parts or branch parts in the growth process of the hepatic vascular tree based on a pruning algorithm, and outputting an extraction result of the hepatic vascular tree when the total growth energy tends to be stable and keeps extremely small. By adopting the technical scheme, the important requirements of clinical application can be met, and the quantitative analysis and further exploration of the internal rule of blood vessel growth can be promoted.
Description
Technical Field
The invention relates to the field of medical image processing, in particular to a liver blood vessel extraction method and system based on global automatic growth and a computer readable storage medium.
Background
The liver is an important organ of human metabolism, and the prevalence rate of the liver is high throughout the year. The liver vessel tree spreads over the whole liver of a human body, and in order to ensure safe implementation of a disease diagnosis and treatment scheme, information such as a mutual position relationship between a focus and peripheral vessel tissues, a blood supply condition, a vein structure of the vessel tree and the like needs to be judged and made clear. From the perspective of actual clinical diagnosis and application, accurate analysis of the liver vessel tree and accurate definition of the topological structure of the liver vessel tree have important values for determining and implementing liver disease diagnosis and treatment schemes, and have key guiding significance for further liver function division, vessel watershed analysis, liver registration and the like according to vessel structure information. The blood vessel tree has high structural complexity, and in addition, the noise of image data has obvious influence on the imaging effect of the tiny blood vessels, the local variability of the blood vessels, poor visibility of the blood vessel endings and the like, so that the automatic blood vessel extraction is always an image processing work with a high difficulty coefficient for many years.
From the biological evolution perspective, the hepatic vascular tree is a very representative biological vascular transmission network, follows the law of murry and the principle of energy minimization, and belongs to a typical biological vascular system satisfying the optimized transmission. The growth, bifurcation and expansion of blood vessels satisfy the principle of energy minimization, where energy is a non-negative function induced by an objective subject. Therefore, how to combine the energy minimum principle to solve the problem of solving and optimizing the non-negative function related to the blood vessel, develop and fuse the blood vessel growth information as a novel image processing method of prior, and explore and model the scientific mechanism of the blood vessel growth is the long-term research focus in the field.
The existing automatic extraction method of the liver vessel tree generally comprises the steps of firstly carrying out vessel segmentation, namely searching for the boundary of vessel tissues, and then carrying out related work of three-dimensional reconstruction, central line extraction, topological structure construction and the like, wherein the situations of vessel under-segmentation and vessel discontinuity usually occur, and the subsequent analysis work of the topological structure of the vessel tree is directly influenced. In addition, for the application of deep learning to vessel segmentation, a large number of vessel training sets need to be drawn, the workload is huge, and meanwhile, in the process of drawing data, due to the high complexity of vessels, a high-quality data set is difficult to obtain, so that the accuracy and stability of the liver vessel tree extraction result based on the deep learning method are difficult to guarantee.
Therefore, a novel liver blood vessel extraction method based on global automatic growth is needed, which can fuse blood vessel growth energy information, so that the extraction result of the blood vessel tree is continuous, topological and better follows the satisfied physical law.
Disclosure of Invention
In order to overcome the technical defects, the invention aims to provide a liver blood vessel extraction method, a system and a computer readable storage medium based on global automatic growth, which can solve the important requirements of clinical application and also contribute to the quantitative analysis and further research of the intrinsic rules of blood vessel growth.
The invention discloses a liver blood vessel extraction method based on global automatic growth, which comprises the following steps:
acquiring a medical image, and segmenting the medical image to obtain a liver region image;
calibrating a starting point of a hepatic vascular tree in the hepatic region image, simulating the growth process of the hepatic vascular tree based on a cost measurement function and a fast random tree expansion algorithm, and evaluating the total growth energy of the hepatic vascular tree according to the cost measurement function in the growth process;
and removing unreasonable connection parts or branch parts in the growth process of the hepatic vascular tree based on a pruning algorithm, and outputting an extraction result of the hepatic vascular tree when the total growth energy tends to be stable and keeps extremely small.
Preferably, the step of acquiring a medical image and segmenting the medical image to obtain an image of the liver region comprises:
acquiring a medical image comprising a coronary image or a magnetic resonance angiography image;
aiming at an image sequence of a coronary artery image or a magnetic resonance blood vessel imaging image, an interactive lobe rough segmentation algorithm is adopted, an area in the liver in a medical image is calibrated to be the foreground of the medical image, and an area outside the liver is the background of the medical image;
a segmentation algorithm is performed on the medical image and masked to obtain an image of the liver region.
