CN111539917A - Blood vessel segmentation method, system, terminal and storage medium based on coarse and fine granularity fusion - Google Patents
Blood vessel segmentation method, system, terminal and storage medium based on coarse and fine granularity fusion Download PDFInfo
- Publication number
- CN111539917A CN111539917A CN202010272913.4A CN202010272913A CN111539917A CN 111539917 A CN111539917 A CN 111539917A CN 202010272913 A CN202010272913 A CN 202010272913A CN 111539917 A CN111539917 A CN 111539917A
- Authority
- CN
- China
- Prior art keywords
- blood vessel
- region
- vessel
- segmentation result
- fine
- 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.)
- Granted
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
- G06T7/11—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
- G06T7/187—Segmentation; Edge detection involving region growing; involving region merging; involving connected component labelling
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20212—Image combination
- G06T2207/20221—Image fusion; Image merging
-
- 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
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
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 application provides a blood vessel segmentation method, a system, a terminal and a storage medium based on coarse and fine granularity fusion, wherein the method comprises the following steps: acquiring a medical image, and performing blood vessel thickness and granularity segmentation on the medical image to obtain a rough blood vessel segmentation result and a fine blood vessel segmentation result; determining a potential blood vessel fracture area according to the rough blood vessel segmentation result and the fine blood vessel segmentation result; determining a non-vascular rupture region in the potential vascular rupture region and removing the non-vascular rupture region from the potential vascular rupture region to obtain a preserved vascular rupture region; fusing the reserved blood vessel fracture area with the fine blood vessel segmentation result to repair the blood vessel fracture area in the fine blood vessel segmentation result to obtain a final blood vessel segmentation result; the method and the device fuse the coarse and fine granularity segmentation results, repair the blood vessel fracture area in fine blood vessel segmentation by using the coarse blood vessel segmentation result, avoid introducing redundant noise in the coarse blood vessel segmentation result, improve the blood vessel segmentation effect and ensure the blood vessel connectivity.
Description
Technical Field
The present application relates to the field of medical imaging and computer-aided technologies, and in particular, to a method, a system, a terminal, and a storage medium for vessel segmentation based on coarse and fine particle size fusion.
Background
Stroke, also known as cerebrovascular accident, is a disorder of cerebral blood circulation, and its fatality rate is second to heart diseases and higher than cancer. At present, the incidence and mortality of stroke in China are still in an ascending trend, and the mortality of old people after suffering from stroke is doubled compared with that of young people. Cerebral apoplexy is divided into ischemic stroke and hemorrhagic stroke, wherein the ischemic stroke accounts for about 80 percent. One of the main causes of ischemic stroke is the stenosis of the lumen and the hemodynamic changes caused by atherosclerotic vulnerable plaque of the head and neck arteries, and the carotid artery is the main blood supply artery of cerebral circulation as the upstream blood vessel of the intracranial artery. Therefore, the head and neck blood vessel segmentation performed by the head and neck artery angiography image (CTA) is of great significance for finding risks in time.
In modern medical image analysis, the separation of blood vessels from the original medical image is an important basis for various pathological analyses. The vessel segmentation based on the deep learning model is the current mainstream method, and compared with the traditional vessel segmentation method, the method has the advantages of good effect and robust performance. However, the existing blood vessel segmentation methods generally have the condition of blood vessel segmentation fracture, and the blood vessel segmentation fracture seriously influences the post-processing effect and pathological analysis.
Therefore, there is a need for a vessel segmentation method, system, terminal and storage medium based on coarse-fine particle fusion, which can effectively reduce vessel segmentation fracture and improve vessel segmentation effect.
Disclosure of Invention
Aiming at the defects of the prior art, the application provides a blood vessel segmentation method, a system, a terminal and a storage medium based on coarse-fine granularity fusion, and solves the problems that the blood vessel segmentation fracture exists in the blood vessel segmentation method in the prior art, and the post-processing effect and pathological analysis are seriously influenced.
In order to solve the above technical problem, in a first aspect, the present application provides a method for segmenting a blood vessel based on coarse-fine particle size fusion, including:
acquiring a medical image, and performing blood vessel thickness and granularity segmentation on the medical image to obtain a rough blood vessel segmentation result and a fine blood vessel segmentation result;
determining a potential blood vessel fracture area according to the rough blood vessel segmentation result and the fine blood vessel segmentation result;
determining a non-vascular rupture region in the potential vascular rupture region and removing the non-vascular rupture region from the potential vascular rupture region to obtain a preserved vascular rupture region;
and fusing the reserved blood vessel fracture area and the fine blood vessel segmentation result to repair the blood vessel fracture area in the fine blood vessel segmentation result to obtain a final blood vessel segmentation result.
Optionally, the acquiring a medical image map and performing blood vessel coarse-fine granularity segmentation on the medical image map to obtain a coarse blood vessel segmentation result and a fine blood vessel segmentation result includes:
acquiring a medical image map;
and performing vessel segmentation on the same medical image by training two deep learning models with different segmentation accuracies or adopting different segmentation thresholds, and obtaining a rough vessel segmentation result and a fine vessel segmentation result of the medical image.
