CN111798438B - Intravascular ultrasound image intelligent segmentation method and system - Google Patents

Intravascular ultrasound image intelligent segmentation method and system Download PDF

Info

Publication number
CN111798438B
CN111798438B CN202010656056.8A CN202010656056A CN111798438B CN 111798438 B CN111798438 B CN 111798438B CN 202010656056 A CN202010656056 A CN 202010656056A CN 111798438 B CN111798438 B CN 111798438B
Authority
CN
China
Prior art keywords
image
contour
lumen
segmentation
ivus
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.)
Active
Application number
CN202010656056.8A
Other languages
Chinese (zh)
Other versions
CN111798438A (en
Inventor
于波
王钊
曾强
侯静波
贾海波
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Panorama Scientific Co ltd
Original Assignee
Panorama Scientific Co ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Panorama Scientific Co ltd filed Critical Panorama Scientific Co ltd
Priority to CN202010656056.8A priority Critical patent/CN111798438B/en
Publication of CN111798438A publication Critical patent/CN111798438A/en
Application granted granted Critical
Publication of CN111798438B publication Critical patent/CN111798438B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

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
    • 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
    • G06T2207/10101—Optical tomography; Optical coherence tomography [OCT]
    • 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
    • G06T2207/10132—Ultrasound image
    • 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
    • G06T2207/20081—Training; Learning
    • 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)
  • 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)
  • Ultra Sonic Daignosis Equipment (AREA)

Abstract

The invention discloses an intelligent segmentation method and system for intravascular ultrasound images, which comprises the following steps: acquiring an IVUS image and an OCT image in a blood vessel; registering the IVUS image and the OCT image; acquiring an OCT image lumen segmentation result through lumen segmentation; mapping the OCT image lumen segmentation result to the IVUS image to obtain an IVUS image lumen segmentation initial contour, putting the IVUS image lumen segmentation initial contour into a movable contour model, and minimizing an energy function of the movable contour model to obtain an IVUS image lumen segmentation processing contour; and acquiring the regional gradient of the neighborhood in the IVUS image lumen segmentation processing contour, searching a local optimal point, and optimizing the IVUS image lumen segmentation processing contour to obtain an IVUS image lumen segmentation result. The invention utilizes the high-resolution OCT image to generate the initialization contour, the quality of the initialization contour is higher than that of the initialization contour directly generated in the IVUS image, and the final segmentation result is closer to a real boundary.

