CN116453700A - Multi-level hierarchical structural typing method based on vascular characteristics - Google Patents

Multi-level hierarchical structural typing method based on vascular characteristics Download PDF

Info

Publication number
CN116453700A
CN116453700A CN202310437404.6A CN202310437404A CN116453700A CN 116453700 A CN116453700 A CN 116453700A CN 202310437404 A CN202310437404 A CN 202310437404A CN 116453700 A CN116453700 A CN 116453700A
Authority
CN
China
Prior art keywords
blood vessel
lesion
vascular
lesions
bifurcation
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202310437404.6A
Other languages
Chinese (zh)
Inventor
吕江
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.)
Jiangsu Jinma Yangming Information Technology Co ltd
Original Assignee
Jiangsu Jinma Yangming Information Technology 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 Jiangsu Jinma Yangming Information Technology Co ltd filed Critical Jiangsu Jinma Yangming Information Technology Co ltd
Priority to CN202310437404.6A priority Critical patent/CN116453700A/en
Publication of CN116453700A publication Critical patent/CN116453700A/en
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/50—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for simulation or modelling of medical disorders
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems

Landscapes

  • Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Public Health (AREA)
  • Data Mining & Analysis (AREA)
  • Medical Informatics (AREA)
  • Biomedical Technology (AREA)
  • Pathology (AREA)
  • Theoretical Computer Science (AREA)
  • Epidemiology (AREA)
  • General Health & Medical Sciences (AREA)
  • Primary Health Care (AREA)
  • Physics & Mathematics (AREA)
  • Databases & Information Systems (AREA)
  • Probability & Statistics with Applications (AREA)
  • Artificial Intelligence (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Apparatus For Radiation Diagnosis (AREA)

Abstract

The invention relates to the field of computer aided design, in particular to a multi-level hierarchical structural typing method based on vascular features, which comprises the following steps: s1, constructing layering typing of bifurcation lesions based on abnormal blood vessel characteristics; s2, extracting characteristics of vascular bifurcation lesions; the method specifically comprises the following steps: s2a, acquiring profile data of a bifurcated blood vessel to be processed, and extracting characteristics of the bifurcated lesion of the blood vessel, wherein the extracted characteristics comprise the diameter of the blood vessel, the stenosis degree of the blood vessel, the stenosis length and the included angle of the blood vessel; s2b, segmenting the bifurcated blood vessel, and establishing a Markov decision to screen blood vessel characteristic parameter data; s3, selecting effective characteristics to identify the bifurcation lesion type. The invention can intuitively express the characteristics of vascular lesions, assists doctors in diagnosing the types of vascular bifurcation lesions, and has important significance for determining the interventional treatment strategy of the bifurcation lesions.