Preferably, the step of calibrating the starting point of the hepatic vascular tree in the liver region image, simulating the growth process of the hepatic vascular tree based on a cost metric function and a fast random tree expansion algorithm, and estimating the total growth energy of the hepatic vascular tree according to the cost metric function in the growth process comprises the following steps:
observing an image sequence of the liver region image to mark a starting point of a liver blood vessel tree in the liver region image;
initializing a graph structure G ═ (V, E), wherein the graph structure G represents a growth tree of a global auto-growth liver vessel tree, V represents a node of the growth tree, and E represents an edge of the growth tree;
according to the pixel distribution of the graph structure G, sampling randomly to obtain a sampling point r, and searching to obtain a nearest point c of the sampling point r in the graph structure G;
carrying out connection conflict detection on the sampling point r and the nearest point c, returning to the previous step if connection conflict exists, forming a connection set if connection conflict does not exist, and continuing to the next step;
searching the neighborhood of the nearest neighbor point c to obtain a neighborhood point set, traversing in the neighborhood point set, and performing discrete expression according to the following cost measurement function: ∈ ═ i (x) dl (x).
And calculating and comparing cost measurement of the point in the connecting line set and the nearest point c, thereby updating a graph structure G, wherein epsilon represents the cost required by connecting the sampling point r and the nearest point c, l (x) represents a connecting line subdivision difference value, x is a point for subdivision interpolation, and I (x) is the corresponding image gray value.
Preferably, the step of removing the unreasonable connection part or the bifurcation part in the growth process of the hepatic vessel tree based on the pruning algorithm, and outputting the extraction result of the hepatic vessel tree when the total growth energy tends to be stable and remains extremely small, comprises:
carrying out statistics and evaluation on the mode, the average and the cost measurement of pixel values on the connecting lines in the connecting line set in the updated graph structure G to form a statistical result, carrying out pruning operation on the connecting lines with abnormal statistical results based on a pruning algorithm, removing the corresponding sampling points r, the nearest points c and the connecting lines, and updating the graph structure G again;
when the cost measure of the whole global automatically-grown graph structure G is in a fluctuation state, repeating: and (3) carrying out connection conflict detection on the sampling point r and the nearest point c, returning to the previous step if connection conflict exists, forming a connection set if connection conflict does not exist, continuing the next step and the subsequent steps, and after multiple iterations, outputting the extraction result of the hepatic vessel tree until the cost measurement of the graph structure G is converged.
The invention also discloses a liver blood vessel extraction system based on global automatic growth, which comprises the following components:
the acquisition module acquires a medical image and segments the medical image to acquire a liver region image;
the calibration module is used for calibrating the starting point of the hepatic vascular tree in the liver region image, simulating the growth process of the hepatic vascular tree based on a cost measurement function and a fast random tree expansion algorithm, and evaluating the total growth energy of the hepatic vascular tree according to the cost measurement function in the growth process;
and the output module is used for removing the unreasonable connection part or the bifurcation part in the growth process of the liver vessel tree based on the pruning algorithm, and outputting the extraction result of the liver vessel tree when the total growth energy tends to be stable and keeps extremely small.
Preferably, the acquisition module comprises:
an acquisition unit that acquires a medical image including a coronary artery image or a magnetic resonance blood vessel imaging image;
the segmentation unit is used for calibrating an area in the liver in the medical image into the foreground of the medical image and calibrating an area outside the liver into the background of the medical image by adopting an interactive liver lobe rough segmentation algorithm aiming at an image sequence of a coronary artery image or a magnetic resonance blood vessel imaging image;
and the processing unit is used for executing a segmentation algorithm on the medical image and performing mask processing to obtain a liver region image.
Preferably, the calibration unit comprises:
the calibration unit is used for observing the image sequence of the liver region image so as to calibrate the starting point of the liver blood vessel tree in the liver region image;
an initialization unit, which initializes a graph structure G ═ (V, E), the graph structure G represents a growth tree of a liver blood vessel tree which is grown automatically in the whole, wherein V represents a node of the growth tree, and E represents an edge of the growth tree;
the sampling unit is used for randomly sampling to obtain sampling points r according to the pixel distribution of the graph structure G and searching to obtain the nearest points c of the sampling points r in the graph structure G;
the detection unit is used for detecting the connection conflict between the sampling point r and the nearest point c, returning to the previous step if the connection conflict exists, forming a connection set if the connection conflict does not exist, and continuing to the next step;
and the updating unit is used for searching the neighborhood of the nearest neighbor point c to obtain a neighborhood point set, traversing in the neighborhood point set and discretely expressing the cost metric function according to the following steps: ∈ ═ i (x) dl (x).