Description

Intravascular ultrasound image intelligent segmentation method and system
Technical Field
The invention relates to cardiovascular interventional surgery, belongs to the field of medical image segmentation, and particularly relates to an intravascular ultrasound image intelligent segmentation method and system.
Background
In recent years, intravascular ultrasound (IVUS) imaging has become more and more widely used in clinical diagnosis and interventional therapy of vascular diseases, particularly coronary heart disease. The IVUS image can reflect the change of the blood vessel inner cavity, and can also reflect the information of the cross section structure of the blood vessel including plaque, the thickness of the blood vessel wall, the plaque components and the like. Generally, IVUS images contain three layers of annular structures: the inner layer is expressed as a strong echo bright ring and is histologically composed of an inner membrane and an inner elastic membrane; the middle layer is a middle membrane and is expressed as a low echo dark zone; the outer layer is a bright strong echogenic zone and is histologically composed of an outer membrane and an outer membrane elastic membrane. The IVUS image is segmented, mainly to extract the vessel lumen boundary and the tunica media boundary, and the current clinically commonly used means is manual segmentation, which is generally segmented frame by an experienced doctor. Generally, each IVUS image sequence contains thousands of images, the manual segmentation method is time-consuming, labor-intensive, and has poor repeatability, and the segmentation result also depends on the experience and subjective factors of the segmenter to a great extent.
When the digital image processing method is adopted to automatically segment the IVUS image, the problems of high noise, low resolution, image artifacts, calcified shadows and the like of the IVUS image are faced. The current algorithms for IVUS image segmentation are mainly graph search, active contour model, etc. Compared with the traditional conventional image segmentation algorithm, the methods are greatly improved, but the methods are not completely automatic, manual intervention is still needed, and how to solve the problem that the full-automatic IVUS image segmentation becomes a research hotspot in the field.
Patent CN103886599A discloses a blood vessel ROI segmentation method based on intravascular ultrasound images, which segments the lumen region of a blood vessel and the lumen membrane contour initialization contour, and then completes the segmentation of a blood vessel plaque by using the prior information of the lumen region information. However, the method has more empirical parameters for initializing the active contour model, is not accurate enough for the initial segmentation of the vessel lumen region, and has poor robustness of the final segmentation result.
Patent CN107909585A discloses a method for segmenting intima in blood vessels by intravascular ultrasound image, which uses a fusion net deep learning segmentation model to extract intima region in blood vessels. However, the method has a large requirement on data, ten thousand labeled images are needed for training an effective segmentation model, and the segmentation effect of the model trained by using a small amount of data is even inferior to that of the traditional segmentation algorithm.
Disclosure of Invention
The invention aims to solve the technical problems that in the prior art, when an active contour model is adopted for initialization, more experience parameters are utilized, initial segmentation is not accurate enough, so that the robustness of the final segmentation result is poor, or the requirement of training data is large when a fusion Net deep learning segmentation model is used, or manual segmentation consumes time, is large in workload and poor in repeatability, and the segmentation result also depends on the experience and subjective factors of a segmenter.
The invention is realized by the following technical scheme:
an intelligent segmentation method for intravascular ultrasound images comprises the following steps: s1: acquiring intravascular optical coherence tomography and intravascular ultrasound images; the intravascular ultrasound image is a first modality image, namely an IVUS image; the intravascular optical coherence tomography is a second modality image, namely an OCT image; s2: registering the first modality image and the second modality image; s3: obtaining a lumen segmentation result of the second modality image through lumen segmentation; s4: mapping the second modality image lumen segmentation result to the first modality image to obtain a first modality image lumen segmentation initial contour; s5: the first modality image lumen segmentation initial contour is placed into a movable contour model, and an energy function of the movable contour model is minimized to obtain a first modality image lumen segmentation processing contour; s6: acquiring the regional gradient of the neighborhood in the first modal image lumen segmentation processing contour, searching a local optimal point, optimizing the first modal image lumen segmentation processing contour to obtain a first modal image lumen segmentation result, wherein the first modal image lumen segmentation result comprises a first modal image lumen inner membrane contour and an outer membrane contour.