Optionally, the determining a potential vessel fracture region according to the coarse vessel segmentation result and the fine vessel segmentation result includes:
making a difference between the rough blood vessel segmentation result and the fine blood vessel segmentation result, and setting voxels with difference values smaller than zero as zero to obtain redundant blood vessel parts in the rough blood vessel segmentation result;
and analyzing the connected domain of the redundant blood vessel part in the rough blood vessel segmentation result, and dividing mutually connected voxel points into the same region to form a plurality of potential blood vessel fracture regions.
Optionally, the determining a non-vascular fracture region in the potential vascular fracture region and removing the non-vascular fracture region from the potential vascular fracture region to obtain a preserved vascular fracture region includes:
determining non-vascular rupture regions in the potential vascular rupture regions;
removing the non-vascular rupture region from the potential vascular rupture region to yield a preserved vascular rupture region;
wherein the non-vascular rupture region comprises: small area vascular noise, vessel wall area, terminal unwanted branch vessel area, and stuck vessel branches.
Optionally, the method for determining the small region vascular noise includes:
acquiring the number of voxels of each region of the potential vessel fracture region;
and judging whether the number of voxels in the region is less than a preset number, if so, judging that the region is small-region vascular noise.
Optionally, the method for determining the vascular wall region includes:
acquiring the total number of voxel points of each region of the potential vessel fracture region;
judging whether each individual voxel point in each region of the blood vessel fracture region has adjacent blood vessel voxels in the fine blood vessel segmentation result, if so, acquiring the number of voxel points of the adjacent blood vessel voxels in the region;
calculating the ratio of the number of voxel points of the voxels of the adjacent vessels in the region to the total number of the voxel points in the region;
and judging whether the voxel point ratio of the region exceeds a preset threshold value, if so, judging that the region is a vascular wall region.
Optionally, the method for determining the terminal redundant branch vessel region includes:
obtaining voxel points of adjacent vessel voxels in each region of the vessel fracture region;
and judging whether the voxel points of the regions only belong to the same connected region, if so, judging that the connected region is the only adjacent surface of the terminal blood vessel branch and the fine blood vessel, and judging that the region is the terminal redundant branch blood vessel region.
Optionally, the method for determining an adhered blood vessel branch includes:
obtaining voxel points of adjacent vessel voxels in each region of the vessel fracture region;
obtaining a vessel voxel point adjacent to a voxel point in a vessel fracture region in a fine vessel segmentation result;
and judging whether the voxel points in the fine blood vessel segmentation result only belong to the same communicated region, if so, judging that the blood vessel branch is connected with the mutually communicated regions in the fine blood vessel segmentation, namely two ends of a U-shaped blood vessel or a plurality of adjacent blood vessels, and judging that the regions are adhered blood vessel branches.
Optionally, the fusing the remaining blood vessel fracture region with the refined blood vessel segmentation result to repair the blood vessel fracture region in the refined blood vessel segmentation result, so as to obtain a final blood vessel segmentation result, including:
and merging the fine blood vessel segmentation result with the reserved blood vessel fracture area to obtain a final blood vessel segmentation result.
In a second aspect, the present application further provides a vessel segmentation system based on coarse-fine particle fusion, including:
the segmentation unit is configured to acquire a medical image map and perform blood vessel thickness-granularity segmentation on the medical image map to obtain a rough blood vessel segmentation result and a fine blood vessel segmentation result;
a determining unit configured to determine a potential vessel rupture region according to the coarse vessel segmentation result and the fine vessel segmentation result;
a removal unit configured to determine a non-vascular rupture region of the potential vascular rupture regions and remove the non-vascular rupture region from the potential vascular rupture region to obtain a preserved vascular rupture region;
and the fusion unit is configured to fuse the reserved blood vessel fracture region and the fine blood vessel segmentation result to repair the blood vessel fracture region in the fine blood vessel segmentation result to obtain a final blood vessel segmentation result.
Optionally, the obtaining unit is specifically configured to:
acquiring a medical image map;
and performing vessel segmentation on the same medical image by training two deep learning models with different segmentation accuracies or adopting different segmentation thresholds, and obtaining a rough vessel segmentation result and a fine vessel segmentation result of the medical image.
Optionally, the determining unit is specifically configured to:
making a difference between the rough blood vessel segmentation result and the fine blood vessel segmentation result, and setting voxels with difference values smaller than zero as zero to obtain redundant blood vessel parts in the rough blood vessel segmentation result;
and analyzing the connected domain of the redundant blood vessel part in the rough blood vessel segmentation result, and dividing mutually connected voxel points into the same region to form a plurality of potential blood vessel fracture regions.
Optionally, the removing unit is specifically configured to:
determining non-vascular rupture regions in the potential vascular rupture regions;
removing the non-vascular rupture region from the potential vascular rupture region to yield a preserved vascular rupture region;
wherein the non-vascular rupture region comprises: small area vascular noise, vessel wall area, terminal unwanted branch vessel area, and stuck vessel branches.