Description

Multi-level hierarchical structural typing method based on vascular characteristics
Technical Field
The invention relates to the technical field of computer aided design, in particular to a multi-level hierarchical structural typing method based on vascular features.
Background
In recent years, the number of cases of bifurcated lesion intervention has shown an increasing trend. Before interventional therapy, a doctor needs to observe angiographic images to solve vascular stenosis, and the interventional technology is judged according to personal experience, but the visual estimation or manual measurement of an expert has differences and errors due to subjective judgment. At present, a plurality of lesion typing methods with wide clinical application are available, for example, according to whether a bifurcation lesion is found or not, although intervention risks of some bifurcation lesions can be predicted, the method has guiding significance but is insufficient in evaluating a plurality of characteristics of the bifurcation lesions; the method for parting according to the bifurcation angle of the branch vessel aims at guiding operations such as vascular ducts and the like, and is effectively but mostly used for preliminary parting and evaluation of bifurcation lesions; the method for parting according to the distribution characteristics of the plaque has the advantages of emphasizing the importance of the lesion distribution to the bifurcation lesions, but not including all bifurcation lesions; the method of systematic typing according to the lesion site is capable of covering all bifurcation lesions, but does not emphasize the importance of the branch reference diameter. Although the above methods are widely used, most of them are proposed in the age of bare metal stents, and are not suitable for treating complex lesions at present. In addition, many classification methods are cumbersome, memory and use are not suitable, and some classification methods involve more characteristics, but some classification methods have little relation with interventional therapy decision-making and lack of relation between actual lesion characteristics.
The traditional identification work of many medical images is based on a manual processing method, and the accuracy rate is greatly dependent on the expertise of medical staff. In the process of image auxiliary diagnosis, the judging levels of the imaging examination results by different imaging doctors are different, certain subjectivity and error exist in the judgment of the micro lesions, and the misjudgment and misdiagnosis phenomenon often occurs, so that adverse effects are caused to patients. Meanwhile, the technical efficiency of selecting the vascular stent in a mode of manual judgment by doctors is low, and time and labor are wasted.
Disclosure of Invention
The invention aims to provide a multi-level hierarchical structural type typing method based on vascular characteristics, aiming at the problem that the relevance between the identified vascular disease change typing and an interventional therapy strategy is not high in the background art.
The technical scheme of the invention is as follows: a multi-level hierarchical structural typing method based on vascular features comprises the following steps:
s1, constructing layering typing of bifurcation lesions based on abnormal blood vessel characteristics; the method specifically comprises the following steps: s1a, determining the position type of the vascular lesion by using Medina typing; s1b, based on lesion position types, selecting a lesion parting design element to further classify blood vessels by combining important features related to bifurcation lesions and interventional therapy decisions; s1c, setting corresponding blood vessel characteristic classification conditions according to disease transformation type design factors, and constructing a multi-level blood vessel disease transformation type method;
s2, extracting characteristics of vascular bifurcation lesions; the method specifically comprises the following steps: s2a, acquiring profile data of a bifurcated blood vessel to be processed, and extracting characteristics of the bifurcated lesion of the blood vessel, wherein the extracted characteristics comprise the diameter of the blood vessel, the stenosis degree of the blood vessel, the stenosis length and the included angle of the blood vessel; s2b, segmenting the bifurcated blood vessel, and establishing a Markov decision to screen blood vessel characteristic parameter data;
s3, selecting effective characteristics to identify the bifurcation lesion type.