The invention realizes accurate segmentation of the inner cavity and the outer cavity of the IVUS image by analyzing respective imaging characteristics of Optical Coherence Tomography (OCT) and intravascular ultrasound (IVUS). Firstly, an OCT image and an IVUS image are simultaneously acquired in a blood vessel, because the OCT image has high resolution, the lumen outline of the OCT image is extracted through lumen segmentation, then the lumen outline of the OCT image is mapped into a corresponding IVUS image to be used as an initial outline of an active outline model, the energy of the active outline model is minimized to enable the initial outline to be continuously deformed, so that an inner membrane area and an outer membrane area of the IVUS image are extracted, finally, the extracted inner membrane outline and the extracted outer membrane outline are accurately optimized, the regional gradient is calculated, the original point is replaced by a local optimal point, and a final segmentation result is obtained.
When the active contour model is used, the high-resolution OCT image is used for generating the initialized contour, the segmentation precision is higher than that of the initialized contour directly generated in the IVUS image, and the final segmentation result is closer to a real boundary.
The invention relates to a method for detecting an inner cavity and an outer cavity of an IVUS image by initializing a movable contour model according to a corresponding OCT image lumen segmentation result. Compared with the traditional conventional method, the method directly uses the active contour model or the improved algorithm thereof to carry out the full-automatic segmentation of the IVUS image, the method greatly improves the segmentation precision, and greatly simplifies the calculation model on the calculation complexity. Compared with the traditional medical image segmentation algorithm, the method can adaptively segment the inner cavity and the outer cavity of the IVUS image, and the segmentation precision is higher than that of the traditional algorithm.
Further, the active contour model in step S5 is as follows:
Figure BDA0002576796300000021
wherein E isint(i) Denotes the internal energy, Eext(i) Representing the external energy and N representing the total number of points in the initial contour.
Further, the internal energy Eint(i) The active contour model is kept continuous and smooth in the deformation process, and the normalization process is as follows:
Figure BDA0002576796300000031
wherein, ciRepresenting the ith active contour point,
Figure BDA0002576796300000032
representing the average distance between adjacent active contour points, maxd representing the maximum distance between adjacent active contour points, the values being updated in each iteration, α and β representing weight parameters whose value ranges from [0, 1%]。
Further, the external energy Eext(i) Is an external force ensuring the convergence of the active contour model, determines the activityMoving direction of the contour model, which is defined as follows:
Figure BDA0002576796300000033
wherein, I (x)i,yi) Represents a pixel (x)i,yi) Is determined by the gray-scale value of (a),
Figure BDA0002576796300000034
represents a pixel (x)i,yi) Gamma and lambda represent weight parameters, whose value ranges from 0,1]。
Further, in step S3, the lumen is segmented by a deep learning algorithm.
Further, the deep learning algorithm comprises a graph cut method.
Further, in step S1, the intravascular IVUS image and the OCT image are simultaneously acquired by the cardiovascular interventional imaging system.
In another implementation manner of the present invention, an intravascular ultrasound image intelligent segmentation system includes: cardiovascular intervention image module: for acquiring an OCT image and an IVUS image within a vessel; a segmentation processing module: for registering the IVUS image and the OCT image and establishing a mapping model of the IVUS image and the OCT image; acquiring the OCT image lumen segmentation result by a graph cutting method; mapping the OCT image lumen segmentation result to the IVUS image to obtain an IVUS image lumen segmentation initial contour; placing the IVUS image lumen segmentation initial contour into a movable contour model, and minimizing an energy function of the movable contour model to obtain an IVUS image lumen segmentation processing contour; acquiring the regional gradient of the neighborhood in the IVUS image lumen segmentation processing contour, searching a local optimal point, and optimizing the IVUS image lumen segmentation processing contour to obtain an IVUS image lumen segmentation result, wherein the IVUS image lumen segmentation result comprises an IVUS image lumen inner membrane contour and an IVUS image lumen outer membrane contour; an output display module: and the segmentation result is used for displaying the lumen intima contour and the adventitia contour of the IVUS image.
Further, the OCT image is intravascular optical coherence tomography, and the IVUS image is an intravascular ultrasound image.
Compared with the prior art, the invention has the following advantages and beneficial effects:
compared with the traditional medical image segmentation algorithm, the method can adaptively segment the inner cavity and the outer cavity of the IVUS image, and the segmentation precision is higher than that of the traditional algorithm. The method does not need a large amount of data for training and manual operation, segments the OCT image synchronously acquired with the IVUS image by processing, maps the segmented OCT image into the IVUS image, and generates the initialized contour by using the high-resolution OCT image when using the active contour model, so that the quality of the initialized contour is higher than that of the initialized contour directly generated in the IVUS image, and the final segmentation result is closer to a real boundary. The invention has smaller calculation data amount and higher segmentation precision, and promotes the development of the blood vessel ultrasonic image segmentation technology.
Drawings
The accompanying drawings, which are included to provide a further understanding of the embodiments of the invention and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the invention and together with the description serve to explain the principles of the invention. In the drawings:
FIG. 1 is a schematic view of the process of the present invention;
FIG. 2 is a schematic diagram of a method for extracting image contours by Graph Cuts;
FIG. 3 is a diagram of an implementation of the active contour model to extract the inner and outer membranes of an IVUS image;
FIG. 4 is a schematic view of an IVUS image;
fig. 5 is a schematic diagram of an OCT image.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail below with reference to examples and accompanying drawings, and the exemplary embodiments and descriptions thereof are only used for explaining the present invention and are not meant to limit the present invention.
Example 1
This embodiment 1 is an intelligent segmentation method for intravascular ultrasound images, and proposes a method for detecting an inner lumen and an outer lumen of an IVUS image by initializing an active contour model using a segmentation result of a lumen corresponding to an OCT image in order to realize high-precision full-automatic segmentation of the inner lumen and the outer lumen of the IVUS image. Compared with the conventional method, the method directly uses the active contour model or the improved algorithm thereof to perform the full-automatic segmentation of the IVUS image, the method of the embodiment 1 has the advantages that the segmentation precision is greatly improved, and the calculation model is greatly simplified in the calculation complexity.
The specific steps of this example 1 are:
1. acquiring an IVUS image and an OCT image in a blood vessel; as shown in fig. 4 and 5;
2. registering the IVUS image and the OCT image and establishing a mapping model of the IVUS image and the OCT image;
3. performing lumen segmentation through a deep learning algorithm, wherein the preferred deep learning algorithm is a graph segmentation method, and acquiring the OCT image lumen segmentation result;
4. mapping the OCT image lumen segmentation result to the IVUS image to obtain an IVUS image lumen segmentation initial contour;
5. placing the IVUS image lumen segmentation initial contour into a movable contour model, and minimizing an energy function of the movable contour model to obtain an IVUS image lumen segmentation processing contour;
6. obtaining the regional gradient of the neighborhood in the IVUS image lumen segmentation processing contour, searching a local optimal point, optimizing the IVUS image lumen segmentation processing contour to obtain an IVUS image lumen segmentation result, wherein the IVUS image lumen segmentation result comprises an IVUS image lumen inner membrane contour and an outer membrane contour.
When calculating the lumen segmentation of the IVUS image, mapping the segmentation result to the corresponding IVUS image; establishing an IVUS and corresponding OCT image lumen segmentation result mapping model, firstly calculating a high-resolution OCT image lumen segmentation result, then mapping the OCT image lumen segmentation result into an IVUS image to serve as an initial contour of IVUS image segmentation, sending the initial contour into a movable contour model, and calculating an inner cavity segmentation result and an outer cavity segmentation result of the IVUS image by using an energy function of the minimized movable contour model; and calculating the regional gradient of the calculated inner cavity and outer cavity segmentation results of the IVUS image in the neighborhood, searching a local optimal point and further improving the segmentation precision.
The technical solution of this embodiment 1 will be clearly and completely described below with reference to the accompanying drawings.
As shown in fig. 1, in a cardiovascular interventional procedure, intravascular images are simultaneously acquired by an intravascular imaging system such as Optical Coherence Tomography (OCT) and intravascular ultrasound (IVUS) acquisition. And registering the OCT images and the IVUS images correspondingly to obtain the registered OCT images and IVUS images, segmenting the OCT images with higher resolution ratio through a lumen to obtain a lumen contour, transmitting the lumen contour into an IVUS image active contour model as an initial contour, and finally deforming the initial contour by minimizing an energy function of the active contour model to extract the contours of the inner membrane and the outer membrane of the IVUS image.