Optionally, the fusion unit is specifically configured to:
and merging the fine blood vessel segmentation result with the reserved blood vessel fracture area to obtain a final blood vessel segmentation result.
In a third aspect, the present application provides a terminal, comprising:
a processor, a memory, wherein,
the memory is used for storing a computer program which,
the processor is used for calling and running the computer program from the memory so as to make the terminal execute the method of the terminal.
In a fourth aspect, the present application provides a computer storage medium having instructions stored thereon, which when executed on a computer, cause the computer to perform the method of the above aspects.
Compared with the prior art, the method has the following beneficial effects:
the method and the device have the advantages that the coarse and fine granularity segmentation results are fused, the blood vessel fracture area in fine blood vessel segmentation is repaired by using the coarse blood vessel segmentation results, meanwhile, the introduction of redundant noise in the coarse blood vessel segmentation results is avoided, the blood vessel segmentation effect is improved, and the connectivity of blood vessels is ensured.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below, it is obvious that the drawings in the following description are only embodiments of the present application, and for those skilled in the art, other drawings can be obtained according to the provided drawings without creative efforts.
Fig. 1 is a flowchart of a vessel segmentation method based on coarse-fine particle fusion according to an embodiment of the present application;
FIG. 2 is a flowchart of a vessel thickness granularity segmentation provided in an embodiment of the present application;
FIG. 3 is a flow chart for determining a potential vessel rupture region provided by an embodiment of the present application;
FIG. 4 is a flow chart for removing a non-vascular fracture region from a potential vascular fracture region as provided by an embodiment of the present application;
FIG. 5 is a flow chart of the fusion of the preserved vessel fracture region and the fine vessel segmentation result provided in the embodiment of the present application;
fig. 6 is a schematic structural diagram of a vessel segmentation system based on coarse-fine particle fusion according to another embodiment of the present application;
fig. 7 is a schematic structural diagram of a terminal system according to an embodiment of the present application.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are some embodiments of the present application, but not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
Referring to fig. 1, fig. 1 is a flowchart of a vessel segmentation method based on coarse-fine particle fusion according to an embodiment of the present application, where the method 100 includes:
s101: acquiring a medical image, and performing blood vessel thickness and granularity segmentation on the medical image to obtain a rough blood vessel segmentation result and a fine blood vessel segmentation result;
s102: determining a potential blood vessel fracture area according to the rough blood vessel segmentation result and the fine blood vessel segmentation result;
s103: determining a non-vascular rupture region in the potential vascular rupture region and removing the non-vascular rupture region from the potential vascular rupture region to obtain a preserved vascular rupture region;
s104: and fusing the reserved blood vessel fracture area and the fine blood vessel segmentation result to repair the blood vessel fracture area in the fine blood vessel segmentation result to obtain a final blood vessel segmentation result.
In addition, the conventional vessel segmentation method based on a single deep learning model or a segmentation threshold is difficult to balance the vessel segmentation accuracy and the vessel connectivity, and the segmentation result is often not ideal. The result of the fine vessel segmentation is generally good in segmentation accuracy, but a vessel is likely to be broken. The vessel connectivity of the rough vessel segmentation result is better, but more additional noises such as terminal vessel branches and adhesion vessel branches are included. Therefore, the method and the device fuse the coarse and fine granularity segmentation results, repair the blood vessel fracture area in fine blood vessel segmentation by using the coarse blood vessel segmentation result, avoid introducing redundant noise in the coarse blood vessel segmentation result, improve the blood vessel segmentation effect and ensure the connectivity of the blood vessel.
Based on the foregoing embodiment, as an optional embodiment, the S101 acquires a medical image map, and performs vessel thickness granularity segmentation on the medical image map to obtain a coarse vessel segmentation result and a fine vessel segmentation result, including:
acquiring a medical image map;
and performing vessel segmentation on the same medical image by training two deep learning models with different segmentation accuracies or adopting different segmentation thresholds, and obtaining a rough vessel segmentation result and a fine vessel segmentation result of the medical image.
Specifically, as shown in fig. 2, fig. 2 is a flowchart of blood vessel thickness granularity segmentation provided in the embodiment of the present application. After the medical image is obtained, the same medical image can be subjected to vessel segmentation by training two deep learning models with different segmentation accuracies, namely a high-accuracy segmentation model and a low-accuracy segmentation model, so that a fine vessel segmentation result and a rough vessel segmentation result are obtained respectively. After the medical image is obtained, the same medical image can be subjected to vessel segmentation by setting different segmentation threshold values and respectively adopting a high segmentation threshold value and a low segmentation threshold value to respectively obtain a fine vessel segmentation result and a rough vessel segmentation result.
Based on the foregoing embodiment, as an optional embodiment, the S102 determining a potential blood vessel fracture region according to the coarse blood vessel segmentation result and the fine blood vessel segmentation result includes:
making a difference between the rough blood vessel segmentation result and the fine blood vessel segmentation result, and setting voxels with difference values smaller than zero as zero to obtain redundant blood vessel parts in the rough blood vessel segmentation result;
and analyzing the connected domain of the redundant blood vessel part in the rough blood vessel segmentation result, and dividing mutually connected voxel points into the same region to form a plurality of potential blood vessel fracture regions.
