CN110310256A - Coronary stenosis detection method, device, computer equipment and storage medium - Google Patents
Coronary stenosis detection method, device, computer equipment and storage medium Download PDFInfo
- Publication number
- CN110310256A CN110310256A CN201910464256.0A CN201910464256A CN110310256A CN 110310256 A CN110310256 A CN 110310256A CN 201910464256 A CN201910464256 A CN 201910464256A CN 110310256 A CN110310256 A CN 110310256A
- Authority
- CN
- China
- Prior art keywords
- network
- image
- coronary
- coronary artery
- stenosis
- 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
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
- 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
- G06T2207/20084—Artificial neural networks [ANN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30101—Blood vessel; Artery; Vein; Vascular
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiology & Medical Imaging (AREA)
- Quality & Reliability (AREA)
- Apparatus For Radiation Diagnosis (AREA)
Abstract
A kind of coronary stenosis detection method, device, computer equipment and storage medium provided by the present application, are input to coronary stenosis detection model by the image to be detected that will acquire, obtain testing result;Coronary stenosis detection model therein includes core network, segmentation network and narrow analysis network, and the output of core network is connect with the input of segmentation network and narrow analysis network respectively;Above-mentioned testing result includes coronary artery segmentation result and coronary stenosis result.It can be seen that by the structure of above-mentioned coronary stenosis detection model, the coronary stenosis detection model includes segmentation network and narrow analysis network, therefore, above-mentioned coronary stenosis detection model is a kind of multi task model, it can be realized simultaneously and image to be detected of input is split, and stenosis detection is carried out to image to be detected of input, obtain the testing result comprising at least two types, i.e. coronary artery segmentation result and coronary stenosis result.
Description
Technical field
This application involves technical field of medical detection more particularly to a kind of coronary stenosis detection method, device, computer to set
Standby and storage medium.
Background technique
With the development of medical detection technique, when carrying out Indexs measure to a variety of organs, for example, coronary artery is (referred to as
Coronary artery) Indexs measure, how fast and accurately to obtain coronary artery testing result becomes the problem of now comparing concern.
Currently, different types of coronary artery testing result can be obtained based on different models.Specifically, electronics can be passed through
The CTA image of Computed tomography (Computed Tomography angiography, CTA) equipment acquisition coronary artery
Data further can carry out coronary artery segmentation using CTA image data of the coronary artery parted pattern to acquisition, obtain coronary artery point
It cuts as a result, and positioning to obtain coronary artery using CTA image data progress narrow zone of the coronary stenosis zone location model to acquisition
The specific location of narrow zone.
But that there are committed memories is big for the method for above-mentioned determining coronary artery testing result, and it is slow to obtain coronary artery testing result speed
The problem of.
Summary of the invention
In a first aspect, a kind of coronary stenosis detection method, which comprises
Obtain image to be detected;
Image to be detected is input to coronary stenosis detection model, obtains testing result;Coronary stenosis detection model includes
Core network, segmentation network and narrow analysis network;The output of core network respectively with segmentation network and narrow analysis network
Input connection;Testing result includes coronary artery segmentation result and coronary stenosis result.
Narrow analysis network includes sorter network and Recurrent networks in one of the embodiments,;Coronary stenosis result packet
Include narrow classification and stenosis.
Coronary artery segmentation result includes probability respondence figure and coronary artery segmented image in one of the embodiments,.
Coronary stenosis detection model further includes pond layer in one of the embodiments, for defeated according to probability respondence figure
Region of interest area image out.
The output of pond layer is connect with the input of sorter network and Recurrent networks respectively in one of the embodiments,;Point
Class network is used to carry out classification processing to region of interest area image, obtains narrow classification;Recurrent networks are used for narrow to belonging to
Region of interest area image carries out stenosis estimation, obtains stenosis.
Core network from image to be detected for extracting characteristics of image in one of the embodiments,.
Coronary artery segmented image includes the position of key point in one of the embodiments,;Method further include:
The position of coronary artery center line in image to be detected is obtained according to probability respondence figure;
According to the position of the position of coronary artery center line and key point, coronary artery segmentation result is obtained.
Second aspect, a kind of coronary stenosis detection device, described device include:
Module is obtained, for obtaining image to be detected;
Detection module obtains testing result for image to be detected to be input to coronary stenosis detection model;Coronary stenosis
Detection model includes core network, segmentation network and narrow analysis network;The output of core network respectively with segmentation network and narrow
The input connection of narrow analysis network;Testing result includes coronary artery segmentation result and coronary stenosis result.
The third aspect, a kind of computer equipment, including memory and processor, the memory are stored with computer journey
Sequence, the processor realize coronary stenosis detection side described in first aspect any embodiment when executing the computer program
Method.
Fourth aspect, a kind of computer readable storage medium are stored thereon with computer program, the computer program quilt
Coronary stenosis detection method described in first aspect any embodiment is realized when processor executes.
A kind of coronary stenosis detection method, device, computer equipment and storage medium provided by the present application, by will acquire
To image to be detected be input to coronary stenosis detection model, obtain testing result;Coronary stenosis detection model therein includes
Core network, segmentation network and narrow analysis network, the output of core network respectively with segmentation network and narrow analysis network
Input connection;Above-mentioned testing result includes coronary artery segmentation result and coronary stenosis result.By above-mentioned coronary stenosis detection model
It includes segmentation network and narrow analysis network, therefore, above-mentioned coronary stenosis inspection that structure, which can be seen that the coronary stenosis detection model,
Survey model is a kind of multi task model, can be realized simultaneously and is split to image to be detected of input, and to input to
Detection image carries out stenosis detection, obtains the testing result comprising at least two types results, i.e. coronary artery segmentation result and coronary artery
Narrow result.In addition, due to segmentation network and it is narrow analysis network share Backbone network inside parameter, cause divide network and
Narrow analysis network has stronger relevance, therefore can mutually assist improving when the network structure as optimizing mutual
Learning performance reduces the risk of over-fitting, thus the case where being more in line with practical application, and then improve the detection essence of testing result
Degree.