As shown in fig. 2, the principle of the Graph cut method (Graph Cuts) to solve the OCT image lumen profile is shown. Graph cut methods associate an image segmentation problem with the min cut problem of the graph. Firstly, an undirected graph G is constructed, wherein the undirected graph G is equal to < V, E > represents an image to be segmented, and V and E respectively represent a set of vertexes and edges. The Graph Cuts Graph is provided with two vertexes and edges, wherein the first common vertex corresponds to each pixel in the image, and the connection of every two adjacent vertexes is an edge and is recorded as n-links; besides the image pixels, there are two other terminal vertices, denoted as s (source) and t (sink), and each common vertex and the two terminal vertices are connected to form a second edge, denoted as t-links. Each edge in the graph has a non-negative weight w, which can be understood as cost. A cut is a subset C of the edge set E in the graph, and the cost of the cut is the sum of the weights of all the edges of the edge subset C. And if the sum of all the weights of the edges of one cut is minimum, the cut is called the minimum cut, namely the final segmentation result. The minimum cut is calculated as shown in the following equation:
E(A)＝λ·R(A)+B(A)
wherein e (a) represents a cut value, λ represents a balance factor, r (a) represents a prior penalty term, and b (a) represents a region similarity penalty term.
As shown in FIG. 3, a method and steps for extracting the intima and adventitia of an IVUS image using an active contour model are shown. And mapping the OCT image segmentation result into a corresponding IVUS image, and transmitting the result into the active contour model as an initial contour. Setting an energy function of the active contour model under the interaction of internal energy and external energy, wherein the contour is continuously deformed by minimizing the energy function, and the active contour model is stopped at a target contour when the final energy reaches a global minimum value, wherein the energy function of the active contour model is shown in the following formula;
Figure BDA0002576796300000061
wherein Eint(i) Denotes the internal energy, Eext(i) Representing the external energy and N representing the total number of points in the initial contour. Internal energy Eint(i) The active contour is kept continuous and smooth in the deformation process, and the normalization process is as follows:
Figure BDA0002576796300000062
wherein c isiRepresenting the ith active contour point and,
Figure BDA0002576796300000063
and maxd respectively represents the average distance and the maximum distance between adjacent active contour points, the values are updated in each iteration, alpha and beta represent weight parameters, and the value range is [0,1 ]]. External energy Eext(i) Is an external force for ensuring the convergence of the movable contour, determines the moving direction of the movable contour, and is defined as formula (4).
Figure BDA0002576796300000064
Wherein I (x)i,yi) And
Figure BDA0002576796300000065
respectively represent pixels (x)i,yi) Gamma and lambda represent weight parameters, with a value range of [0, 1%]。
And finally, calculating the regional gradient value of all the points of the obtained segmentation result in the neighborhood range, and taking the locally optimal point to obtain the optimized inner cavity and outer cavity segmentation results.
Example 2
This embodiment 2 is an intravascular ultrasound image intelligent segmentation system, which includes a cardiovascular interventional image module, a segmentation processing contour, and an output display module.
The cardiovascular intervention image module is used for acquiring an OCT image and an IVUS image in a blood vessel.
The segmentation processing module is used for registering the IVUS image and the OCT image and establishing a mapping model of the IVUS image and the OCT image; acquiring an OCT image lumen segmentation result by a graph cutting method; mapping the OCT image lumen segmentation result to the IVUS image to obtain an IVUS image lumen segmentation initial contour; placing the IVUS image lumen segmentation initial contour into a movable contour model, and minimizing an energy function of the movable contour model to obtain an IVUS image lumen segmentation processing contour; obtaining the regional gradient of the neighborhood in the IVUS image lumen segmentation processing contour, searching a local optimal point, optimizing the IVUS image lumen segmentation processing contour, and obtaining an IVUS image lumen segmentation result, wherein the IVUS image lumen segmentation result comprises an IVUS image lumen intima contour and an IVUS image lumen adventitia contour.
The output display module is used for displaying the segmentation result of the lumen intima contour and the adventitia contour of the IVUS image.
In this embodiment 2, the OCT image is an intravascular optical coherence tomography image, and the IVUS image is an intravascular ultrasound image.
The above-mentioned embodiments are intended to illustrate the objects, technical solutions and advantages of the present invention in further detail, and it should be understood that the above-mentioned embodiments are merely exemplary embodiments of the present invention, and are not intended to limit the scope of the present invention, and any modifications, equivalent substitutions, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.