Specifically, as shown in fig. 3, fig. 3 is a flowchart for determining a potential blood vessel fracture region according to an embodiment of the present application. According to the method, the blood vessel regions which only appear in the rough blood vessel segmentation result but not in the fine blood vessel segmentation result are regarded as potential blood vessel fracture regions, redundant blood vessel parts in the rough blood vessel segmentation result can be obtained by making a difference between the rough blood vessel segmentation result and the fine blood vessel segmentation result, the redundant blood vessel parts in the rough blood vessel segmentation result are subjected to connected domain analysis, mutually connected voxel points are divided into the same region, and a plurality of potential blood vessel fracture regions can be formed.
Based on the foregoing embodiment, as an alternative embodiment, the S103 determines a non-vascular fracture region in the potential vascular fracture region, and removes the non-vascular fracture region from the potential vascular fracture region to obtain a preserved vascular fracture region, including:
determining non-vascular rupture regions in the potential vascular rupture regions;
removing the non-vascular rupture region from the potential vascular rupture region to yield a preserved vascular rupture region;
wherein the non-vascular rupture region comprises: small area vascular noise, vessel wall area, terminal unwanted branch vessel area, and stuck vessel branches.
Specifically, as shown in fig. 4, fig. 4 is a flow chart for removing a non-vascular rupture region from a potential vascular rupture region according to an embodiment of the present application. Since the non-vascular rupture region includes: after the small-area blood vessel noise, the blood vessel wall area, the tail-end redundant branch blood vessel area, the adhered blood vessel branches and the like are judged firstly, the small-area blood vessel noise, the blood vessel wall area, the tail-end redundant branch blood vessel area and the adhered blood vessel branches are removed from the potential blood vessel fracture area respectively, and then the reserved blood vessel fracture area can be obtained. The method for determining the small-area blood vessel noise, the blood vessel wall area, the terminal redundant branch blood vessel area and the adhered blood vessel branch is as follows:
and (3) judging the blood vessel noise of the small region: because the potential blood vessel fracture area is obtained by directly performing a difference method, more fragmentary noises, namely small-area blood vessel noises, can be obtained. Acquiring the number of voxels of each region of the potential blood vessel fracture region by using all the obtained potential blood vessel fracture regions; and judging whether the number of voxels in the region is less than a preset number, if so, judging that the region is small-region vascular noise.
Blood vessel wall area determination: the vessels in the coarse vessel segmentation result are usually thicker than those in the fine vessel segmentation result, so that there are more empty vessel regions, i.e. vessel wall regions, after the difference. Acquiring the total number of voxel points of each region of the potential blood vessel fracture regions by using all the obtained potential blood vessel fracture regions; judging whether each individual voxel point in each region of the blood vessel fracture region has adjacent blood vessel voxels in the fine blood vessel segmentation result, if so, acquiring the number of voxel points of the adjacent blood vessel voxels in the region; calculating the ratio of the number of voxel points of the voxels of the adjacent vessels in the region to the total number of the voxel points in the region; and judging whether the voxel point ratio of the region exceeds a preset threshold value, if so, judging that the region is a vascular wall region.
Judging the region of the terminal redundant branch blood vessel: compared with the fine vessel segmentation result, the vessel in the coarse vessel segmentation result usually introduces many redundant elongated branches at the vessel end, that is, the terminal redundant branched vessel. The method comprises the steps of obtaining voxel points of adjacent vessel voxels in each region of a vessel fracture region by using all obtained potential vessel fracture regions; and judging whether the voxel points of the regions only belong to the same connected region, if so, judging that the connected region is the only adjacent surface of the terminal blood vessel branch and the fine blood vessel, and judging that the region is the terminal redundant branch blood vessel region.
Judging the blood vessel branches with adhesion: there is also a special class of vessel branches in the coarse vessel segmentation result, which just connect the two ends of the "U" shaped vessel or several adjacent vessels in the fine vessel segmentation result. Such a special structure allows it to be located just in multiple neighbours to the delicate vessels, i.e. adherent vessel branches. The method comprises the steps of obtaining voxel points of adjacent vessel voxels in each region of a vessel fracture region by using all obtained potential vessel fracture regions; obtaining a vessel voxel point adjacent to a voxel point in a vessel fracture region in a fine vessel segmentation result; and judging whether the voxel points in the fine blood vessel segmentation result only belong to the same communicated region, if so, judging that the blood vessel branch is connected with the mutually communicated regions in the fine blood vessel segmentation, namely two ends of a U-shaped blood vessel or a plurality of adjacent blood vessels, and judging that the regions are adhered blood vessel branches.
Based on the foregoing embodiment, as an alternative embodiment, the S104 fuses the preserved blood vessel fracture region with the fine blood vessel segmentation result to repair the blood vessel fracture region in the fine blood vessel segmentation result, so as to obtain a final blood vessel segmentation result, including:
and merging the fine blood vessel segmentation result with the reserved blood vessel fracture area to obtain a final blood vessel segmentation result.