Detailed description of the invention
Fig. 1 is a kind of schematic diagram of internal structure for computer equipment that one embodiment provides;
Fig. 2 is a kind of flow chart for coronary stenosis detection method that one embodiment provides;
Fig. 3 is a kind of schematic diagram for network structure that one embodiment provides;
Fig. 4 is a kind of schematic diagram for network structure that one embodiment provides;
Fig. 5 is a kind of schematic diagram for network structure that one embodiment provides;
Fig. 6 is a kind of schematic diagram for network structure that one embodiment provides;
Fig. 7 is a kind of schematic diagram for network structure that one embodiment provides;
Fig. 8 is a kind of schematic diagram for network structure that one embodiment provides;
Fig. 9 is a kind of flow chart for training method that one embodiment provides;
Figure 10 is a kind of flow chart for coronary stenosis detection method that one embodiment provides;
Figure 11 is a kind of structural schematic diagram for coronary stenosis detection device that one embodiment provides.
Specific embodiment
It is with reference to the accompanying drawings and embodiments, right in order to which the objects, technical solutions and advantages of the application are more clearly understood
The application is further elaborated.It should be appreciated that specific embodiment described herein is only to explain the application, and do not have to
In restriction the application.
Coronary stenosis detection method provided by the present application, can be applied in computer equipment as shown in Figure 1, the calculating
Machine equipment can be terminal, and internal structure chart can be as shown in Figure 1.The computer equipment includes being connected by system bus
Processor, memory, network interface, display screen and input unit.Wherein, the processor of the computer equipment is for providing calculating
And control ability.The memory of the computer equipment includes non-volatile memory medium, built-in storage.The non-volatile memories are situated between
Matter is stored with operating system and computer program.The built-in storage is operating system and computer in non-volatile memory medium
The operation of program provides environment.The network interface of the computer equipment is used to communicate with external terminal by network connection.It should
To realize a kind of coronary stenosis detection method when computer program is executed by processor.The display screen of the computer equipment can be
Liquid crystal display or electric ink display screen, the input unit of the computer equipment can be the touch covered on display screen
Layer, is also possible to the key being arranged on computer equipment shell, trace ball or Trackpad, can also be external keyboard, touch-control
Plate or mouse etc..
It will be understood by those skilled in the art that structure shown in Fig. 1, only part relevant to application scheme is tied
The block diagram of structure does not constitute the restriction for the computer equipment being applied thereon to application scheme, specific computer equipment
It may include perhaps combining certain components or with different component layouts than more or fewer components as shown in the figure.
Embodiment will be passed through below and in conjunction with attached drawing specifically to the technical side of the technical solution of the application and the application
How case, which solves above-mentioned technical problem, is described in detail.These specific embodiments can be combined with each other below, for phase
Same or similar concept or process may repeat no more in certain embodiments.
Fig. 2 is a kind of flow chart for coronary stenosis detection method that one embodiment provides.The executing subject of the present embodiment
For computer equipment as shown in Figure 1, the present embodiment is what is involved is computer equipment using coronary stenosis detection model to be checked
The detailed process of altimetric image progress stenosis detection.As shown in Fig. 2, this method comprises:
S101, image to be detected is obtained.
Wherein, image to be detected indicates the image currently detected, is a kind of figure comprising coronary artery type structure
Picture can specifically include the structure of left coronary artery blood vessel, right coronary artery blood vessel or the other tissues or organ adjacent with coronary artery.It should be to
Detection image can include but is not limited to conventional CT image, CTA image, MRI image, PET-MRI image etc., and the present embodiment is to this
Without limitation.In practical applications, computer equipment can be scanned the coronary artery structure of human body by connecting scanning device
Obtain image to be detected.Optionally, computer equipment can also directly from database or from internet downloading included
Image to be detected of coronary artery structure, with no restrictions to this present embodiment.
S102, image to be detected is input to coronary stenosis detection model, obtains testing result;Coronary stenosis detection model
Including core network, segmentation network and narrow analysis network;The output of core network respectively with segmentation network and narrow analysis net
The input of network connects;Testing result includes coronary artery segmentation result and coronary stenosis result.
Wherein, coronary stenosis detection model is used to carry out image to be detected the segmentation of coronary artery and the detection of coronary stenosis,
To obtain testing result, and the testing result can include coronary artery segmentation result and coronary stenosis result simultaneously.Above-mentioned coronary artery
Segmentation result may include the image of various description segmentation results, such as segmented image, probability respondence figure etc.;Above-mentioned coronary stenosis
Result may include various description coronary stenosis testing results or value, such as the position of the instruction information of narrow classification, narrow zone
It sets, the valuation of the size of narrow zone, stenosis etc..
Above-mentioned core network from image to be detected for extracting characteristics of image, to obtain the spy of the image to be detected
Levy image.Core network is specifically as follows a kind of convolutional neural networks, can specifically include and is not limited to V-NET, U-NET,
The convolutional neural networks such as VGG, ResNet, DenseNet, with no restrictions to this present embodiment.Core network in the present embodiment is
V-NET network.Segmentation network obtains above-mentioned coronary artery segmentation for being split processing to the characteristic image that core network exports
As a result.The coronary artery structure on characteristic image that narrow analysis network is used to export core network carries out narrow analysis, thus
To above-mentioned coronary stenosis as a result, in practical applications, which can be a kind of sorter network, it is also possible to one
Kind Recurrent networks, can also include sorter network and Recurrent networks simultaneously.