Claims (8)

1. An intelligent segmentation method for intravascular ultrasound images is characterized by comprising the following steps:
s1: acquiring intravascular optical coherence tomography and intravascular ultrasound images; the intravascular ultrasound image is a first modality image, namely an IVUS image; the intravascular optical coherence tomography is a second modality image, namely an OCT image;
s2: registering the first modality image and the second modality image;
s3: obtaining a lumen segmentation result of the second modality image through lumen segmentation;
s4: establishing a mapping model of the first modality image and the second modality image, and mapping the lumen segmentation result of the second modality image to the first modality image to obtain a lumen segmentation initial contour of the first modality image;
s5: the first modality image lumen segmentation initial contour is placed into a movable contour model, and an energy function of the movable contour model is minimized to obtain a first modality image lumen segmentation processing contour;
s6: acquiring the regional gradient of the neighborhood in the first modal image lumen segmentation processing contour, searching a local optimal point, optimizing the first modal image lumen segmentation processing contour to obtain a first modal image lumen segmentation result, wherein the first modal image lumen segmentation result comprises a first modal image lumen inner membrane contour and an outer membrane contour.
2. The intelligent segmentation method for ultrasound image in blood vessel as claimed in claim 1, wherein the active contour model in step S5 is as follows:
Figure FDA0003656453150000011
wherein, Eint(i) Denotes the internal energy, Eext(i) Indicating external energyQuantity, N, represents the total number of points in the initial contour.
3. The intelligent segmentation method for intravascular ultrasound images according to claim 2, wherein the internal energy E isint(i) The active contour model is kept continuous and smooth in the deformation process, and the normalization process is as follows:
Figure FDA0003656453150000012
wherein, ciRepresenting the ith active contour point,
Figure FDA0003656453150000013
represents the average distance between adjacent active contour points, maxd represents the maximum distance between adjacent active contour points, the value will be updated in each iteration, α and β represent weight parameters, the value range is [0,1 ]]。
4. The intelligent segmentation method for intravascular ultrasound images according to claim 2, wherein the external energy E isext(i) The external force for ensuring the convergence of the active contour model determines the moving direction of the active contour model, which is defined as follows:
Figure FDA0003656453150000014
wherein, I (x)i,yi) Represents a pixel (x)i,yi) Is determined by the gray-scale value of (a),
Figure FDA0003656453150000015
represents a pixel (x)i,yi) Gamma and lambda represent weight parameters, whose value ranges from 0,1]。
5. The intelligent segmentation method for ultrasound images in blood vessels according to claim 1, wherein in step S3, the lumen segmentation is performed by a depth learning algorithm.
6. The intelligent segmentation method for ultrasound images in blood vessels according to claim 5, wherein the deep learning algorithm comprises graph cut method.
7. The intelligent segmentation method for ultrasound image in blood vessel as claimed in claim 1, wherein in step S1, IVUS image and OCT image in blood vessel are obtained simultaneously by cardiovascular interventional imaging system.
8. An intravascular ultrasound image intelligent segmentation system, comprising:
cardiovascular intervention image module: for acquiring an OCT image and an IVUS image within a vessel;
a segmentation processing module: for registering the IVUS image and the OCT image and establishing a mapping model of the IVUS image and the OCT image; acquiring the OCT image lumen segmentation result by a graph cutting method; mapping the OCT image lumen segmentation result to the IVUS image to obtain an IVUS image lumen segmentation initial contour; placing the IVUS image lumen segmentation initial contour into a movable contour model, and minimizing an energy function of the movable contour model to obtain an IVUS image lumen segmentation processing contour; acquiring the regional gradient of the neighborhood in the IVUS image lumen segmentation processing contour, searching a local optimal point, and optimizing the IVUS image lumen segmentation processing contour to obtain an IVUS image lumen segmentation result, wherein the IVUS image lumen segmentation result comprises an IVUS image lumen inner membrane contour and an IVUS image lumen outer membrane contour;
an output display module: and the segmentation result is used for displaying the lumen intima contour and the adventitia contour of the IVUS image.
CN202010656056.8A 2020-07-09 2020-07-09 Intravascular ultrasound image intelligent segmentation method and system Active CN111798438B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010656056.8A CN111798438B (en) 2020-07-09 2020-07-09 Intravascular ultrasound image intelligent segmentation method and system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010656056.8A CN111798438B (en) 2020-07-09 2020-07-09 Intravascular ultrasound image intelligent segmentation method and system