And calculating and comparing cost measurement of the point in the connecting line set and the nearest point c, thereby updating a graph structure G, wherein epsilon represents the cost required by connecting the sampling point r and the nearest point c, l (x) represents a connecting line subdivision difference value, x is a point for subdivision interpolation, and I (x) is the corresponding image gray value.
Preferably, the output module includes:
the pruning unit is used for carrying out statistics and evaluation on the mode, the average and the cost measurement of the pixel values on the connecting lines in the connecting line set in the updated graph structure G to form a statistical result, carrying out pruning operation on the connecting lines with abnormal statistical result based on a pruning algorithm, removing the corresponding sampling points r, the nearest points c and the connecting lines, and updating the graph structure G again;
and the iteration unit is used for repeating the following steps when the cost metric of the whole global automatic growth graph structure G is in a fluctuation state: and (3) carrying out connection conflict detection on the sampling point r and the nearest point c, returning to the previous step if connection conflict exists, forming a connection set if connection conflict does not exist, continuing the next step and the subsequent steps, and after multiple iterations, outputting the extraction result of the hepatic vessel tree until the cost measurement of the graph structure G is converged.
The invention also discloses a computer readable storage medium having stored thereon a computer program which, when executed by a processor, performs the steps as described above.
After the technical scheme is adopted, compared with the prior art, the method has the following beneficial effects:
1. the completely unnecessary calculation domain space can be reduced, and the algorithm efficiency is optimized;
2. aiming at the fine tubular and multi-branched tissues such as a blood vessel tree, the method is different from the realization process of searching the boundary of the blood vessel to the inside in the prior art, and can realize better following of the growth rule of the blood vessel and guarantee the continuity of the blood vessel by the realization process of extracting the central skeleton of the blood vessel to the boundary;
3. the automatic degree is high, and the calculation result is more reasonable and stable.
Drawings
FIG. 1 is a schematic flow chart of a method for extracting hepatic blood vessels according to a preferred embodiment of the present invention;
FIG. 2 is a schematic illustration of a medical image in accordance with a preferred embodiment of the present invention;
FIG. 3 is a schematic illustration of an image of a liver region in accordance with a preferred embodiment of the present invention;
fig. 4 is a diagram illustrating a calculation manner of a cost metric function according to a preferred embodiment of the present invention.
Detailed Description
The advantages of the invention are further illustrated in the following description of specific embodiments in conjunction with the accompanying drawings.
Reference will now be made in detail to the exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, like numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the exemplary embodiments below are not intended to represent all implementations consistent with the present disclosure. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present disclosure, as detailed in the appended claims.
The terminology used in the present disclosure is for the purpose of describing particular embodiments only and is not intended to be limiting of the disclosure. As used in this disclosure and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It should also be understood that the term "and/or" as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items.
It is to be understood that although the terms first, second, third, etc. may be used herein to describe various information, such information should not be limited to these terms. These terms are only used to distinguish one type of information from another. For example, first information may also be referred to as second information, and similarly, second information may also be referred to as first information, without departing from the scope of the present disclosure. The word "if" as used herein may be interpreted as "at … …" or "when … …" or "in response to a determination", depending on the context.
In the description of the present invention, it is to be understood that the terms "longitudinal", "lateral", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", and the like, indicate orientations or positional relationships based on those shown in the drawings, and are used merely for convenience of description and for simplicity of description, and do not indicate or imply that the referenced devices or elements must have a particular orientation, be constructed in a particular orientation, and be operated, and thus, are not to be construed as limiting the present invention.
In the description of the present invention, unless otherwise specified and limited, it is to be noted that the terms "mounted," "connected," and "connected" are to be interpreted broadly, and may be, for example, a mechanical connection or an electrical connection, a communication between two elements, a direct connection, or an indirect connection via an intermediate medium, and specific meanings of the terms may be understood by those skilled in the art according to specific situations.
In the following description, suffixes such as "module", "component", or "unit" used to denote elements are used only for facilitating the explanation of the present invention, and have no specific meaning in themselves. Thus, "module" and "component" may be used in a mixture.
Referring to fig. 1, a liver blood vessel extraction method based on global autoregression in accordance with a preferred embodiment of the present invention includes the following steps:
s100: acquiring a medical image, and segmenting the medical image to obtain a liver region image;
referring to fig. 2, a rough segmentation is performed on any one of the medical images of the liver slice image, and the obtained liver region image is the portion shown in fig. 3, which includes the liver portion.