In addition, the coronary stenosis detection model in the present embodiment may include network structure as shown in Figure 3, including
Core network, segmentation network and narrow analysis network, and the output of core network respectively with segmentation network and narrow analysis network
Input connection.
In the present embodiment, segmentation and the hat to image to be detected are realized using coronary stenosis detection model as shown in Figure 3
Arteries and veins stenosis detection can be to be checked by this when detailed process includes: that the step of computer equipment is based on S101 obtains image to be detected
Altimetric image is input to core network and carries out feature extraction, obtains the characteristic image of the detection image;Computer equipment is further
This feature image is input to segmentation network carry out coronary artery structure dividing processing, obtain coronary artery segmentation result;It calculates simultaneously
Characteristic image can be input to narrow analysis network and carry out coronary stenosis analysis by machine equipment, obtain coronary stenosis result.
Optionally, the coronary stenosis detection model in the present embodiment can also include network structure as shown in Figure 4, wherein
Including core network, segmentation network and narrow analysis network, and the output of core network respectively with segmentation network and narrow analysis
The input of network connects;The output of segmentation network is connect with the input of narrow analysis network.
Specifically, being realized using coronary stenosis detection model as shown in Figure 4 narrow to the segmentation of image to be detected and coronary artery
Narrow detection can be by the mapping to be checked when detailed process includes: that the step of computer equipment is based on S101 obtains image to be detected
Feature extraction is carried out as being input to core network, obtains the characteristic image of the detection image;Computer equipment is again by this feature figure
The dividing processing that coronary artery structure is carried out as being input to segmentation network, obtains coronary artery segmentation result;Further, computer equipment can
Coronary stenosis analysis is carried out so that coronary artery segmentation result and features described above image are input to narrow analysis network, obtains coronary stenosis
As a result.Optionally, computer equipment can also first be handled coronary artery segmentation result and characteristic image, be obtained on characteristic image
Then image comprising needing to be detected coronary artery region will need the image for being detected coronary artery region to be input to narrow analysis net again
Network carries out the narrow analysis in the coronary artery region, obtains coronary stenosis result.
To sum up, coronary stenosis detection method provided by the above embodiment, is input to by the image to be detected that will acquire
Coronary stenosis detection model, obtains testing result;Coronary stenosis detection model therein includes core network, segmentation network and narrow
Narrow analysis network, the output of core network are connect with the input of segmentation network and narrow analysis network respectively;Above-mentioned testing result
Including coronary artery segmentation result and coronary stenosis result.The coronary artery is narrow it can be seen from the structure of above-mentioned coronary stenosis detection model
Narrow detection model includes segmentation network and narrow analysis network, and therefore, above-mentioned coronary stenosis detection model is a kind of multitask mould
Type can be realized simultaneously and be split to image to be detected of input, and carry out stenosis detection to image to be detected of input,
Obtain the testing result comprising at least two types results, i.e. coronary artery segmentation result and coronary stenosis result.In addition, due to segmentation
Parameter inside network and narrow analysis network share Backbone network, causes to divide network and narrow analysis network with stronger
Relevance, therefore can mutually assist improving mutual learning performance when optimizing such network structure, reduce over-fitting
Risk, thus the case where being more in line with practical application, and then improve the detection accuracy of testing result.
Present invention also provides three kinds of network structures to realize the coronary stenosis detection to image to be detected, specifically includes as follows
Content:
In the first application scenarios, on the basis of network structure shown in Fig. 3, present invention also provides a kind of network knots
Structure, as shown in figure 5, the narrow analysis network in network structure includes sorter network and Recurrent networks.It is right under the applicable cases
The coronary stenosis result answered includes narrow classification and stenosis.
Wherein, whether coronary artery region of the narrow classification for describing to include in image to be detected belongs to narrow zone, specifically
It can be indicated with text, number, letter etc..Stenosis is used to describe the narrow journey in the coronary artery region for including in image to be detected
Degree, can be a specific numerical value, or the height of a percentages, the numerical value represents the height of stenosis
It is low.The function of interpretive classification network on the basis of the network structure shown in Fig. 3 embodiment, the sorter network are used for the spy to input
Specific coronary artery region in sign image is classified, and judges whether the coronary artery region belongs to narrow zone, if so, output indicates
The classification results of "Yes", if it is not, then exporting the classification results for indicating "No".Correspondingly, Recurrent networks are used for the feature to input
Specific coronary artery region in image carries out narrow analysis, obtains the stenosis in the coronary artery region, is equivalent to estimating for stenosis
Value.
In the application scenarios, image to be detected is split or is detected using network structure shown in fig. 5, specific mistake
Journey, which includes: computer equipment, to be input to core network for obtained image to be detected and carries out feature extraction, and characteristic image is obtained;Again
This feature image is further input to the dividing processing that segmentation network carries out coronary artery structure, obtains coronary artery segmentation result;Together
When computer equipment characteristic image can be input to sorter network carry out classification processing, obtain narrow classification as a result, and
Characteristic image is input to Recurrent networks and carries out stenosis estimation, obtains the valuation of stenosis.
In second of application scenarios, on the basis of network structure shown in Fig. 4, present invention also provides a kind of network knots
Structure as shown in fig. 6, the narrow analysis network in network structure includes sorter network and Recurrent networks, and divides the output of network
The input of link sort network and Recurrent networks simultaneously.Under the applicable cases, corresponding coronary stenosis result includes narrow point
Class and stenosis.
