CN115409758A - Method for comprehensively and quantitatively evaluating angiostenosis from contrast images - Google Patents
Method for comprehensively and quantitatively evaluating angiostenosis from contrast images Download PDFInfo
- Publication number
- CN115409758A CN115409758A CN202110584158.8A CN202110584158A CN115409758A CN 115409758 A CN115409758 A CN 115409758A CN 202110584158 A CN202110584158 A CN 202110584158A CN 115409758 A CN115409758 A CN 115409758A
- Authority
- CN
- China
- Prior art keywords
- stenosis
- blood vessel
- image
- vessel
- tree
- 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
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
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/181—Segmentation; Edge detection involving edge growing; involving edge linking
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Geometry (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiology & Medical Imaging (AREA)
- Quality & Reliability (AREA)
- Apparatus For Radiation Diagnosis (AREA)
Abstract
The invention discloses a method for comprehensively and quantitatively evaluating vascular stenosis from a contrast image, which comprises an image preprocessing module, an automatic stenosis evaluation module and an interactive stenosis evaluation module. The preprocessing module is combined with TOF, UM, CLAHE and Hessian matrix multi-scale enhancement algorithms to improve image quality and highlight vascular structures. The automatic stenosis evaluation module extracts a blood vessel tree framework and a blood vessel segmentation by using a provided novel tracking algorithm, extracts a blood vessel outline by using a CV model, measures the diameter and stenosis degree of each position of the blood vessel tree and detects a stenosis part. The interactive stenosis assessment module selects a target blood vessel section by a user, extracts a skeleton of the target blood vessel section and an extraction contour of a CV model by utilizing a proposed tracking algorithm based on an energy function, measures the diameter and the stenosis degree of each part of the blood vessel section and detects the stenosis part. The invention combines automatic and interactive stenosis assessment, and can perform batch automatic detection on contrast images and quantitative assessment on detail of specific images.
Description
Technical Field
The invention relates to a method for comprehensively and quantitatively evaluating coronary angiography image angiostenosis, which comprises a full-automatic evaluation method and an interactive evaluation method and belongs to the field of medical image processing.
Background
Cardiovascular diseases are currently one of the major chronic diseases recognized worldwide to cause human death. In recent years, the number of cardiovascular diseases and deaths has been increasing, and these diseases are beginning to be treated. Coronary angiography is a common and effective method for diagnosing coronary heart disease, is regarded as the 'gold standard' for diagnosing coronary heart disease, and is widely applied to clinical diagnosis of coronary heart disease. The human arterial blood vessel is composed of soft tissue and is normally not visible under X-rays. However, if a substance opaque to X-rays is injected into the coronary artery and the coronary artery region is irradiated with X-rays, the arterial blood vessel can be developed under X-rays, and a doctor can find the location and extent of the coronary artery stenosis from the image and decide the treatment plan based on the location and extent.
At present, most of assessment methods for determining the position of coronary lesion and judging the coronary stenosis degree are based on subjective measurement of doctors, and a large amount of repeated work not only reduces the work efficiency, but also has more subjectivity, so that the judgment result is not accurate enough. Therefore, in order to improve the efficiency and accuracy of coronary heart disease detection, it is necessary to accurately extract the vascular skeleton and contour from the contrast image and measure the diameter and stenosis degree of each blood vessel by using an image processing technique.
Because the difference between the vascular structure and the characteristics of background gray scale, morphology and the like in the contrast image is obvious, the skeleton and the outline of the blood vessel can be accurately extracted by comprehensively considering the characteristics by using a proper method. And the obtained skeleton and contour information can be used for measuring the diameter and the stenosis degree of each part of the blood vessel so as to accurately evaluate the stenosis of the blood vessel.
However, most of the current methods for evaluating the blood vessel stenosis adopt a single global blood vessel stenosis detection method, and in practical applications, medical staff may need to perform further quantitative analysis on a specific blood vessel segment in the blood vessel tree. Meanwhile, due to the design defects of various technologies, the result accuracy needs to be improved.
Disclosure of Invention
The invention provides a comprehensive quantitative evaluation method for angiostenosis of a contrast image, aiming at the current situations that the coronary angiography manual detection efficiency and the accuracy are low, and the related image processing technology is single and not accurate enough in the actual medical scene, and the comprehensive quantitative evaluation method comprises a full-automatic evaluation method and an interactive evaluation method. In a practical application scene, the method can be used for carrying out efficient and automatic global angiostenosis assessment on the coronary angiography image, medical staff can interactively select a certain blood vessel section and carry out further quantitative analysis in a targeted manner, the two methods are complementary, the methods have quite strong robustness, and the clinical diagnosis and treatment requirements of cardiovascular diseases can be well met. The method mainly comprises three parts of image preprocessing, automatic stenosis assessment and interactive assessment.