Preferably, in S1a, traversing and screening the characteristic parameter set of the separated blood vessel to obtain the stenosis degree of the blood vessel, judging whether the stenosis degree is more than 50% by adopting Medina blood vessel typing, respectively marking whether the blood vessel is provided with lesions or not by using 1 and 0, and determining the position type of the lesions of the blood vessel; in S1b, based on the blood vessel lesion type marked in S1a, combining important features related to bifurcation lesions and interventional therapy decisions, dividing the design factors of lesion classification into four aspects of lesion part features, lesion branch number, whether lesions are true and side branch protection factors, and further classifying blood vessels.
Preferably, in S1b, the lesion site feature refers to whether a main branch has a lesion; the number of lesion branches refers to the number of lesions present in the vessel branches; whether the lesion is truly indicative of whether the diameter of the vessel of the main stent containing the lesion is greater than 2mm; the side branch protection factor refers to whether the diameter of a branch lesion is larger than 2mm and the length of the branch lesion is larger than 10mm when the lesion exists in the blood vessel branch, and if the conditions are met, the blood vessel lesion side branch needs to be protected.
Preferably, in S2, the parameters of the vascular lesion include the diameter of each segment of the blood vessel, the stenosis degree of the blood vessel, the length parameter of the stenosis and the bifurcation angle; the diameters of the blood vessel segments are calculated by extracting the central line of the blood vessel, determining the diameter of the far end and the near end according to the change of the diameters of the blood vessel segments by segmenting the blood vessel, finding a narrow area, calculating the stenosis degree of the blood vessel, calculating the arc of the narrow area of the blood vessel according to the length parameter of the narrow area of the blood vessel and the far end and the near end diameter of the narrow area of the blood vessel, determining the main branch of the blood vessel according to Murray law by the bifurcation angle, finding the position of the branch of the blood vessel, and calculating the branch included angle.
Preferably, the calculation formula of the parameters of the vascular lesions is as follows:
(1) Diameter of each segment of blood vessel: wherein D is I Represents the diameter of the first branch, n represents the total number of the diameters of the blood vessels of the section, d i Diameter at each point:
(2) Vascular stenosis: s is S roi Represents the degree of stenosis, d A And d B Is the diameter of the far and near ends of the stenosis, d C Is the narrowest diameter:
(3) Length of stenosis: the central line length of blood vessel is determined by the pixel point sequence [ (x) i ，y i )]Wherein i= [1, n]Arc length is L c ：
(4) Bifurcation included angle: the center line corresponding point set is P i+m ，P i 、P i+n 、P i+m P i For the fitted branches, θ nm Is the angle between branches m and n:
preferably, in S3, path selection and feature selection operations are performed on the blood vessel feature parameter data through markov decision, so as to obtain an empirical sequence track formed by each blood vessel classification, and the empirical sequence track is used as a recognition basis of the blood vessel feature data, a lesion segment screening strategy is established, important features in a blood vessel lesion segment are extracted, and the blood vessel lesion type is recognized through a correspondence between a layering parting method and the blood vessel lesion features.
Preferably, the empirical sequence trajectory formed by each vessel classification is obtained by Markov decision and is written as: { S, A, r }, A represents the action set, and the expression is A= { a 1 ,a 2 ,...a n And }, wherein a i Representing actions taken in time state i; with corresponding r as the prize, i.e. in state s i Executing action a i The prize r obtained by the transition to state s' thereafter i ＝R(s i +a i ) Finally, an empirical track sequence consisting of S, A and r is generated.
Preferably, in S3, the blood vessel features are analyzed by using markov decision, so that each segment of blood vessel segment has a unique experience track, the blood vessel bifurcation structure is segmented by comparing with the normal blood vessel features, and the abnormal features are selected to form corresponding feature combinations, wherein the feature combinations are defined as follows: MDP (U, X, p, r) = { U, a, r }.