Publications (2)

Publication Number Publication Date
CN111798438A CN111798438A (en) 2020-10-20
CN111798438B true CN111798438B (en) 2022-07-01

Family

ID=72811397

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010656056.8A Active CN111798438B (en) 2020-07-09 2020-07-09 Intravascular ultrasound image intelligent segmentation method and system

Country Status (1)

Country Link
CN (1) CN111798438B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113643302B (en) * 2021-07-21 2023-10-27 上海交通大学 Active contour model-based unsupervised medical image segmentation method and system

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101964118A (en) * 2010-09-30 2011-02-02 华北电力大学（保定） Three-dimensional segmentation method for intravascular ultrasound image sequence
CN104376549A (en) * 2014-11-20 2015-02-25 华北电力大学（保定） Intravascular ultrasound image and intravascular-OCT image fusing method
CN111291736A (en) * 2020-05-07 2020-06-16 南京景三医疗科技有限公司 Image correction method and device and medical equipment

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE102005037427A1 (en) * 2005-08-08 2007-02-15 Siemens Ag Method for recording and evaluating vascular examination data

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101964118A (en) * 2010-09-30 2011-02-02 华北电力大学（保定） Three-dimensional segmentation method for intravascular ultrasound image sequence
CN104376549A (en) * 2014-11-20 2015-02-25 华北电力大学（保定） Intravascular ultrasound image and intravascular-OCT image fusing method
CN111291736A (en) * 2020-05-07 2020-06-16 南京景三医疗科技有限公司 Image correction method and device and medical equipment

Also Published As

Publication number Publication date
CN111798438A (en) 2020-10-20

Similar Documents

Publication Publication Date Title
US12079921B2 (en) System and method for image-based object modeling using multiple image acquisitions or reconstructions
Leung et al. Automated border detection in three-dimensional echocardiography: principles and promises
US7925064B2 (en) Automatic multi-dimensional intravascular ultrasound image segmentation method
US9514530B2 (en) Systems and methods for image-based object modeling using multiple image acquisitions or reconstructions
CN101964118B (en) Three-dimensional segmentation method for intravascular ultrasound image sequence
CN110047086B (en) Automatic carotid intimal thickness measuring method and system
CN109003280B (en) Method for segmenting intima in blood vessel by two-channel intravascular ultrasonic image
CN111798438B (en) Intravascular ultrasound image intelligent segmentation method and system
CN113689441B (en) Left ventricle ultrasonic dynamic segmentation method based on DeepLabV network
CN109919098B (en) Target object identification method and device
CN113470060B (en) Coronary artery multi-angle curved surface reconstruction visualization method based on CT image
CN113057677B (en) Heart image modeling method, system and equipment for fusing ultrasonic image and CT image
CN112070778A (en) Multi-parameter extraction method based on intravascular OCT and ultrasound image fusion
CN116994113A (en) Automatic segmentation of liver and tumor in CT image based on residual UNet and efficient multi-scale attention method
CN114565623B (en) Lung vessel segmentation method, device, storage medium and electronic equipment
CN116363332A (en) Prostate ultrasonic correction method based on medical image segmentation and feature matching
CN115564817A (en) Three-dimensional blood vessel image determination method, device, equipment and storage medium
CN113592802B (en) Mitral valve annular displacement automatic detection system based on ultrasonic image
CN112669256B (en) Medical image segmentation and display method based on transfer learning
LU505348B1 (en) A deep learning-based method for CT pulmonary angiography cardiac chamber segmentation and quantification
CN118298004B (en) Heart function assessment method and system based on three-dimensional echocardiography
CN112669439B (en) Method for establishing intracranial angiography enhanced three-dimensional model based on transfer learning
CN111768391B (en) Full-automatic heart function analysis method, device, computer equipment and storage medium based on CT image
Bharali et al. Cardiac motion estimation from echocardiographic image sequence using unsupervised active contour tracker
Almeida et al. Semi-automatic left-atrial segmentation from volumetric ultrasound using B-spline explicit active surfaces

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
GR01 Patent grant
GR01 Patent grant