The image preprocessing part realizes image denoising, improves image contrast, enhances a tubular structure in the image, obviously increases the difference between the blood vessel tree and the background, and weakens the interference of a non-tubular target in the background.
The automatic stenosis assessment part utilizes a CV model to realize the extraction of the outline of the blood vessel tree, provides a novel tracking algorithm to realize the extraction of a blood vessel framework (a blood vessel central line), the detection of a bifurcation point and the segmentation of the blood vessel, carries out the diameter measurement and the stenosis degree assessment of all parts of the blood vessel on the basis, can assess all the stenosis parts of the whole blood vessel tree, and has no artificial interference in the whole process.
The interactive stenosis assessment part also utilizes a CV model to realize the extraction of the blood vessel tree outline, and after a section of blood vessel is interactively selected by a user, a second tracking algorithm based on an energy function is provided to realize the extraction of the skeleton of the blood vessel section, so that the diameter measurement and the stenosis degree assessment of each part of the blood vessel section are carried out, and all the stenosis parts of the blood vessel section are detected.
The comprehensive quantitative evaluation of the vascular stenosis has the advantages that:
1. the invention realizes the minimization of manual operation during detection. For the automatic stenosis assessment part, the whole process does not need human interference; for interactive evaluation, a user only needs to manually select an initial point and a termination point to determine a target blood vessel section, and other operations are fully automatic;
2. compared with the existing tracking method such as self-adaptive geometric tracking, the two tracking algorithms provided by the invention have higher accuracy and stronger robustness, and can more accurately extract the blood vessel tree skeleton or the specific target blood vessel section skeleton;
3. the invention realizes better balance between the calculation complexity and the evaluation effect, can efficiently detect the vascular stenosis and is convenient for clinical application;
4. the quantitative processing of the invention comprises diameter measurement and stenosis degree evaluation at each position, which can intuitively and accurately present the relative stenosis condition at each position of the blood vessel to the user, and has incomparable advantages compared with manual estimation.
Drawings
FIG. 1 is a flowchart of the automatic stenosis assessment method proposed by the present invention.
FIG. 2 is a flowchart of the interactive stenosis assessment method of the present invention.
Fig. 3 is a schematic diagram of the initial tracking direction proposed by the present invention.
Fig. 4 is a schematic diagram of the tracking direction proposed by the present invention.
FIG. 5 is a schematic diagram of the center line of tracking point adjustment according to the present invention.
Fig. 6 is a schematic diagram of bifurcation point detection proposed by the present invention.
Fig. 7 is a schematic view of a vessel segment according to the present invention.
Fig. 8 is a schematic diagram of the diameter measurement proposed by the present invention.
Detailed Description
The advantages and spirit of the present invention can be further understood by the following detailed description of the invention and the accompanying drawings.
In step S101, coronary angiography image data is read, and the data may be DICOM serial slice images or single two-dimensional angiography images.
Step S102, carrying out image preprocessing on the contrast image, firstly carrying out denoising by using an ROF algorithm, then generating a new enhanced edge image from the old image by using an UM algorithm, further increasing the contrast of the image by using contrast-limited adaptive histogram equalization, and finally highlighting the vascular structure by adopting a multi-scale image enhancement method based on an image Hessian matrix.
The following is the same pre-operation of the automated stenosis assessment as the interactive stenosis assessment component.
And step S103, applying the CV model to the preprocessed image to obtain a blood vessel tree contour.
Step S104, detecting a blood vessel ridge point (local brightness maximum point) in the preprocessed image, and detecting an (x, y) pixel point satisfying the following condition:
the remaining automated stenosis assessment component is as follows.
Step S105, randomly selecting seed points from the obtained ridge pointsAnd performing initial tracking, wherein the initial tracking direction can be obtained from the blood vessel gray scale information near the seed point. Specifically, toAs the center of a circle, the radius isSearching for the gray level maximum point on the circleThe positive tracking direction is expressed as:
after the forward tracking angle is obtained, the backward tracking angle can be at the forward tracking angleIn the opposite direction ofCentered search rangeSearching local maximum point on corresponding arcIn the same way, the tracking of the reverse direction can be obtained。
This step can be represented by fig. 1.
Step S106, tracking from the current point to the next tracking point, which is a main link of the tracking algorithm. The current tracking direction is from the last tracking pointTo the current pointThe direction of (c) determines:
Wherein the content of the first and second substances,is thatIs determined by the gray-scale value of (a),is a given threshold value for the value of the threshold,is thatThe number of the tracking points around the target,is also a threshold.
This process can be represented by figure 2.