In the application scenarios, image to be detected is split or is detected using network structure shown in fig. 6, specific mistake
Journey, which includes: computer equipment, to be input to core network for obtained image to be detected and carries out feature extraction, and characteristic image is obtained;Meter
It calculates machine equipment and this feature image is input to the dividing processing that segmentation network carries out coronary artery structure again, obtain coronary artery segmentation result;
Later, on the one hand, coronary artery segmentation result and characteristic image are input to sorter network simultaneously and carry out classification processing by computer equipment,
Obtain narrow classification as a result, on the other hand, coronary artery segmentation result and characteristic image are inputted recurrence net by computer equipment simultaneously
Network carries out stenosis estimation, obtains the valuation of stenosis.Optionally, computer equipment can also first divide above-mentioned coronary artery
As a result handled with characteristic image, obtain on characteristic image comprising need be detected region image, then will need again by
The image of detection zone is separately input into sorter network and Recurrent networks carry out the region instruction coronary artery classification and narrow journey
Degree detection, correspondence obtain narrow classification and stenosis.
By foregoing description as it can be seen that in Fig. 3-arbitrary network structure shown in fig. 6, divide network and narrow analysis network
The output feature that (or sorter network with Recurrent networks) has shared core network, illustrates coronary artery segmentation and coronary stenosis analysis is
It is mutually related, in particular, the input of the narrow analysis network of output connection of the segmentation network in Fig. 4 and Fig. 6, makes narrow analysis
The relevance of network is stronger, therefore can mutually assist improving mutual learning performance when optimizing above-mentioned network structure, reduces
The over-fitting risk being easy to produce in multitask network structure, and then improve the detection accuracy of above-mentioned network structure.Actually answering
In, optionally, coronary artery segmentation result includes probability respondence figure and coronary artery segmented image.Coronary artery segmented image therein indicates warp
Image after over-segmentation processing, comprising may include the left branch after over-segmentation in coronary artery structure, such as the image on the image
Coronary artery blood vessel and right branch coronary artery blood vessel can also include the crosspoint on the coronary artery blood vessel after over-segmentation or other keys
Point.Probability respondence figure is used to indicate the probability that each pixel on image to be detected or characteristic image is expressed as coronary artery.
In the third application scenarios, it is based on network structure shown in fig. 6, above-mentioned coronary stenosis detection model further includes pond
Change layer, as shown in Figure 7.Wherein, pond layer is used to export region of interest area image, region of interest area image according to probability respondence figure
For the characteristic image comprising whole coronary artery structures or part-crown kank structure.It is solved on the basis of the network structure shown in Fig. 6 embodiment
The function of pond layer is released, which is used to carry out region of interest area image according to characteristic image of the area-of-interest to input
It extracts, to obtain region of interest area image.
Correspondingly, the output of above-mentioned pond layer is connect with the input of sorter network and Recurrent networks respectively;In this case,
Sorter network is used to carry out classification processing to region of interest area image, obtains narrow classification;Recurrent networks are used for narrow to belonging to
Region of interest area image carry out stenosis estimation, obtain the valuation of stenosis.
Under above-mentioned applicable cases, side is split or examined to image to be detected using network structure shown in Fig. 7, specifically
Process, which includes: computer equipment, to be input to core network for obtained image to be detected and carries out feature extraction, and characteristic image is obtained;
This feature image is further input to the dividing processing that segmentation network carries out coronary artery structure, obtains probability respondence figure;So
The coronary artery probability value that computer equipment is indicated by analyzing the probability respondence figure afterwards, obtains comprising needing detected coronary artery structure
Area-of-interest, then the area-of-interest and features described above image are input in the layer of pond progress region of interest area image
Extraction, obtain comprising need be detected coronary artery structure region of interest area image.Then, computer equipment can be by the sense
Interest area image is input to the stenosis estimation that Recurrent networks carry out the region, obtains the valuation of stenosis, and will
The region of interest area image is input to the narrow classification that sorter network carries out the region, with identify the region whether belong to it is narrow,
Obtain the result of narrow classification.
The network structure that above-described embodiment is related to is realized using pond layer and is carried out to core network output characteristic image
The feature extraction of area-of-interest image block allows the narrow analysis network being attached thereto targetedly to area-of-interest figure
As the narrow analysis of coronary artery structure progress for including in block, the accurate fixed of narrow analysis is further improved.
In the 4th kind of application scenarios, it is based on network structure shown in Fig. 7, above-mentioned coronary stenosis detection model further includes volume
Lamination, as shown in Figure 8.On the basis of the network structure shown in Fig. 7 embodiment be added convolutional layer function, the convolutional layer for pair
Region of interest area image carries out depth characteristic extraction, obtains the characteristic image of the region of interest area image.Correspondingly, above-mentioned convolution
The input of layer is connect with pond layer, and the output of convolutional layer is connect with the input of sorter network and Recurrent networks respectively;In the situation
Under, sorter network is used to carry out classification processing to the characteristic image of region of interest area image, obtains narrow classification;Recurrent networks are used
In carrying out stenosis estimation to the characteristic image for belonging to narrow region of interest area image, the valuation of stenosis is obtained.
Under above-mentioned applicable cases, side is split or examined to image to be detected using network structure shown in Fig. 8, specifically
Process, which includes: computer equipment, to be input to core network for obtained image to be detected and carries out feature extraction, and characteristic image is obtained;
This feature image is further input to the dividing processing that segmentation network carries out coronary artery structure, obtains probability respondence figure;So
The coronary artery probability value that computer equipment is indicated by analyzing the probability respondence figure afterwards, obtains comprising needing detected coronary artery structure
Area-of-interest, then the area-of-interest and features described above image are input in the layer of pond and carry out region of interest area image
It extracts, obtains the region of interest area image comprising needing to be detected coronary artery structure.Then, computer equipment is by the area-of-interest
Image is input to convolutional layer and carries out depth characteristic extraction, obtains the characteristic image of the region of interest area image;Finally, computer is set
It is standby that the characteristic image of the region of interest area image is input to the estimation that Recurrent networks carry out the stenosis in the region again, it obtains
The valuation of stenosis, and the characteristic image of the region of interest area image is input to sorter network and carries out the narrow of the region
Whether classification belongs to narrow, obtains the result of narrow classification to identify the region.