Compared with the prior art, the invention has the following beneficial technical effects: a multi-level hierarchical structural typing method for vascular features is disclosed. Firstly, acquiring contour data of a bifurcated blood vessel to be processed, and extracting the characteristics of vascular lesions; then, determining the position type of the vascular lesion by a Medina typing method according to the parameters of the vascular stenosis degree, and constructing a multi-level bifurcation lesion typing method by combining the significant characteristics of bifurcation lesions and stent technology, so as to further classify the bifurcation lesions; finally, taking the established multi-level disease transformation typing method as a guiding idea, and carrying out path selection and feature selection operation on the blood vessel feature data through a Markov decision process to realize efficient identification of the blood vessel bifurcation lesion type so as to achieve the purpose of assisting a doctor in carrying out effective and rapid diagnosis. The invention not only can help identify the type of the bifurcation vascular lesion, but also can cover various bifurcation lesion characteristics, and has important significance for guiding interventional therapy operation.
Drawings
FIG. 1 is a flow chart of an embodiment of the present invention;
FIG. 2 is a schematic diagram of hierarchical typing of bifurcation lesions according to the present invention;
fig. 3 is a schematic diagram of path selection for determining a type of vascular lesion according to the present invention.
Detailed Description
Example 1
As shown in fig. 1, the multi-level hierarchical structural typing method based on vascular features provided by the invention comprises the following steps:
s1, constructing layering typing of bifurcation lesions based on abnormal blood vessel characteristics; the method specifically comprises the following steps: s1a, determining the position type of the vascular lesion by using Medina typing; s1b, based on lesion position types, selecting a lesion parting design element to further classify blood vessels by combining important features related to bifurcation lesions and interventional therapy decisions; s1c, setting corresponding blood vessel characteristic classification conditions according to disease transformation type design factors, and constructing a multi-level blood vessel disease transformation type method; the bifurcation lesions classifying step of the Medina classification according to the vasculopathy elements is as follows:
(1) Lesion features. Lefevre et al consider bifurcation lesions distribution characteristics but are more classified. Depending on whether there is a lesion in the main branch as the primary feature, medina is classified into a bifurcation lesion feature containing the main branch and a vascular lesion containing only the side branch.
(2) Number of lesion branches. The number of vascular side branch lesions affects the number of stent interventions, and the two types are divided into single main branch lesions, main branch lesions involving branches, and single (multi) side branch lesions, respectively, according to the number of side branch lesions.
(3) Whether the lesion is true. The reference diameter of the blood vessel bifurcation influences whether a stent is placed or not, and the relation between the reference diameter of the blood vessel containing the lesion of the main body and 2mm is judged, and the blood vessel containing the lesion of the main body is divided into true vascular lesions and pseudo vascular lesions; further judging the lesions of the main branch and the branch in turn for the lesions of the blood vessel of the affected branch, and dividing the lesions into false affected branch lesions, false bifurcation lesions and true bifurcation lesions; the single-side branch and the multi-side branch bifurcation lesions are sequentially traversed by the blood vessel reference diameters of different branches, and are divided into false-side branch lesions, true-side branch lesions, false-side branch lesions and true-side branch lesions.
(4) Edge support protection factors. For the true bifurcation lesions and true bifurcation lesions in the above-mentioned typing, the true bifurcation lesions are classified into normal true bifurcation lesions, long-side bifurcation lesions, difficult bifurcation lesions and complex bifurcation lesions by entering side branches according to whether the branch reference diameter is greater than 2mm, whether the length of the narrow area is greater than 10mm, and whether the bifurcation included angle is greater than 70 or less than 40.
S2, extracting characteristics of vascular bifurcation lesions; the method specifically comprises the following steps: s2a, acquiring profile data of a bifurcated blood vessel to be processed, and extracting characteristics of the bifurcated lesion of the blood vessel, wherein the extracted characteristics comprise the diameter of the blood vessel, the stenosis degree of the blood vessel, the stenosis length and the included angle of the blood vessel; s2b, segmenting the bifurcated blood vessel, and establishing a Markov decision to screen blood vessel characteristic parameter data;