The tracking points may be deviated from the center of the blood vessel during the tracking process, so that each tracking point needs to be adjusted to the center line of the blood vessel by the blood vessel contour obtained in step S103. The specific operation is to calculate the normal of the current tracking direction and find the intersection point of the normal and the blood vessel contourCan adjust the tracking pointComprises the following steps:
accordingly, the tracking direction is adjusted to:
this process can be represented by figure 3.
In step S107, bifurcation point detection is required at each tracking point. In particular, in the radianRadius ofThe ridge points are searched in the fan-shaped annular area, and for each ridge point, if the following conditions are met, the ridge point is a bifurcation point:
this process can be represented by figure 4.
Because the diameters of all blood vessel sections (main trunk and branch) of the blood vessel tree are greatly different, the blood vessel section treatment is needed for facilitating the subsequent stenosis assessment. The vessel segmentation is completely based on tracking point sequence information recorded in the tracking process, and a vessel segment is formed between any two adjacent cut-off points (a bifurcation point and a segment tracking stop point) in the tracking point sequence.
This process can be represented by figure 5.
Step S108, according to the adjusted blood vessel skeleton information and the extracted blood vessel contour, the diameter and the stenosis degree of the target blood vessel tracking point can be measured. The measuring method comprises the following steps: for a certain tracking pointIn other words, we utilizeAndthe 3 adjacent tracking points (less than 3 using all tracking points) are fitted to a straight lineFurther, the corresponding passing point is obtainedNormal toAnd taking the distance between the normal line and two intersection points of the blood vessel contour as the diameter of the blood vessel at the point.
This process can be represented by figure 6.
For a certain blood vessel section, traversing all tracking points on the central line of the current blood vessel section from the starting point, and calculating the diameter corresponding to each tracking pointAnd average diameter of all tracking points of the vessel segmentTracing pointsThe degree of stenosis of the blood vessel is:
step S109, measure the diameter marker stenosis at each site of the vessel tree. The operation is as follows:
1. preprocessing the obtained segmentation result, and removing blood vessel segments which are too short and not suitable for evaluation;
2. traversing all the blood vessel sections, and solving the diameter and the stenosis degree of each tracking point on each blood vessel section;
3. given a threshold value T, trace points with a stenosis degree less than T are marked, which is the stenosis of the segment.
The remaining interactive stenosis assessment section follows.
In step S110, the user manually selects a start point and a stop point to determine the target blood vessel segment.
And step S111, finding ridge points adjacent to the start point and the stop point. According to our step S104, a set of initial Ridge Points (RP) is obtained. The start and end points for tracking can be expressed as:
and step S112, tracking. To be provided withThe tracking starts as a seed point. The point on the search arc that maximizes the following energy function value is selected as the next tracking point:
then, the tracking point and the tracking point are judgedIs less than a given threshold. If the distance is within a threshold, we consider the two points to be close enough, tracking can be stopped, otherwise, it continues.
In step S113, a correct tracking route is selected. Step S112 is executed in the forward and backward tracking directions respectively to obtain forward and backward paths, and the end tracking point is selected to be closerThe route of (1).
Step S114, the diameter and the stenosis degree of each tracking point on the blood vessel section are obtained, a threshold value T is given, and the tracking point with the stenosis degree smaller than T is marked, namely the stenosis part of the blood vessel section.
Although the present invention has been described with reference to preferred embodiments, the above examples should not be construed as limiting the scope of the present invention, and any modifications, equivalents and improvements that are within the spirit and principle of the present invention should be included in the scope of the claims of the present invention.
Claims (4)
1. A method for comprehensively and quantitatively evaluating the stenosis of a blood vessel from a contrast image comprises three key modules, wherein the three key modules comprise:
(1) The image preprocessing module is used for denoising an original image by using an ROF algorithm, generating a new enhanced edge image from an old image by using an UM algorithm, increasing the contrast of the image by using contrast-limited adaptive histogram equalization and highlighting the vascular structure in the image by using a multiscale image enhancement method based on an image Hessian matrix;
(2) An automated stenosis assessment module: the method comprises the steps of extracting a blood vessel tree contour by using a CV model, extracting a blood vessel tree framework (blood vessel center line), detecting a bifurcation point and segmenting a blood vessel by using a novel tracking algorithm provided by the invention, measuring the diameter and evaluating the stenosis degree of each part of the blood vessel tree by combining blood vessel framework information and blood vessel contour information, and detecting all the stenosis parts of the whole blood vessel tree;
(3) An interactive stenosis assessment module: the CV model is used for realizing the extraction of the contour of the blood vessel tree, the tracking algorithm based on the energy function provided by the invention is used for realizing the extraction of the skeleton of the target blood vessel section, the diameter measurement and the stenosis degree evaluation of each part of the blood vessel section are carried out by combining the centerline information and the contour information of the blood vessel section, and all the stenosis parts of the blood vessel section are detected.