The network structure that above-described embodiment is related to is realized using convolutional layer to the further of area-of-interest image block
Depth characteristic is extracted, and so that the feature extracted is more adapted to the analysis demand of narrow analysis network, is strengthened narrow analysis
The relevance of network, segmentation network and core network, to improve the detection accuracy of whole network structure.
By foregoing description it is found that coronary stenosis detection model is the network model obtained in advance by computer equipment training,
Therefore, present invention also provides a kind of method for training the coronary stenosis detection model, Fig. 9 is one kind that one embodiment provides
The flow chart of training method, the present embodiment are related to computer equipment according to multiple sample images, and with the corresponding gold of sample image
Standard picture is supervision message, to the process that initial coronary stenosis detection model is trained, as shown in figure 9, the process includes:
S201, sample image is obtained.
Wherein, sample image indicates the image currently used when needing to be trained, with described in aforementioned S101 to
The type of detection image is identical, and particular content can refer to explanation above-mentioned, does not repeat burdensome explanation herein.
S202, using the corresponding goldstandard image of sample image as supervision message, it is narrow that sample image is input to initial coronary artery
Narrow detection model is trained, and obtains coronary stenosis detection model.
Wherein, goldstandard image is the image after label, be can be in advance by computer equipment using different
The image formed after the various forms structure for including in coronary artery structure is marked in label on sample image, for example, coronary artery structure
In left branch coronary artery, right branch coronary artery, the crosspoint of coronary artery etc..It should be noted that when marking the crosspoint of coronary artery, it can
Crosspoint is first further carried out expansive working, and the crosspoint after expansion is marked with same label, if expansion
The corresponding coordinate value in crosspoint afterwards has exceeded coronary artery blood vessel location, then the corresponding pixel that will exceed is labeled as back
Scape.In addition, computer equipment can also mark in advance the boundary in coronary stenosis region on sample image, and according in blood vessel
The stenosis for the narrow zone that the cross-sectional area estimation of heart line marks, optionally, computer equipment can be combined with DSA and examine
The stenosis of the narrow zone marked on disconnected report sample estimates image.Other than above-mentioned labeling method, it can also wrap
The enterprising line flag of coronary artery blood vessel of the different sections using different labels on sample image is included, and is existed using different labels
It is marked on different crosspoints or key point on sample image, the mode about label can be true according to practical application request
Fixed, to this, the present embodiment does not limit.After completing all labels according to the method described above in computer equipment, by the figure after label
As being used as goldstandard image.
It is when computer equipment gets multiple sample images and corresponding goldstandard image, this is more in the present embodiment
A sample image is input in initial coronary stenosis detection model, exports the corresponding segmented image of sample image and/or narrow inspection
Altimetric image adjusts initial hat then according to the difference between the segmented image/of output or stenosis detection image and goldstandard image
The parameter of arteries and veins stenosis detection model, is trained, until the target loss function convergence or defeated of initial coronary stenosis detection model
Until segmented image/or stenosis detection image goldstandard image corresponding with input sample image out is almost the same, instructed
The coronary stenosis detection model perfected, to be used in the detection process described in above-mentioned Fig. 2.
It is related to the target loss function of initial coronary stenosis detection model, the target loss during above-mentioned training
Function uses the loss function of multitask, therefore the target loss function includes first-loss function, the second loss function, third
Loss function.Wherein, first-loss function can be obtained according to the output result of segmentation network;Second loss function can basis
The output result of sorter network obtains;Third loss function can be obtained according to the output result of Recurrent networks.
Target loss function in the present embodiment is first-loss function, the second loss function and third loss function
The sum of weighted accumulation.It can specifically be indicated with following relational expression (1):
Lz=Ls+ λ1*Lc+λ2*Lr (1);
Wherein, Lz indicates target loss function;Ls indicates first-loss function;Lc indicates the second loss function;Lr is indicated
Third loss function;λ1Presentation class weight;λ2It indicates to return weight.
There is also a kind of application scenarios, i.e., when in above-mentioned coronary artery segmented image including the position of key point, in computer
Equipment obtain through the foregoing embodiment include probability respondence figure and coronary artery segmented image coronary artery segmentation result, and including narrow
After the coronary stenosis result of classification and stenosis, computer equipment can also be performed following step and obtain coronary artery center line
Position and coronary artery segmentation result, as shown in Figure 10, which includes:
S301, the position of coronary artery center line in image to be detected is obtained according to probability respondence figure.
It, can be further in the probability respondence figure when computer equipment gets probability respondence figure in the present embodiment
In probability respondence curve or curved surface in middle description each cross-sectional area of blood vessel, it is highest to select probability respondence in each cross-sectional area
Point is used as control point, and each control point that is linked in sequence, and the curve connected into is the coronary artery center line of part.Later, computer is set
Standby one end of choosing on the coronary artery center line is as initial point, and point carries out surrounding half through neighbouring relations or centered on this endpoint
Line segment tracking in diameter region, to supplement or extend the end of coronary artery center line, to obtain complete coronary artery center line.Later
Computer equipment can also be smoothed obtained coronary artery center line using smoothing algorithm, to exclude exception control point.
S302, according to the position of coronary artery center line and the position of key point, obtain coronary artery segmentation result.
When computer equipment gets coronary artery center line, that is, the position of the coronary artery center line is got, then
The coronary artery center line can be carried out segment processing, specifically will by computer equipment according to the key point position got
Blood vessel indicated by line segment between two key points is as one section of coronary artery blood vessel after segmentation.When getting according to the method described above
After multistage blood vessel, that is, the coronary artery segmentation result of coronary artery structure is got.