Specifically, as shown in fig. 5, fig. 5 is a flowchart for fusing the preserved blood vessel fracture region and the fine blood vessel segmentation result provided in the embodiment of the present application. And merging the fine blood vessel segmentation result with all the reserved blood vessel fracture areas, namely directly fusing the reserved blood vessel fracture areas and the fine blood vessel segmentation result to repair the blood vessel fracture areas in the fine blood vessel segmentation, so as to obtain a final blood vessel segmentation result and ensure the connectivity of the final blood vessel segmentation result.
Referring to fig. 6, fig. 6 is a schematic structural diagram of a vessel segmentation system based on coarse-fine particle size fusion according to an embodiment of the present application, the system 600 includes:
a segmentation unit 601 configured to obtain a medical image map, and perform coarse-fine granularity segmentation on the medical image map to obtain a coarse blood vessel segmentation result and a fine blood vessel segmentation result;
a determining unit 602 configured to determine a potential vessel rupture region according to the coarse vessel segmentation result and the fine vessel segmentation result;
a removal unit 603 configured to determine a non-vascular rupture region of the potential vascular rupture regions and remove the non-vascular rupture region from the potential vascular rupture region to obtain a preserved vascular rupture region;
a fusion unit 604 configured to fuse the remaining vessel rupture region with the fine vessel segmentation result to repair the vessel rupture region in the fine vessel segmentation result, resulting in a final vessel segmentation result.
Based on the foregoing embodiment, as an optional embodiment, the obtaining unit 601 is specifically configured to:
acquiring a medical image map;
and performing vessel segmentation on the same medical image by training two deep learning models with different segmentation accuracies or adopting different segmentation thresholds, and obtaining a rough vessel segmentation result and a fine vessel segmentation result of the medical image.
Based on the foregoing embodiment, as an optional embodiment, the determining unit 602 is specifically configured to:
making a difference between the rough blood vessel segmentation result and the fine blood vessel segmentation result, and setting voxels with difference values smaller than zero as zero to obtain redundant blood vessel parts in the rough blood vessel segmentation result;
and analyzing the connected domain of the redundant blood vessel part in the rough blood vessel segmentation result, and dividing mutually connected voxel points into the same region to form a plurality of potential blood vessel fracture regions.
Based on the foregoing embodiment, as an optional embodiment, the removing unit 603 is specifically configured to:
determining non-vascular rupture regions in the potential vascular rupture regions;
removing the non-vascular rupture region from the potential vascular rupture region to yield a preserved vascular rupture region;
wherein the non-vascular rupture region comprises: small area vascular noise, vessel wall area, terminal unwanted branch vessel area, and stuck vessel branches.
Based on the foregoing embodiment, as an optional embodiment, the fusion unit 604 is specifically configured to:
and merging the fine blood vessel segmentation result with the reserved blood vessel fracture area to obtain a final blood vessel segmentation result.
Referring to fig. 7, fig. 7 is a schematic structural diagram of a terminal system 700 according to an embodiment of the present disclosure, where the terminal system 700 can be used to execute the software multi-language display and input synchronization switching method according to the embodiment of the present disclosure.
The terminal system 700 may include: a processor 701, a memory 702, and a communication unit 703. The components communicate via one or more buses, and those skilled in the art will appreciate that the architecture of the servers shown in the figures is not intended to be limiting, and may be a bus architecture, a star architecture, a combination of more or less components than those shown, or a different arrangement of components.
The memory 702 may be used for storing instructions executed by the processor 701, and the memory 702 may be implemented by any type of volatile or non-volatile storage terminal or combination thereof, such as Static Random Access Memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk. The execution instructions in memory 702, when executed by processor 701, enable terminal system 700 to perform some or all of the steps in the method embodiments described below.
The processor 701 is a control center of the storage terminal, connects various parts of the entire electronic terminal using various interfaces and lines, and performs various functions of the electronic terminal and/or processes data by operating or executing software programs and/or modules stored in the memory 702 and calling data stored in the memory. The processor may be composed of an Integrated Circuit (IC), for example, a single packaged IC, or a plurality of packaged ICs connected with the same or different functions. For example, the processor 701 may include only a Central Processing Unit (CPU). In the embodiment of the present invention, the CPU may be a single operation core, or may include multiple operation cores.
A communication unit 703, configured to establish a communication channel so that the storage terminal can communicate with other terminals. And receiving user data sent by other terminals or sending the user data to other terminals.
The present application also provides a computer storage medium, wherein the computer storage medium may store a program, and the program may include some or all of the steps in the embodiments provided by the present invention when executed. The storage medium may be a magnetic disk, an optical disk, a read-only memory (ROM) or a Random Access Memory (RAM).
According to the method and the device, the coarse and fine granularity segmentation results are fused, the rough blood vessel segmentation results are utilized to repair the blood vessel fracture area in the fine blood vessel segmentation, meanwhile, the introduction of redundant noise in the rough blood vessel segmentation results is avoided, the blood vessel segmentation effect is improved, and the connectivity of blood vessels is ensured.