2. The method of claim 1, wherein the image pre-processing module is capable of improving contrast image quality, enhancing vascular structures, highlighting vascular ridge features, and suppressing non-tubular background.
3. The method of claim 1, wherein the automatic stenosis evaluating module is capable of automatically extracting a contour of the vessel tree and obtaining a plurality of seed points located on a centerline of the vessel, extracting a skeleton of the vessel tree (the centerline of the vessel) from a certain seed point, identifying a bifurcation point of the vessel, segmenting the vessel, and automatically performing diameter measurement and stenosis degree evaluation at each segment of the vessel tree, thereby detecting a location of stenosis of the vessel tree.
4. The method for comprehensive quantitative assessment of vascular stenosis as claimed in claim 1, wherein the interactive stenosis assessment module can perform extraction of the profile of the target vessel segment and extraction of the central line, and automatically perform diameter measurement and stenosis degree assessment at various positions of the vessel segment, thereby detecting the stenosis position of the target vessel segment.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110584158.8A CN115409758A (en) | 2021-05-27 | 2021-05-27 | Method for comprehensively and quantitatively evaluating angiostenosis from contrast images |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110584158.8A CN115409758A (en) | 2021-05-27 | 2021-05-27 | Method for comprehensively and quantitatively evaluating angiostenosis from contrast images |
Publications (1)
Publication Number | Publication Date |
---|---|
CN115409758A true CN115409758A (en) | 2022-11-29 |
Family
ID=84155987
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110584158.8A Pending CN115409758A (en) | 2021-05-27 | 2021-05-27 | Method for comprehensively and quantitatively evaluating angiostenosis from contrast images |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN115409758A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117274502A (en) * | 2023-11-17 | 2023-12-22 | 北京唯迈医疗设备有限公司 | Image processing method and device for assisting interventional operation |
-
2021
- 2021-05-27 CN CN202110584158.8A patent/CN115409758A/en active Pending
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117274502A (en) * | 2023-11-17 | 2023-12-22 | 北京唯迈医疗设备有限公司 | Image processing method and device for assisting interventional operation |
CN117274502B (en) * | 2023-11-17 | 2024-03-01 | 北京唯迈医疗设备有限公司 | Image processing method and device for assisting interventional operation |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109166124B (en) | Retinal blood vessel morphology quantification method based on connected region | |
US10614573B2 (en) | Method for automatically recognizing liver tumor types in ultrasound images | |
CN109829942B (en) | Automatic quantification method for retinal vessel diameter of fundus image | |
CN110807764A (en) | Lung cancer screening method based on neural network | |
CN111657883B (en) | Coronary artery SYNTAX score automatic calculation method and system based on sequence radiography | |
CN110222759A (en) | A kind of Plaque Vulnerability in Coronary Artery automatic recognition system | |
CN111369528B (en) | Coronary artery angiography image stenosis region marking method based on deep convolutional network | |
WO2021208739A1 (en) | Method and apparatus for evaluating blood vessel in fundus color image, and computer device and medium | |
CN109146854A (en) | A kind of analysis method of Lung neoplasm and pulmonary vascular association relationship | |
CN115969310A (en) | System and method for measuring pathological change characteristics of hypertensive retinopathy | |
CN113240684B (en) | Method for converting mask data extracted to contain coronary artery central line into central line point set | |
CN112233789A (en) | Regional feature fusion type hypertensive retinopathy classification method | |
CN115049807B (en) | Method, device and server for establishing pulmonary vessel model | |
Maaliw et al. | An enhanced segmentation and deep learning architecture for early diabetic retinopathy detection | |
CN111145155B (en) | Meibomian gland identification method | |
CN115409758A (en) | Method for comprehensively and quantitatively evaluating angiostenosis from contrast images | |
CN116524548B (en) | Vascular structure information extraction method, device and storage medium | |
CN113974667B (en) | Automatic positioning device and method for TAVI preoperative key target | |
CN109559317B (en) | Lung nodule segmentation method based on CT image | |
CN116363311A (en) | Coronary Leiden score calculation and risk classification method and system | |
CN115456950B (en) | Automatic identification and positioning method, equipment and storage medium for puncture blood vessel | |
CN116152492A (en) | Medical image segmentation method based on multi-attention fusion | |
US8422754B2 (en) | Medical imaging system | |
US11049241B2 (en) | Qualification method of lung vessel based on lobe | |
Ashame et al. | Abnormality Detection in Eye Fundus Retina |
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 |