Although should be understood that each step in Fig. 2 and Fig. 9, Figure 10 flow chart is successively shown according to the instruction of arrow
Show, but these steps are not that the inevitable sequence according to arrow instruction successively executes.Unless expressly state otherwise herein, this
There is no stringent sequences to limit for the execution of a little steps, these steps can execute in other order.Moreover, Fig. 2 and Fig. 9,
Figure 10 at least part step may include that perhaps these sub-steps of multiple stages or stage are not necessarily multiple sub-steps
Completion is executed in synchronization, but can be executed at different times, the execution in these sub-steps or stage sequence is not yet
Necessarily successively carry out.
In one embodiment, as Figure 11 shows, a kind of coronary stenosis detection device is provided, comprising: obtain 11 He of module
Detection module 12, in which:
Module 11 is obtained, for obtaining image to be detected;
Detection module 12 obtains testing result for image to be detected to be input to coronary stenosis detection model;Coronary artery is narrow
Narrow detection model includes core network, segmentation network and narrow analysis network;The output of core network respectively with segmentation network and
The input connection of narrow analysis network;Testing result includes coronary artery segmentation result and coronary stenosis result.
Specific about coronary stenosis detection limits the limit that may refer to above for a kind of coronary stenosis detection method
Fixed, details are not described herein.Modules in above-mentioned coronary stenosis detection device can fully or partially through software, hardware and its
Combination is to realize.Above-mentioned each module can be embedded in the form of hardware or independently of in the processor in computer equipment, can also be with
It is stored in the memory in computer equipment in a software form, in order to which processor calls the above modules of execution corresponding
Operation.
In one embodiment, a kind of computer equipment, including memory and processor are provided, is stored in memory
Computer program, the processor perform the steps of when executing computer program
Obtain image to be detected;
Image to be detected is input to coronary stenosis detection model, obtains testing result;Coronary stenosis detection model includes
Core network, segmentation network and narrow analysis network;The output of core network respectively with segmentation network and narrow analysis network
Input connection;Testing result includes coronary artery segmentation result and coronary stenosis result.
A kind of computer equipment provided by the above embodiment, implementing principle and technical effect and above method embodiment class
Seemingly, details are not described herein.
In one embodiment, a kind of computer readable storage medium is provided, computer program is stored thereon with, is calculated
Machine program also performs the steps of when being executed by processor
Obtain image to be detected;
Image to be detected is input to coronary stenosis detection model, obtains testing result;Coronary stenosis detection model includes
Core network, segmentation network and narrow analysis network;The output of core network respectively with segmentation network and narrow analysis network
Input connection;Testing result includes coronary artery segmentation result and coronary stenosis result.
A kind of computer readable storage medium provided by the above embodiment, implementing principle and technical effect and the above method
Embodiment is similar, and details are not described herein.
Those of ordinary skill in the art will appreciate that realizing all or part of the process in above-described embodiment method, being can be with
Relevant hardware is instructed to complete by computer program, the computer program can be stored in a non-volatile computer
In read/write memory medium, the computer program is when being executed, it may include such as the process of the embodiment of above-mentioned each method.Wherein,
To any reference of memory, storage, database or other media used in each embodiment provided herein,
Including non-volatile and/or volatile memory.Nonvolatile memory may include read-only memory (ROM), programming ROM
(PROM), electrically programmable ROM (EPROM), electrically erasable ROM (EEPROM) or flash memory.Volatile memory may include
Random access memory (RAM) or external cache.By way of illustration and not limitation, RAM is available in many forms,
Such as static state RAM (SRAM), dynamic ram (DRAM), synchronous dram (SDRAM), double data rate SDRAM (DDRSDRAM), increase
Strong type SDRAM (ESDRAM), synchronization link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM
(RDRAM), direct memory bus dynamic ram (DRDRAM) and memory bus dynamic ram (RDRAM) etc..
Each technical characteristic of above embodiments can be combined arbitrarily, for simplicity of description, not to above-described embodiment
In each technical characteristic it is all possible combination be all described, as long as however, the combination of these technical characteristics be not present lance
Shield all should be considered as described in this specification.
The embodiments described above only express several embodiments of the present invention, and the description thereof is more specific and detailed, but simultaneously
It cannot therefore be construed as limiting the scope of the patent.It should be pointed out that coming for those of ordinary skill in the art
It says, without departing from the inventive concept of the premise, various modifications and improvements can be made, these belong to protection of the invention
Range.Therefore, the scope of protection of the patent of the invention shall be subject to the appended claims.
Claims (10)
1. a kind of coronary stenosis detection method, which is characterized in that the described method includes:
Obtain image to be detected;
Described image to be detected is input to coronary stenosis detection model, obtains testing result;The coronary stenosis detection model
Including core network, segmentation network and narrow analysis network;The output of the core network respectively with the segmentation network and institute
The input connection of narrow analysis network is stated, the testing result includes coronary artery segmentation result and coronary stenosis result.
2. the method according to claim 1, wherein the narrow analysis network includes sorter network and recurrence net
Network;The coronary stenosis result includes narrow classification and stenosis.
3. method according to claim 1 or 2, which is characterized in that the coronary artery segmentation result include probability respondence figure and
Coronary artery segmented image.
4. according to the method described in claim 3, using it is characterized in that, the coronary stenosis detection model further includes pond layer
According to probability respondence figure output region of interest area image.
5. according to the method described in claim 4, it is characterized in that, the output of the pond layer respectively with the sorter network and
The input of the Recurrent networks connects;The sorter network is used to carry out classification processing to the region of interest area image, obtains
The narrow classification;The Recurrent networks are used to obtain to narrow region of interest area image progress stenosis estimation is belonged to
The stenosis.
6. the method according to claim 1, wherein the core network from described image to be detected for mentioning
Take characteristics of image.