S200: calibrating a starting point of a hepatic vascular tree in the hepatic region image, simulating the growth process of the hepatic vascular tree based on a cost measurement function and a fast random tree expansion algorithm, and evaluating the total growth energy of the hepatic vascular tree according to the cost measurement function in the growth process;
the starting point (namely the root of the tree shape or the inflow port understood as liver blood) of a liver vessel tree (also called a vessel or a hierarchical vessel tree) in the liver region image is calibrated on the liver region image, and the growth process of the liver vessel book from the starting point is simulated by combining a cost measurement function and a fast random tree expansion algorithm. It will be appreciated that the objective of the cost metric function (or called cost function, loss function) is to minimize the solution under constrained conditions. The fast random spanning tree algorithm (or called fast spanning random tree) is a path planning algorithm under a known map, and in the embodiment, all scalable paths from a starting point can be generated quickly by using the algorithm. In this embodiment, all possible growth paths of the hepatic vessel tree are simulated, so that the growth process of the hepatic vessel tree is simulated. And meanwhile, evaluating the total growing energy caused by the growing track of each hepatic vessel tree in the growing process according to the cost metric function.
S300: removing unreasonable connection parts or branch parts in the growth process of the hepatic vascular tree based on a pruning algorithm, and outputting the extraction result of the hepatic vascular tree when the total growth energy tends to be stable and keeps extremely small
The pruning algorithm is as follows: in the process of solving the practical problem, some points and child nodes thereof obviously do not accord with the theme, and the search is not needed at all. Then the spatio-temporal complexity is optimized by adding a decision condition such that this subtree is not entered during the search. This optimization technique is called pruning algorithm because it is very much like cutting redundant branches on a tree. The pruning algorithm is utilized to remove unreasonable connection parts or bifurcation parts in the growth process of the liver blood vessel tree until the total growth energy tends to be stable and is kept extremely small, which means that the growth process of the liver blood vessel book firstly conforms to natural laws, and secondly, the cost of the growth path is minimum, the liver blood vessel tree should conform to the actual liver blood vessel tree of the focus, so that the liver blood vessel extraction result basically conforming to the actual condition of the focus can be output.
By fusing image information and growth energy information and simulating the global automatic growth of the blood vessel tree based on the cost measurement function, the extraction result is more continuous, topology-preserving and follows the physical law of blood vessel growth expansion.
In a preferred embodiment, step S100 includes:
s110: acquiring a medical image comprising a coronary image or a magnetic resonance angiography image;
the coronary artery image is a CTA image, and the magnetic resonance angiography image is an MRA image, wherein part of the magnetic resonance angiography image comprises a liver region.
S120: aiming at an image sequence of a coronary artery image or a magnetic resonance blood vessel imaging image, an interactive lobe rough segmentation algorithm is adopted, an area in the liver in a medical image is calibrated to be the foreground of the medical image, and an area outside the liver is the background of the medical image;
the image sequence may also be referred to as DICOM sequence, which is the key to medical imaging and stores all information for a single diagnosis (patient information + image data), and all information can be obtained by reading and parsing. The DICOM sequence in this embodiment has a size of 512 × 512 × 160.
S130: and executing a segmentation algorithm on the medical image, and performing mask processing to obtain a liver region image.
Mask processing, also called image masking, refers to the process of masking (wholly or partially) an image to be processed with a selected image, graphic or object to control the area or process of image processing, so as to obtain an image of the liver area with the background removed.
Preferably or optionally, step S200 comprises:
s210: observing an image sequence of the liver region image to mark a starting point of a liver blood vessel tree in the liver region image;
this step S210 can be done manually, and the calibrated point is denoted xinit.
S220: initializing a graph structure G ═ (V, E), wherein the graph structure G represents a growth tree of a global auto-growth liver vessel tree, V represents a node of the growth tree, and E represents an edge of the growth tree;
it can be understood that the graph structure G is initialized without the image gridding spatial constraint, i.e. the graph structure G is borderless in the initial state.
S230: according to the pixel distribution of the graph structure G, sampling randomly to obtain a sampling point r, and searching to obtain a nearest point c of the sampling point r in the graph structure G;
it is understood that the nearest pixel point c may be the nearest pixel point adjacent to the sampling point r, or the first pixel point after crossing the blocking area.