s3, selecting effective characteristics to identify the bifurcation lesion type. Specifically, path selection and feature selection operation are carried out on blood vessel feature parameter data through Markov decision, an experience sequence track formed by each blood vessel classification is obtained, a lesion segment screening strategy is established as a recognition basis of the blood vessel feature data, important features in a blood vessel lesion segment are extracted, and the blood vessel lesion type is recognized through a corresponding relation between a layering parting method and the blood vessel lesion features. The empirical sequence trajectory formed by each vessel classification is obtained by markov decision and is noted as: { S, A, r }, A represents the action set, and the expression is A= { a 1 ,a 2 ,...a n And }, wherein a i Representing actions taken in time state i; with corresponding r as the prize, i.e. in state s i Executing action a i The prize r obtained by the transition to state s' thereafter i ＝R(s i +a i ) Finally, an empirical track sequence consisting of S, A and r is generated. Analyzing the blood vessel characteristics by using Markov decision, achieving that each segment of blood vessel segment has an experience track, segmenting the blood vessel bifurcation structure by comparing with the normal blood vessel characteristics, selecting abnormal characteristics to form corresponding characteristic combinations, wherein the definition of the characteristic combinations is as follows: MDP (U, X, p, r) = { U, a, r }.
Example two
Compared with the first embodiment, in the S2, the vascular lesion parameters comprise the diameter of each segment of blood vessel, the stenosis degree of the blood vessel, the length parameter of the stenosis and the bifurcation angle; the diameters of the blood vessel segments are calculated by extracting the central line of the blood vessel, determining the diameter of the far end and the near end according to the change of the diameters of the blood vessel segments by segmenting the blood vessel, finding a narrow area, calculating the stenosis degree of the blood vessel, calculating the arc of the narrow area of the blood vessel according to the length parameter of the narrow area of the blood vessel and the far end and the near end diameter of the narrow area of the blood vessel, determining the main branch of the blood vessel according to Murray law by the bifurcation angle, finding the position of the branch of the blood vessel, and calculating the branch included angle.
The calculation formula of the parameters of the vascular lesions is as follows:
(1) Diameter of each segment of blood vessel: wherein D is I Represents the diameter of the first branch, n represents the total number of the diameters of the blood vessels of the section, d i Diameter at each point:
(2) Blood vessel stenosis degree: s is S roi Represents the degree of stenosis, d A And d B Is the diameter of the far and near ends of the stenosis, d C Is the narrowest diameter:
(3) Length of stenosis: the central line length of blood vessel is determined by the pixel point sequence [ (x) i ，y i )]Wherein i= [1, n]Arc length is L c ：
(4) Bifurcation included angle: the center line corresponding point set is P i+m ，P i 、P i+n 、P i+m P i For the fitted branches, θ nm Is the angle between branches m and n:
example III
As shown in fig. 2-3, in the multi-level hierarchical structural typing method based on vascular features provided by the present invention, compared with the first embodiment, in S2b, a vascular bifurcation structure track which is subjected to all states and behaviors is obtained by using vascular data, and important features to be examined are identified by vascular segments, and whether a learning process is terminated or not is judged as a state, as shown in fig. 3, definition of U is as follows:
X i ＝{x i1 ,x i2 ,...x in (v) is the data in each step of the target vessel segment, which is used as input data, herein U 1 The obtained experience track sequence comprises the following specific steps:
step1: at this time the state is u 0 Action a is performed 0 (whether or not narrow: yes), state transitions to u 1 (stenosis position) a prize of 1 is obtained (stenosis degree: less than 50% prize is-1; greater than 50% prize is 1).
Step2: at this time the state is u 1 Action a is performed 1 (stenosis position: middle), state transition to u 2 (stenosis degree) gets a prize of 1 (stenosis position: front prize of-1; rear prize of 0; middle prize of 1).
Step3: at this time the state is u 2 Action a is performed 2 (stenosis degree: greater than 50%), state transition to u 3 (average diameter) a prize of 1 is obtained (stenosis: less than 50% prize of-1; greater than 50% prize of 1).
Step4: at this time the state is at u 3 Action a is performed 3 (average diameter: greater than 2 mm), transition of state to u 4 (end), a prize value of 1 is obtained (prize-1 for average diameters less than 2mm and prize 1 for average diameters greater than 2 mm).
X i ＝{x i1 ,x i2 ,...x in (v) is the data in each step of the target vessel segment, which is used as input data, herein U 1 The obtained empirical trace sequence trace is: u (u) 0 (whether or not it is narrow), a 0 (Yes), r 0 ，u 1 (extent of stenosis), a 1 (greater than 50%), r 1 ，u 2 (average diameter), a 2 (greater than 2 mm), r 2 ，u 3 (termination).