The embodiments are described in a progressive manner in the specification, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other. For the system provided by the embodiment, the description is relatively simple because the system corresponds to the method provided by the embodiment, and the relevant points can be referred to the method part for description.
The principles and embodiments of the present application are explained herein using specific examples, which are provided only to help understand the method and the core idea of the present application. It should be noted that, for those skilled in the art, it is possible to make several improvements and modifications to the present application without departing from the principle of the present application, and such improvements and modifications also fall within the scope of the claims of the present application.
It is further noted that, in the present specification, relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.
Claims (12)
1. A blood vessel segmentation method based on coarse and fine granularity fusion is characterized by comprising the following steps:
acquiring a medical image, and performing blood vessel thickness and granularity segmentation on the medical image to obtain a rough blood vessel segmentation result and a fine blood vessel segmentation result;
determining a potential blood vessel fracture area according to the rough blood vessel segmentation result and the fine blood vessel segmentation result;
determining a non-vascular rupture region in the potential vascular rupture region and removing the non-vascular rupture region from the potential vascular rupture region to obtain a preserved vascular rupture region;
and fusing the reserved blood vessel fracture area and the fine blood vessel segmentation result to repair the blood vessel fracture area in the fine blood vessel segmentation result to obtain a final blood vessel segmentation result.
2. The method for vessel segmentation based on coarse-fine particle fusion as claimed in claim 1, wherein the obtaining of the medical image map and the vessel coarse-fine particle segmentation of the medical image map to obtain the coarse vessel segmentation result and the fine vessel segmentation result comprises:
acquiring a medical image map;
and performing vessel segmentation on the same medical image by training two deep learning models with different segmentation accuracies or adopting different segmentation thresholds, and obtaining a rough vessel segmentation result and a fine vessel segmentation result of the medical image.
3. The vessel segmentation method based on coarse-fine granularity fusion according to claim 1, wherein the determining a potential vessel fracture region according to the coarse vessel segmentation result and the fine vessel segmentation result comprises:
making a difference between the rough blood vessel segmentation result and the fine blood vessel segmentation result, and setting voxels with difference values smaller than zero as zero to obtain redundant blood vessel parts in the rough blood vessel segmentation result;
and analyzing the connected domain of the redundant blood vessel part in the rough blood vessel segmentation result, and dividing mutually connected voxel points into the same region to form a plurality of potential blood vessel fracture regions.
4. The vessel segmentation method based on coarse-fine particle fusion according to claim 1, wherein the determining of the non-vessel fracture region in the potential vessel fracture region and the removing of the non-vessel fracture region from the potential vessel fracture region result in a preserved vessel fracture region comprises:
determining non-vascular rupture regions in the potential vascular rupture regions;
removing the non-vascular rupture region from the potential vascular rupture region to yield a preserved vascular rupture region;
wherein the non-vascular rupture region comprises: small area vascular noise, vessel wall area, terminal unwanted branch vessel area, and stuck vessel branches.
5. The method for vessel segmentation based on coarse-fine particle fusion according to claim 4, wherein the method for determining the small region vessel noise comprises:
acquiring the number of voxels of each region of the potential vessel fracture region;
and judging whether the number of voxels in the region is less than a preset number, if so, judging that the region is small-region vascular noise.
6. The method for vessel segmentation based on coarse-and-fine-grained fusion according to claim 4, wherein the method for determining the vessel wall region comprises:
acquiring the total number of voxel points of each region of the potential vessel fracture region;
judging whether each individual voxel point in each region of the blood vessel fracture region has adjacent blood vessel voxels in the fine blood vessel segmentation result, if so, acquiring the number of voxel points of the adjacent blood vessel voxels in the region;
calculating the ratio of the number of voxel points of the voxels of the adjacent vessels in the region to the total number of the voxel points in the region;
and judging whether the voxel point ratio of the region exceeds a preset threshold value, if so, judging that the region is a vascular wall region.
7. The vessel segmentation method based on coarse-fine particle fusion according to claim 4, wherein the method for determining the terminal redundant branch vessel region comprises:
obtaining voxel points of adjacent vessel voxels in each region of the vessel fracture region;
and judging whether the voxel points of the regions only belong to the same connected region, if so, judging that the connected region is the only adjacent surface of the terminal blood vessel branch and the fine blood vessel, and judging that the region is the terminal redundant branch blood vessel region.
8. The method for vessel segmentation based on coarse-fine particle fusion according to claim 4, wherein the method for determining the stuck vessel branch comprises:
obtaining voxel points of adjacent vessel voxels in each region of the vessel fracture region;
obtaining a vessel voxel point adjacent to a voxel point in a vessel fracture region in a fine vessel segmentation result;
and judging whether the voxel points in the fine blood vessel segmentation result only belong to the same communicated region, if so, judging that the blood vessel branch is connected with the mutually communicated regions in the fine blood vessel segmentation, namely two ends of a U-shaped blood vessel or a plurality of adjacent blood vessels, and judging that the regions are adhered blood vessel branches.