S240: carrying out connection conflict detection on the sampling point r and the nearest point c, returning to the previous step if connection conflict exists, forming a connection set if connection conflict does not exist, and continuing to the next step;
and detecting connection conflict, namely detecting whether the two nodes can not be directly connected or not so as to represent the correlation and the connection smoothness of the sampling point r and the nearest point c.
S250: searching the neighborhood of the nearest neighbor point c to obtain a neighborhood point set, traversing in the neighborhood point set, and performing discrete expression according to the following cost measurement function: ∈ ═ i (x) dl (x).
And calculating and comparing cost measurement of the point in the connecting line set and the nearest point c, thereby updating a graph structure G, wherein epsilon represents the cost required by connecting the sampling point r and the nearest point c, l (x) represents a connecting line subdivision difference value, x is a point for subdivision interpolation, and I (x) is the corresponding image gray value.
Referring to fig. 4, in this step, the graph structure G is updated in a connection manner with a smaller update cost, so that the graph structure G is always kept extremely small in the iteration process.
Further, step S300 includes:
s310: carrying out statistics and evaluation on the mode, the average and the cost measurement of pixel values on the connecting lines in the connecting line set in the updated graph structure G to form a statistical result, carrying out pruning operation on the connecting lines with abnormal statistical results based on a pruning algorithm, removing the corresponding sampling points r, the nearest points c and the connecting lines, and updating the graph structure G again;
s320: when the cost measure of the whole global automatically-grown graph structure G is in a fluctuating state, i.e. the cost of the graph structure G does not tend to be stable, the following steps are repeated: and (3) carrying out connection conflict detection on the sampling point r and the nearest point c, returning to the previous step if connection conflict exists, forming a connection set if connection conflict does not exist, continuing the next step and the subsequent steps, and after multiple iterations, outputting the extraction result of the hepatic vessel tree until the cost measurement of the graph structure G is converged.
The invention also discloses a liver blood vessel extraction system based on global automatic growth, which comprises the following steps: the acquisition module acquires a medical image and segments the medical image to acquire a liver region image; the calibration module is used for calibrating the starting point of the hepatic vascular tree in the liver region image, simulating the growth process of the hepatic vascular tree based on a cost measurement function and a fast random tree expansion algorithm, and evaluating the total growth energy of the hepatic vascular tree according to the cost measurement function in the growth process; and the output module is used for removing the unreasonable connection part or the bifurcation part in the growth process of the liver vessel tree based on the pruning algorithm, and outputting the extraction result of the liver vessel tree when the total growth energy tends to be stable and keeps extremely small.
Preferably, the acquisition module comprises: an acquisition unit that acquires a medical image including a coronary artery image or a magnetic resonance blood vessel imaging image; the segmentation unit is used for calibrating an area in the liver in the medical image into the foreground of the medical image and calibrating an area outside the liver into the background of the medical image by adopting an interactive liver lobe rough segmentation algorithm aiming at an image sequence of a coronary artery image or a magnetic resonance blood vessel imaging image; and the processing unit is used for executing a segmentation algorithm on the medical image and performing mask processing to obtain a liver region image.
Preferably, the calibration unit comprises: the calibration unit is used for observing the image sequence of the liver region image so as to calibrate the starting point of the liver blood vessel tree in the liver region image; an initialization unit, which initializes a graph structure G ═ (V, E), the graph structure G represents a growth tree of a liver blood vessel tree which is grown automatically in the whole, wherein V represents a node of the growth tree, and E represents an edge of the growth tree; the sampling unit is used for randomly sampling to obtain a sampling point r according to the pixel distribution of the graph structure G, and searching to obtain a nearest point c of the sampling point r in the graph structure G; the detection unit is used for detecting the connection conflict between the sampling point r and the nearest point c, returning to the previous step if the connection conflict exists, forming a connection set if the connection conflict does not exist, and continuing to the next step; and the updating unit is used for searching the neighborhood of the nearest neighbor point c to obtain a neighborhood point set, traversing in the neighborhood point set and discretely expressing the cost metric function according to the following steps: e ═ i (x) dl (x), a cost metric is calculated and compared between points in the set of lines and the nearest points c, thereby updating the graph structure G, where e represents the cost required to line the sample points r and nearest points c, l (x) represents line subdivision differences, x is the point of subdivision interpolation, and i (x) is the corresponding image gray value size.