7. according to the described in any item methods of claim 3-6, which is characterized in that the coronary artery segmented image includes key point
Position;The method also includes:
The position of coronary artery center line in described image to be detected is obtained according to the probability respondence figure;
According to the position of the position of the coronary artery center line and the key point, coronary artery segmentation result is obtained.
8. a kind of coronary stenosis detection device, which is characterized in that described device includes:
Module is obtained, for obtaining image to be detected;
Detection module obtains testing result for described image to be detected to be input to coronary stenosis detection model;The coronary artery
Stenosis detection model includes core network, segmentation network and narrow analysis network;The output of the core network respectively with it is described
Segmentation network is connected with the input of the narrow analysis network;The testing result includes coronary artery segmentation result and coronary stenosis knot
Fruit.
9. a kind of computer equipment, including memory and processor, the memory are stored with computer program, feature exists
In the step of processor realizes any one of claims 1 to 7 the method when executing the computer program.
10. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the computer program
The step of method described in any one of claims 1 to 7 is realized when being executed by processor.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910464256.0A CN110310256B (en) | 2019-05-30 | 2019-05-30 | Coronary stenosis detection method, coronary stenosis detection device, computer equipment and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910464256.0A CN110310256B (en) | 2019-05-30 | 2019-05-30 | Coronary stenosis detection method, coronary stenosis detection device, computer equipment and storage medium |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110310256A true CN110310256A (en) | 2019-10-08 |
CN110310256B CN110310256B (en) | 2021-09-21 |
Family
ID=68075665
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910464256.0A Active CN110310256B (en) | 2019-05-30 | 2019-05-30 | Coronary stenosis detection method, coronary stenosis detection device, computer equipment and storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110310256B (en) |
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110866931A (en) * | 2019-11-18 | 2020-03-06 | 东声(苏州)智能科技有限公司 | Image segmentation model training method and classification-based enhanced image segmentation method |
CN110910370A (en) * | 2019-11-21 | 2020-03-24 | 北京理工大学 | CTA image coronary stenosis detection method and device |
CN111012377A (en) * | 2019-12-06 | 2020-04-17 | 北京安德医智科技有限公司 | Echocardiogram heart parameter calculation and myocardial strain measurement method and device |
CN111292314A (en) * | 2020-03-03 | 2020-06-16 | 上海联影智能医疗科技有限公司 | Coronary artery segmentation method, coronary artery segmentation device, image processing system and storage medium |
CN111340794A (en) * | 2020-03-09 | 2020-06-26 | 中山大学 | Method and device for quantifying coronary artery stenosis |
CN111369528A (en) * | 2020-03-03 | 2020-07-03 | 重庆理工大学 | Coronary artery angiography image stenosis region marking method based on deep convolutional network |
CN111383259A (en) * | 2020-03-02 | 2020-07-07 | 联影智能医疗科技(北京)有限公司 | Image analysis method, computer device, and storage medium |
CN111445449A (en) * | 2020-03-19 | 2020-07-24 | 上海联影智能医疗科技有限公司 | Region-of-interest classification method and device, computer equipment and storage medium |
CN111667456A (en) * | 2020-04-28 | 2020-09-15 | 北京理工大学 | Method and device for detecting vascular stenosis in coronary artery X-ray sequence radiography |
CN112967284A (en) * | 2021-04-30 | 2021-06-15 | 平安科技(深圳)有限公司 | Blood vessel image segmentation analysis method, device, equipment and storage medium |
CN114757944A (en) * | 2022-06-13 | 2022-07-15 | 深圳科亚医疗科技有限公司 | Blood vessel image analysis method and device and storage medium |
CN114972221A (en) * | 2022-05-13 | 2022-08-30 | 北京医准智能科技有限公司 | Image processing method and device, electronic equipment and readable storage medium |
CN116616804A (en) * | 2023-07-25 | 2023-08-22 | 杭州脉流科技有限公司 | Method, device, equipment and storage medium for acquiring intracranial arterial stenosis evaluation parameters |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20170265754A1 (en) * | 2013-10-17 | 2017-09-21 | Siemens Healthcare Gmbh | Method and system for machine learning based assessment of fractional flow reserve |
CN108171698A (en) * | 2018-02-12 | 2018-06-15 | 数坤(北京)网络科技有限公司 | A kind of method of automatic detection human heart Coronary Calcification patch |
CN109222980A (en) * | 2018-06-19 | 2019-01-18 | 北京红云智胜科技有限公司 | Method of the measurement coronarogram based on deep learning as blood vessel diameter |
Family Cites Families (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108446662A (en) * | 2018-04-02 | 2018-08-24 | 电子科技大学 | A kind of pedestrian detection method based on semantic segmentation information |
CN109272510B (en) * | 2018-07-24 | 2021-06-04 | 清华大学 | Method for segmenting tubular structure in three-dimensional medical image |
CN109410267B (en) * | 2018-09-30 | 2020-02-07 | 数坤(北京)网络科技有限公司 | Coronary artery segmentation evaluation method and system |
CN109448005B (en) * | 2018-10-31 | 2019-12-27 | 数坤(北京)网络科技有限公司 | Network model segmentation method and equipment for coronary artery |