Example IV
As shown in fig. 2-3, compared with the first embodiment or the third embodiment, in the S3, the multi-level hierarchical structural typing method based on vascular features provided by the invention establishes a lesion feature screening strategy through vascular abnormal features, extracts the abnormal features to form corresponding feature combinations, and uses the established multi-level lesion typing method as a guiding idea to realize efficient recognition of the type of vascular bifurcation lesions.
In order to extract the vascular lesion characteristics of the bifurcation structure, the vascular characteristics and the included angles of the blood vessels are combined to construct a combined model which is represented by gamma, wherein U represents the state set of the bifurcation structure and U 1 、U 2 、U 3 Respectively represent the state set, alpha k As a bifurcation angle function, as shown in the following formula
Further, traversing step length characteristics in each blood vessel segment according to the characteristic that abnormal characteristics of vascular lesions have a definite value range and are limited, screening obtained U characteristics, selecting blood vessel segments with stenosis as an initial set, combining blood vessel segments adjacent to the head-tail stenosis on the basis of the initial set to serve as new segments, and re-extracting characteristics in the new segments to serve as target blood vessel segment screening characteristics; if the integral vascular stenosis characteristics exist in the vascular segment, the integral vascular stenosis characteristics are taken as target screening characteristic segments. After obtaining each segment of target lesion blood vessel segment, updating the blood vessel bifurcation structure feature set as shown in the following formula, wherein U' is the feature after screening.
γ′＝{U′,α k }
The blood vessel segment features are composed of blood vessel segment stenosis related parameters, blood vessel segment diameter and other attributes, and are used for influencing the diagnosis and treatment of vascular lesionsu ij = { λ, β, γ, μ, δ } represents. In order to facilitate the expression of the characteristic attribute, the expression mode of the formed blood vessel characteristic is shown in the following formula; for more concise expression of the characteristics of the blood vessel segment, F is adopted 1 -F 5 Vascular features are expressed. F (F) 1 -F 5 In the above, Y/N is used to indicate whether the narrow is narrow, H/M/T is used to indicate "head/middle/tail", the third bit 1/0 is used to indicate "the narrow degree is more than 50%/less than 50%", the fourth bit 1/0 is used to indicate the diameter is more than 2 mm/less than 2mm, and the fifth bit 1/0 is used to indicate the branch narrow length is more than 10 mm/less than 10mm.
The symbols mean whether λ is narrow, β is narrow, γ is narrow, μ is the mean diameter of the vessel segment, δ is narrow.
And obtaining lesion characteristics of each blood vessel segment through a lesion segment screening model, wherein each segment of blood vessel lesion characteristic identification has a unique track and a corresponding abnormal characteristic result. After the significant feature set of the vascular lesions is obtained, the identification of the bifurcation vascular lesion type is realized by using a layering vascular lesion typing method.
In summary, the invention provides a multi-level hierarchical structural typing method for vascular features. Firstly, acquiring contour data of a bifurcated blood vessel to be processed, and extracting the characteristics of vascular lesions; then, determining the position type of the vascular lesion by a Medina typing method according to the parameters of the vascular stenosis degree, and constructing a multi-level bifurcation lesion typing method by combining the significant characteristics of bifurcation lesions and stent technology, so as to further classify the bifurcation lesions; finally, taking the established multi-level disease transformation typing method as a guiding idea, and carrying out path selection and feature selection operation on the blood vessel feature data through a Markov decision process to realize efficient identification of the blood vessel bifurcation lesion type so as to achieve the purpose of assisting a doctor in carrying out effective and rapid diagnosis. The invention not only can help identify the type of the bifurcation vascular lesion, but also can cover various bifurcation lesion characteristics, and has important significance for guiding interventional therapy operation.
The embodiments of the present invention have been described in detail with reference to the drawings, but the present invention is not limited thereto, and various changes can be made within the knowledge of those skilled in the art without departing from the spirit of the present invention.