Preferably, the output module includes: the pruning unit is used for carrying out statistics and evaluation on the mode, the average and the cost measurement of the pixel values on the connecting lines in the connecting line set in the updated graph structure G to form a statistical result, carrying out pruning operation on the connecting lines with abnormal statistical result based on a pruning algorithm, removing the corresponding sampling points r, the nearest points c and the connecting lines, and updating the graph structure G again; and the iteration unit is used for repeating the following steps when the cost metric of the whole global automatic growth graph structure G is in a fluctuation state: and (3) carrying out connection conflict detection on the sampling point r and the nearest point c, returning to the previous step if connection conflict exists, forming a connection set if connection conflict does not exist, continuing the next step and the subsequent steps, and after multiple iterations, outputting the extraction result of the hepatic vessel tree until the cost measurement of the graph structure G is converged.
The invention also discloses a computer readable storage medium having stored thereon a computer program which, when executed by a processor, performs the steps as described above.
It should be noted that the embodiments of the present invention have been described in terms of preferred embodiments, and not by way of limitation, and that those skilled in the art can make modifications and variations of the embodiments described above without departing from the spirit of the invention.
Claims (9)
1. A liver blood vessel extraction method based on global automatic growth is characterized by comprising the following steps:
acquiring a medical image, and segmenting the medical image to obtain a liver region image;
calibrating the starting point of the hepatic vascular tree in the hepatic region image, simulating the growth process of the hepatic vascular tree based on a cost measurement function and a fast random tree expansion algorithm, and evaluating the total growth energy of the hepatic vascular tree according to the cost measurement function in the growth process;
and removing unreasonable connection parts or branch parts in the growth process of the hepatic vascular tree based on a pruning algorithm, and outputting an extraction result of the hepatic vascular tree when the total growth energy tends to be stable and keeps extremely small.
2. The liver vessel extraction method of claim 1, wherein the step of obtaining a medical image and segmenting the medical image to obtain an image of the liver region comprises:
acquiring a medical image comprising a coronary image or a magnetic resonance angiography image;
aiming at an image sequence of a coronary artery image or a magnetic resonance blood vessel imaging image, an interactive lobe rough segmentation algorithm is adopted, an area in the liver in a medical image is calibrated to be the foreground of the medical image, and an area outside the liver is the background of the medical image;
and executing a segmentation algorithm on the medical image, and performing mask processing to obtain a liver region image.
3. The hepatic vessel extraction method according to claim 1, wherein the step of calibrating the starting point of the hepatic vessel tree in the liver region image, simulating the growth process of the hepatic vessel tree based on a cost metric function and a fast random spanning tree algorithm, and estimating the total energy of the hepatic vessel tree growth according to the cost metric function during the growth process comprises: observing an image sequence of the liver region image to mark a starting point of a liver blood vessel tree in the liver region image;
initializing a graph structure G ═ (V, E), wherein the graph structure G represents a growth tree of a global auto-growing hepatic vascular tree, V represents a node of the growth tree, and E represents an edge of the growth tree;
according to the pixel distribution of the graph structure G, sampling randomly to obtain a sampling point r, and searching to obtain a nearest point c of the sampling point r in the graph structure G;
carrying out connection conflict detection on the sampling point r and the nearest point c, returning to the previous step if connection conflict exists, forming a connection set if connection conflict does not exist, and continuing to the next step;
searching the neighborhood of the nearest neighbor point c to obtain a neighborhood point set, traversing in the neighborhood point set, and performing discrete expression according to the following cost measurement function: epsilon ═ jeep i (x) dl (x)
And calculating and comparing cost measurement of the point in the connecting line set and the nearest point c, thereby updating a graph structure G, wherein epsilon represents the cost required by connecting the sampling point r and the nearest point c, l (x) represents a connecting line subdivision difference value, x is a point for subdivision interpolation, and I (x) is the corresponding image gray value.
4. The hepatic vessel extraction method according to claim 3, wherein the step of removing the unreasonable connection part or the bifurcation part in the growth process of the hepatic vessel tree based on the pruning algorithm, and outputting the extraction result of the hepatic vessel tree when the total growth energy tends to be stable and remains extremely small, comprises:
carrying out statistics and evaluation on the mode, the average and the cost measurement of pixel values on the connecting lines in the connecting line set in the updated graph structure G to form a statistical result, carrying out pruning operation on the connecting lines with abnormal statistical results based on a pruning algorithm, removing the corresponding sampling points r, the nearest points c and the connecting lines, and updating the graph structure G again;
when the cost measure of the whole global automatically-grown graph structure G is in a fluctuation state, repeating: and (3) carrying out connection conflict detection on the sampling point r and the nearest point c, returning to the previous step if connection conflict exists, forming a connection set if connection conflict does not exist, continuing the next step and the subsequent steps, and after multiple iterations, outputting the extraction result of the hepatic vessel tree until the cost measurement of the graph structure G is converged.