9. The vessel segmentation method based on coarse-fine granularity fusion according to claim 1, wherein fusing the preserved vessel fracture region with the fine vessel segmentation result to repair the vessel fracture region in the fine vessel segmentation result to obtain a final vessel segmentation result comprises:
and merging the fine blood vessel segmentation result with the reserved blood vessel fracture area to obtain a final blood vessel segmentation result.
10. A vessel segmentation system based on coarse-fine granularity fusion is characterized by comprising:
the segmentation unit is configured to acquire a medical image map and perform blood vessel thickness-granularity segmentation on the medical image map to obtain a rough blood vessel segmentation result and a fine blood vessel segmentation result;
a determining unit configured to determine a potential vessel rupture region according to the coarse vessel segmentation result and the fine vessel segmentation result;
a removal unit configured to determine a non-vascular rupture region of the potential vascular rupture regions and remove the non-vascular rupture region from the potential vascular rupture region to obtain a preserved vascular rupture region;
and the fusion unit is configured to fuse the reserved blood vessel fracture region and the fine blood vessel segmentation result to repair the blood vessel fracture region in the fine blood vessel segmentation result to obtain a final blood vessel segmentation result.
11. A terminal, comprising:
a processor;
a memory for storing instructions for execution by the processor;
wherein the processor is configured to perform the method of any one of claims 1-9.
12. A computer-readable storage medium, in which a computer program is stored which, when being executed by a processor, carries out the method according to any one of claims 1-9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010272913.4A CN111539917B (en) | 2020-04-09 | 2020-04-09 | Blood vessel segmentation method, system, terminal and storage medium based on coarse-fine granularity fusion |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010272913.4A CN111539917B (en) | 2020-04-09 | 2020-04-09 | Blood vessel segmentation method, system, terminal and storage medium based on coarse-fine granularity fusion |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111539917A true CN111539917A (en) | 2020-08-14 |
| CN111539917B CN111539917B (en) | 2023-08-25 |
Family
ID=71977152
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010272913.4A Active CN111539917B (en) | 2020-04-09 | 2020-04-09 | Blood vessel segmentation method, system, terminal and storage medium based on coarse-fine granularity fusion |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111539917B (en) |
Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030099390A1 (en) * | 2001-11-23 | 2003-05-29 | Xiaolan Zeng | Lung field segmentation from CT thoracic images |
| US20150003697A1 (en) * | 2013-06-28 | 2015-01-01 | International Business Machines Corporation | Bidirectional blood vessel segmentation |
| CN105005998A (en) * | 2015-08-05 | 2015-10-28 | 大连理工大学 | Cerebrovascular image segmentation method based on multi-angle serialized image space feature point set |
| CN106097378A (en) * | 2016-07-24 | 2016-11-09 | 江西理工大学 | A kind of level set retinal vascular images dividing method merging shape prior |
| CN106875375A (en) * | 2016-12-28 | 2017-06-20 | 浙江工业大学 | Three-dimensional blood vessel axis detection method based on tubulose feature enhancing filtering and ridge line tracking |
| US20180028137A1 (en) * | 2016-07-30 | 2018-02-01 | Shanghai United Imaging Healthcare Co., Ltd. | Method and system for extracting lower limb vasculature |
| CN108109159A (en) * | 2017-12-21 | 2018-06-01 | 东北大学 | It is a kind of to increase the retinal vessel segmenting system being combined based on hessian matrixes and region |
| CN108198184A (en) * | 2018-01-09 | 2018-06-22 | 北京理工大学 | The method and system of contrastographic picture medium vessels segmentation |
| CN108428210A (en) * | 2017-02-15 | 2018-08-21 | 浙江京新术派医疗科技有限公司 | Blood-vessel image method for reconstructing and reconstructing device |
| CN108537802A (en) * | 2018-03-30 | 2018-09-14 | 北京灵医灵科技有限公司 | A kind of blood vessel segmentation method and device |
| CN108986106A (en) * | 2017-12-15 | 2018-12-11 | 浙江中医药大学 | Retinal vessel automatic division method towards glaucoma clinical diagnosis |
| CN109584223A (en) * | 2018-11-20 | 2019-04-05 | 北京中科研究院 | Pulmonary vascular dividing method in CT image |
| CN109886982A (en) * | 2019-04-24 | 2019-06-14 | 数坤（北京）网络科技有限公司 | A kind of blood-vessel image dividing method, device and computer memory device |
| CN109903298A (en) * | 2019-03-12 | 2019-06-18 | 数坤（北京）网络科技有限公司 | Restorative procedure, system and the computer storage medium of blood vessel segmentation image fracture |
| CN110827283A (en) * | 2019-10-28 | 2020-02-21 | 北京深睿博联科技有限责任公司 | Head and neck blood vessel segmentation method and device based on convolutional neural network |
-
2020
- 2020-04-09 CN CN202010272913.4A patent/CN111539917B/en active Active
Patent Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030099390A1 (en) * | 2001-11-23 | 2003-05-29 | Xiaolan Zeng | Lung field segmentation from CT thoracic images |