Claims (8)

1. The multi-level hierarchical structural typing method based on the vascular characteristics is characterized by comprising the following steps of:
s1, constructing layering typing of bifurcation lesions based on abnormal blood vessel characteristics; the method specifically comprises the following steps: s1a, determining the position type of the vascular lesion by using Medina typing; s1b, based on lesion position types, selecting a lesion parting design element to further classify blood vessels by combining important features related to bifurcation lesions and interventional therapy decisions; s1c, setting corresponding blood vessel characteristic classification conditions according to disease transformation type design factors, and constructing a multi-level blood vessel disease transformation type method;
s2, extracting characteristics of vascular bifurcation lesions; the method specifically comprises the following steps: s2a, acquiring profile data of a bifurcated blood vessel to be processed, and extracting characteristics of the bifurcated lesion of the blood vessel, wherein the extracted characteristics comprise the diameter of the blood vessel, the stenosis degree of the blood vessel, the stenosis length and the included angle of the blood vessel; s2b, segmenting the bifurcated blood vessel, and establishing a Markov decision to screen blood vessel characteristic parameter data;
s3, selecting effective characteristics to identify the bifurcation lesion type.
2. The multi-level hierarchical structure type typing method based on vascular features of claim 1, wherein in S1a, traversing and screening the characteristic parameter set of the segmented blood vessels to obtain the stenosis degree of the blood vessels, judging whether the stenosis degree is more than 50% by adopting Medina blood vessel typing, respectively marking whether the blood vessels are diseased or not by using 1 and 0, and determining the position type of the vascular lesions; in S1b, based on the blood vessel lesion type marked in S1a, combining important features related to bifurcation lesions and interventional therapy decisions, dividing the design factors of lesion classification into four aspects of lesion part features, lesion branch number, whether lesions are true and side branch protection factors, and further classifying blood vessels.
3. The method of claim 2, wherein in S1b, the lesion site feature refers to whether a main branch has a lesion; the number of lesion branches refers to the number of lesions present in the vessel branches; whether the lesion is truly indicative of whether the diameter of the vessel of the main stent containing the lesion is greater than 2mm; the side branch protection factor refers to whether the diameter of a branch lesion is larger than 2mm and the length of the branch lesion is larger than 10mm when the lesion exists in the blood vessel branch, and if the conditions are met, the blood vessel lesion side branch needs to be protected.
4. The method of claim 1, wherein in S2, the parameters of vascular lesions include diameter of each segment, stenosis degree of the blood vessel, length parameter of the stenosis, and bifurcation angle; the diameters of the blood vessel segments are calculated by extracting the central line of the blood vessel, determining the diameter of the far end and the near end according to the change of the diameters of the blood vessel segments by segmenting the blood vessel, finding a narrow area, calculating the stenosis degree of the blood vessel, calculating the arc of the narrow area of the blood vessel according to the length parameter of the narrow area of the blood vessel and the far end and the near end diameter of the narrow area of the blood vessel, determining the main branch of the blood vessel according to Murray law by the bifurcation angle, finding the position of the branch of the blood vessel, and calculating the branch included angle.
5. The method of claim 4, wherein the calculation formula of the parameters of the vascular lesions is as follows:
(1) Diameter of each segment of blood vessel: wherein D is I Represents the diameter of the first branch, n represents the total number of the diameters of the blood vessels of the section, d i Diameter at each point:
(2) Vascular stenosis: s is S roi Represents the degree of stenosis, d A And d B Is the diameter of the far and near ends of the stenosis, d C Is the narrowest diameter:
(3) Length of stenosis: the central line length of blood vessel is determined by the pixel point sequence [ (x) i ，y i )]Wherein i= [1, n]Arc length is L c ：
(4) Bifurcation included angle: the center line corresponding point set is P i+m ，P i 、P i+n 、P i+m P i For the fitted branches, θ nm Is the angle between branches m and n:
6. the multi-level hierarchical structural type typing method based on vascular features according to claim 1, wherein in S3, path selection and feature selection operations are performed on vascular feature parameter data through markov decision, an empirical sequence track formed by each vascular classification is obtained, the empirical sequence track is used as a recognition basis of the vascular feature data, a lesion segment screening strategy is established, important features in a vascular lesion segment are extracted, and the type of vascular lesions is recognized through a correspondence between a hierarchical typing method and the vascular lesion features.
7. The method of claim 6, wherein the empirical sequence trajectory for each vessel classification is obtained by markov decision and is expressed as: { S, A, r }, A represents the action set, and the expression is A= { a 1 ,a 2 ,...a n And }, wherein a i Representing actions taken in time state i; with corresponding r as the prize, i.e. in state s i Executing action a i Then transfer to the shapeThe prize r obtained for state s i ＝R(s i +a i ) Finally, an empirical track sequence consisting of S, A and r is generated.
8. The method for hierarchical structure typing of blood vessel features according to claim 1, wherein in S3, the blood vessel features are analyzed by using markov decision, so that each segment of blood vessel segment has an unique experience track, the blood vessel bifurcation structure is segmented by comparing with the normal blood vessel features, the abnormal features are selected to form corresponding feature combinations, and the feature combinations are defined as follows: MDP (U, X, p, r) = { U, a, r }.
CN202310437404.6A 2023-04-23 2023-04-23 Multi-level hierarchical structural typing method based on vascular characteristics Pending CN116453700A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310437404.6A CN116453700A (en) 2023-04-23 2023-04-23 Multi-level hierarchical structural typing method based on vascular characteristics