5. A liver vessel extraction system based on global automatic growth, comprising:
the acquisition module acquires a medical image and segments the medical image to acquire a liver region image;
the calibration module is used for calibrating the starting point of the hepatic vascular tree in the liver region image, simulating the growth process of the hepatic vascular tree based on a cost measurement function and a fast random tree expansion algorithm, and evaluating the total growth energy of the hepatic vascular tree according to the cost measurement function in the growth process;
and the output module is used for removing the unreasonable connection part or the bifurcation part in the growth process of the liver vessel tree based on the pruning algorithm, and outputting the extraction result of the liver vessel tree when the total growth energy tends to be stable and keeps extremely small.
6. The hepatic vessel extraction system of claim 5, wherein the acquisition module comprises:
an acquisition unit that acquires a medical image including a coronary artery image or a magnetic resonance blood vessel imaging image;
the segmentation unit is used for calibrating an area in the liver in the medical image into the foreground of the medical image and calibrating an area outside the liver into the background of the medical image by adopting an interactive liver lobe rough segmentation algorithm aiming at an image sequence of a coronary artery image or a magnetic resonance blood vessel imaging image;
and the processing unit is used for executing a segmentation algorithm on the medical image and performing mask processing to obtain a liver region image.
7. The hepatic vascular extraction system of claim 5, wherein the calibration unit comprises:
the calibration unit is used for observing the image sequence of the liver region image so as to calibrate the starting point of the liver blood vessel tree in the liver region image;
an initialization unit which initializes a graph structure G which represents a growth tree of a liver blood vessel tree of a global automatic growth, wherein V represents a node of the growth tree, and E represents an edge of the growth tree;
the sampling unit is used for randomly sampling to obtain a sampling point r according to the pixel distribution of the graph structure G, and searching to obtain a nearest point c of the sampling point r in the graph structure G;
the detection unit is used for detecting the connection conflict between the sampling point r and the nearest point c, returning to the previous step if the connection conflict exists, forming a connection set if the connection conflict does not exist, and continuing to the next step;
the updating unit is used for searching the neighborhood of the nearest neighbor point c to obtain a neighborhood point set, traversing in the neighborhood point set and discretely expressing the cost metric function according to the following steps: epsilon ═ jeep i (x) dl (x)
And calculating and comparing cost measurement of the point in the connecting line set and the nearest point c, thereby updating a graph structure G, wherein epsilon represents the cost required by connecting the sampling point r and the nearest point c, l (x) represents a connecting line subdivision difference value, x is a point for subdivision interpolation, and I (x) is the corresponding image gray value.
8. The hepatic vessel extraction system of claim 7, wherein the output module comprises:
the pruning unit is used for carrying out statistics and evaluation on the mode, the average and the cost measurement of the pixel values on the connecting lines in the connecting line set in the updated graph structure G to form a statistical result, carrying out pruning operation on the connecting lines with abnormal statistical result based on a pruning algorithm, removing the corresponding sampling points r, the nearest points c and the connecting lines, and updating the graph structure G again;
and the iteration unit is used for repeating the following steps when the cost metric of the whole global automatic growth graph structure G is in a fluctuation state: and (3) carrying out connection conflict detection on the sampling point r and the nearest point c, returning to the previous step if connection conflict exists, forming a connection set if connection conflict does not exist, continuing the next step and the subsequent steps, and after multiple iterations, outputting the extraction result of the hepatic vessel tree until the cost measurement of the graph structure G is converged.
9. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of any of claims 1-4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210513891.5A CN114820554A (en) | 2022-05-11 | 2022-05-11 | Liver blood vessel extraction method, system and computer readable storage medium based on global automatic growth |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210513891.5A CN114820554A (en) | 2022-05-11 | 2022-05-11 | Liver blood vessel extraction method, system and computer readable storage medium based on global automatic growth |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114820554A true CN114820554A (en) | 2022-07-29 |
Family
ID=82512650
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210513891.5A Pending CN114820554A (en) | 2022-05-11 | 2022-05-11 | Liver blood vessel extraction method, system and computer readable storage medium based on global automatic growth |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114820554A (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103218776A (en) * | 2013-03-07 | 2013-07-24 | 天津大学 | Non-local depth image super-resolution rebuilding method based on minimum spanning tree (MST) |
| CN103268603A (en) * | 2013-05-08 | 2013-08-28 | 同济大学 | Liver image segregation method based on hierarchy vessel tree division |
| CN104282009A (en) * | 2013-07-02 | 2015-01-14 | 上海联影医疗科技有限公司 | Coronary artery extraction method |
| CN104835112A (en) * | 2015-05-07 | 2015-08-12 | 厦门大学 | Liver multi-phase CT image fusion method |
| CN109996495A (en) * | 2016-11-22 | 2019-07-09 | 皇家飞利浦有限公司 | For the vascular tree standardization of biophysics simulation and/or for the extended simulation of pruned part |
| WO2022000727A1 (en) * | 2020-06-29 | 2022-01-06 | 苏州润迈德医疗科技有限公司 | Ct sequence image-based coronary artery tree and coronary artery entry point obtaining method and system |
-
2022
- 2022-05-11 CN CN202210513891.5A patent/CN114820554A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103218776A (en) * | 2013-03-07 | 2013-07-24 | 天津大学 | Non-local depth image super-resolution rebuilding method based on minimum spanning tree (MST) |
| CN103268603A (en) * | 2013-05-08 | 2013-08-28 | 同济大学 | Liver image segregation method based on hierarchy vessel tree division |
| CN104282009A (en) * | 2013-07-02 | 2015-01-14 | 上海联影医疗科技有限公司 | Coronary artery extraction method |
| CN104835112A (en) * | 2015-05-07 | 2015-08-12 | 厦门大学 | Liver multi-phase CT image fusion method |
| CN109996495A (en) * | 2016-11-22 | 2019-07-09 | 皇家飞利浦有限公司 | For the vascular tree standardization of biophysics simulation and/or for the extended simulation of pruned part |
| WO2022000727A1 (en) * | 2020-06-29 | 2022-01-06 | 苏州润迈德医疗科技有限公司 | Ct sequence image-based coronary artery tree and coronary artery entry point obtaining method and system |
Non-Patent Citations (1)
| Title |
|---|
| 梁礼明等: "基于多特征融合的有监督视网膜血管提取", 计算机学报, vol. 41, no. 11, 30 November 2018 (2018-11-30) * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111640120B (en) | Pancreas CT automatic segmentation method based on significance dense connection expansion convolution network | |
| CN107563983B (en) | Image processing method and medical imaging device | |
| CN107369160B (en) | Choroid neogenesis blood vessel segmentation algorithm in OCT image | |
| CN112508965B (en) | Automatic outline sketching system for normal organs in medical image | |
| CN113744272B (en) | Automatic cerebral artery delineation method based on deep neural network | |
| US9600890B2 (en) | Image segmentation apparatus, medical image device and image segmentation method | |
| JP4248399B2 (en) | Automatic branch labeling method | |
| WO2006000953A1 (en) | Displaying a tracheobronchial tree | |
| US8532358B2 (en) | System and method for segmentation of MR flow data using fluid dynamics and tracking | |
| CN115409859A (en) | Coronary artery blood vessel image segmentation method and device, storage medium and terminal | |
| CN115908297A (en) | Topology knowledge-based blood vessel segmentation modeling method in medical image | |
| Nowinski et al. | A 3D model of human cerebrovasculature derived from 3T magnetic resonance angiography | |
| CN111260704B (en) | Vascular structure 3D/2D rigid registration method and device based on heuristic tree search | |
| CN111861994B (en) | Coronary artery wall image segmentation method, system, storage medium and computer equipment | |
| EP3951433A1 (en) | Line structure extraction device, method and program, and learned model | |
| CN114820554A (en) | Liver blood vessel extraction method, system and computer readable storage medium based on global automatic growth | |
| CN107705289A (en) | A kind of blood vessel simulation method for reconstructing based on matrix topology | |
| CN105654527A (en) | Magnetic resonance imaging reconstruction method and device based on structural dictionary learning | |
| CN114519722A (en) | Carotid artery extraction method based on convolutional neural network | |
| Jiang et al. | An automatic and fast centerline extraction algorithm for virtual colonoscopy | |
| CN111242915B (en) | Vascular 3D/2D registration method and device based on Monte Carlo tree search | |
| CN107689043B (en) | Method for acquiring blood vessel section terminal node and branch node | |
| CN112446893A (en) | Contour segmentation method and device for liver image | |
| CN112862731A (en) | Full-automatic blood vessel extraction method of TOF image | |
| CN113269816A (en) | Regional progressive brain image elastic registration method and system |
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 |