| US20150003697A1 (en) * | 2013-06-28 | 2015-01-01 | International Business Machines Corporation | Bidirectional blood vessel segmentation |
| CN105005998A (en) * | 2015-08-05 | 2015-10-28 | 大连理工大学 | Cerebrovascular image segmentation method based on multi-angle serialized image space feature point set |
| CN106097378A (en) * | 2016-07-24 | 2016-11-09 | 江西理工大学 | A kind of level set retinal vascular images dividing method merging shape prior |
| US20180028137A1 (en) * | 2016-07-30 | 2018-02-01 | Shanghai United Imaging Healthcare Co., Ltd. | Method and system for extracting lower limb vasculature |
| CN106875375A (en) * | 2016-12-28 | 2017-06-20 | 浙江工业大学 | Three-dimensional blood vessel axis detection method based on tubulose feature enhancing filtering and ridge line tracking |
| CN108428210A (en) * | 2017-02-15 | 2018-08-21 | 浙江京新术派医疗科技有限公司 | Blood-vessel image method for reconstructing and reconstructing device |
| CN108986106A (en) * | 2017-12-15 | 2018-12-11 | 浙江中医药大学 | Retinal vessel automatic division method towards glaucoma clinical diagnosis |
| CN108109159A (en) * | 2017-12-21 | 2018-06-01 | 东北大学 | It is a kind of to increase the retinal vessel segmenting system being combined based on hessian matrixes and region |
| CN108198184A (en) * | 2018-01-09 | 2018-06-22 | 北京理工大学 | The method and system of contrastographic picture medium vessels segmentation |
| CN108537802A (en) * | 2018-03-30 | 2018-09-14 | 北京灵医灵科技有限公司 | A kind of blood vessel segmentation method and device |
| CN109584223A (en) * | 2018-11-20 | 2019-04-05 | 北京中科研究院 | Pulmonary vascular dividing method in CT image |
| CN109903298A (en) * | 2019-03-12 | 2019-06-18 | 数坤（北京）网络科技有限公司 | Restorative procedure, system and the computer storage medium of blood vessel segmentation image fracture |
| CN109886982A (en) * | 2019-04-24 | 2019-06-14 | 数坤（北京）网络科技有限公司 | A kind of blood-vessel image dividing method, device and computer memory device |
| CN110827283A (en) * | 2019-10-28 | 2020-02-21 | 北京深睿博联科技有限责任公司 | Head and neck blood vessel segmentation method and device based on convolutional neural network |
Non-Patent Citations (1)
| Title |
|---|
| 张也;张勇德;沙宪政;: "基于多尺度匹配滤波法的视网膜血管分割" * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111539917B (en) | 2023-08-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110298844B (en) | X-ray radiography image blood vessel segmentation and identification method and device | |
| CN111768411A (en) | Coronary artery central line extraction method and device, computer equipment and storage medium | |
| CN112734774B (en) | High-precision fundus blood vessel extraction method, device, medium, equipment and system | |
| CN113138982B (en) | Big data cleaning method | |
| CN113436166A (en) | Intracranial aneurysm detection method and system based on magnetic resonance angiography data | |
| JP6713690B2 (en) | Blood vessel shape constructing apparatus, method thereof, and computer software program | |
| CN110706272B (en) | Blood vessel centerline extraction method and device, electronic equipment and storage medium | |
| CN110634490B (en) | Voiceprint identification method, device and equipment | |
| CN113470102A (en) | Method, device, medium and equipment for measuring fundus blood vessel curvature with high precision | |
| CN112734828A (en) | Method, device, medium and equipment for determining center line of fundus blood vessel | |
| CN111383191B (en) | Image processing method and device for vascular fracture repair | |
| CN113450329B (en) | Microcirculation image blood vessel branch erythrocyte flow rate calculation method and system | |
| CN112734773B (en) | Sub-pixel-level fundus blood vessel segmentation method, device, medium and equipment | |
| CN111539917A (en) | Blood vessel segmentation method, system, terminal and storage medium based on coarse and fine granularity fusion | |
| CN113763331A (en) | Coronary artery dominant type determination method, device, storage medium, and electronic apparatus | |
| CN113077418A (en) | CT image skeleton segmentation method and device based on convolutional neural network | |
| JP7240199B2 (en) | Parameter determination method, parameter determination device, and program | |
| CN114972220B (en) | Image processing method and device, electronic equipment and readable storage medium | |
| CN111383766A (en) | Computer data processing method, device, medium and electronic equipment | |
| CN116091702A (en) | Vascular structure reconstruction method, device, equipment and storage medium | |
| CN114332082A (en) | Definition evaluation method and device, electronic equipment and computer storage medium | |
| CN114387209A (en) | Method, apparatus, medium, and device for fundus structural feature determination | |
| CN111753598A (en) | Face detection method and device | |
| CN117495951B (en) | Intracranial aneurysm positioning method, device, computer equipment and storage medium | |
| CN116958692A (en) | Angiography image key frame selection method, angiography image key frame selection device, angiography image key frame selection equipment and angiography image key frame selection medium |
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 |