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310437404.6A CN116453700A (en) 2023-04-23 2023-04-23 Multi-level hierarchical structural typing method based on vascular characteristics

Publications (1)

Publication Number Publication Date
CN116453700A true CN116453700A (en) 2023-07-18

Family

ID=87131853

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310437404.6A Pending CN116453700A (en) 2023-04-23 2023-04-23 Multi-level hierarchical structural typing method based on vascular characteristics

Country Status (1)

Country Link
CN (1) CN116453700A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117274502A (en) * 2023-11-17 2023-12-22 北京唯迈医疗设备有限公司 Image processing method and device for assisting interventional operation

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117274502A (en) * 2023-11-17 2023-12-22 北京唯迈医疗设备有限公司 Image processing method and device for assisting interventional operation
CN117274502B (en) * 2023-11-17 2024-03-01 北京唯迈医疗设备有限公司 Image processing method and device for assisting interventional operation

Similar Documents

Publication Publication Date Title
CN109166124B (en) Retinal blood vessel morphology quantification method based on connected region
CN108805134B (en) Construction method and application of aortic dissection model
CN109300113B (en) Pulmonary nodule auxiliary detection system and method based on improved convex hull method
CN110652312B (en) Blood vessel CTA intelligent analysis system and application
CN107341812A (en) A kind of sequence Lung neoplasm image partition method based on super-pixel and Density Clustering
CN111598875A (en) Method, system and device for building thyroid nodule automatic detection model
Zeng et al. Liver vessel segmentation and identification based on oriented flux symmetry and graph cuts
CN113011509B (en) Lung bronchus classification method and device, electronic equipment and storage medium
CN116453700A (en) Multi-level hierarchical structural typing method based on vascular characteristics
KR20220108158A (en) Method for Severity Determination of Skin Diseases Based on Percentage of Body Surface Area Covered with Lesions
CN109087310B (en) Meibomian gland texture region segmentation method and system, storage medium and intelligent terminal
Perperidis et al. Automated detection of uninformative frames in pulmonary optical endomicroscopy
CN113160120A (en) Liver blood vessel segmentation method and system based on multi-mode fusion and deep learning
Guo et al. A bone age assessment system for real-world X-ray images based on convolutional neural networks
CN113077887A (en) Automatic quantitative analysis system and interpretation method for white matter lesions of brain
Yu et al. Break: Bronchi reconstruction by geodesic transformation and skeleton embedding
CN115222674A (en) Detection device for intracranial aneurysm rupture risk based on multi-dimensional feature fusion
CN116978543A (en) Artificial intelligent auxiliary marrow tumor pathological diagnosis device
CN114881975A (en) System, method, electronic device, and medium for predicting aneurysm rupture potential
CN114519705A (en) Ultrasonic standard data processing method and system for medical selection and identification
Ding et al. Classification of chromosome karyotype based on faster-rcnn with the segmatation and enhancement preprocessing model
CN117036288A (en) Tumor subtype diagnosis method for full-slice pathological image
TW202326615A (en) Angiography image determination method and angiography image determination device
CN113889238B (en) Image identification method and device, electronic equipment and storage medium
CN114119579A (en) Retina image main blood vessel identification method based on blood vessel structure similarity

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