CN109544566B (en) * | 2018-11-29 | 2022-02-01 | 上海联影医疗科技股份有限公司 | Coronary image segmentation method, device, computer equipment and storage medium |
CN109686426A (en) * | 2018-12-29 | 2019-04-26 | 上海商汤智能科技有限公司 | Medical imaging processing method and processing device, electronic equipment and storage medium |
-
2019
- 2019-05-30 CN CN201910464256.0A patent/CN110310256B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20170265754A1 (en) * | 2013-10-17 | 2017-09-21 | Siemens Healthcare Gmbh | Method and system for machine learning based assessment of fractional flow reserve |
CN108171698A (en) * | 2018-02-12 | 2018-06-15 | 数坤(北京)网络科技有限公司 | A kind of method of automatic detection human heart Coronary Calcification patch |
CN109222980A (en) * | 2018-06-19 | 2019-01-18 | 北京红云智胜科技有限公司 | Method of the measurement coronarogram based on deep learning as blood vessel diameter |
Non-Patent Citations (4)
Title |
---|
MAJD ZREIK 等: "A Recurrent CNN for Automatic Detection and Classification of Coronary Artery Plaque and Stenosis in Coronary CT Angiography", 《IEEE TRANSACTIONS ON MEDICAL IMAGING》 * |
MAJD ZREIK 等: "Automatic Detection and Characterization of Coronary Artery Plaque and Stenosis using a Recurrent Convolutional Neural Network in Coronary CT Angiography", 《ARXIV》 * |
ROBBERT W.VAN HAMERSVELT 等: "Deep learning analysis of left ventricular myocardium in CT angiographic intermediate-degree coronary stenosis improves the diagnostic accuracy for identification of functionally significant stenosis", 《EUROPEAN RADIOLOGY》 * |
李裕: "基于CCA和CCTA图像的冠状动脉狭窄度量化方法研究", 《中国优秀硕士学位论文全文数据库 医药卫生科技辑》 * |
Cited By (24)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110866931A (en) * | 2019-11-18 | 2020-03-06 | 东声(苏州)智能科技有限公司 | Image segmentation model training method and classification-based enhanced image segmentation method |
CN110910370A (en) * | 2019-11-21 | 2020-03-24 | 北京理工大学 | CTA image coronary stenosis detection method and device |
CN110910370B (en) * | 2019-11-21 | 2023-02-28 | 北京理工大学 | CTA image coronary stenosis detection method and device |
CN111012377B (en) * | 2019-12-06 | 2020-11-03 | 北京安德医智科技有限公司 | Echocardiogram heart parameter calculation and myocardial strain measurement method and device |
CN111012377A (en) * | 2019-12-06 | 2020-04-17 | 北京安德医智科技有限公司 | Echocardiogram heart parameter calculation and myocardial strain measurement method and device |
CN111383259B (en) * | 2020-03-02 | 2023-08-22 | 联影智能医疗科技(北京)有限公司 | Image analysis method, computer device, and storage medium |
CN111383259A (en) * | 2020-03-02 | 2020-07-07 | 联影智能医疗科技(北京)有限公司 | Image analysis method, computer device, and storage medium |
CN111292314A (en) * | 2020-03-03 | 2020-06-16 | 上海联影智能医疗科技有限公司 | Coronary artery segmentation method, coronary artery segmentation device, image processing system and storage medium |
CN111292314B (en) * | 2020-03-03 | 2024-05-24 | 上海联影智能医疗科技有限公司 | Coronary artery segmentation method, device, image processing system and storage medium |
CN111369528A (en) * | 2020-03-03 | 2020-07-03 | 重庆理工大学 | Coronary artery angiography image stenosis region marking method based on deep convolutional network |
CN111369528B (en) * | 2020-03-03 | 2022-09-09 | 重庆理工大学 | Coronary artery angiography image stenosis region marking method based on deep convolutional network |
CN111340794A (en) * | 2020-03-09 | 2020-06-26 | 中山大学 | Method and device for quantifying coronary artery stenosis |
CN111340794B (en) * | 2020-03-09 | 2023-07-04 | 中山大学 | Quantification method and device for coronary artery stenosis |
CN111445449A (en) * | 2020-03-19 | 2020-07-24 | 上海联影智能医疗科技有限公司 | Region-of-interest classification method and device, computer equipment and storage medium |
CN111445449B (en) * | 2020-03-19 | 2024-03-01 | 上海联影智能医疗科技有限公司 | Method, device, computer equipment and storage medium for classifying region of interest |
CN111667456B (en) * | 2020-04-28 | 2023-11-14 | 北京理工大学 | Method and device for detecting vascular stenosis in coronary artery X-ray sequence radiography |
CN111667456A (en) * | 2020-04-28 | 2020-09-15 | 北京理工大学 | Method and device for detecting vascular stenosis in coronary artery X-ray sequence radiography |
CN112967284A (en) * | 2021-04-30 | 2021-06-15 | 平安科技(深圳)有限公司 | Blood vessel image segmentation analysis method, device, equipment and storage medium |
CN114972221B (en) * | 2022-05-13 | 2022-12-23 | 北京医准智能科技有限公司 | Image processing method and device, electronic equipment and readable storage medium |
CN114972221A (en) * | 2022-05-13 | 2022-08-30 | 北京医准智能科技有限公司 | Image processing method and device, electronic equipment and readable storage medium |
CN114757944A (en) * | 2022-06-13 | 2022-07-15 | 深圳科亚医疗科技有限公司 | Blood vessel image analysis method and device and storage medium |
CN114757944B (en) * | 2022-06-13 | 2022-08-16 | 深圳科亚医疗科技有限公司 | Blood vessel image analysis method and device and storage medium |
CN116616804B (en) * | 2023-07-25 | 2023-10-13 | 杭州脉流科技有限公司 | Method, device, equipment and storage medium for acquiring intracranial arterial stenosis evaluation parameters |
CN116616804A (en) * | 2023-07-25 | 2023-08-22 | 杭州脉流科技有限公司 | Method, device, equipment and storage medium for acquiring intracranial arterial stenosis evaluation parameters |
Also Published As
Publication number | Publication date |
---|---|
CN110310256B (en) | 2021-09-21 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110310256A (en) | Coronary stenosis detection method, device, computer equipment and storage medium | |
CN110148192B (en) | Medical image imaging method, device, computer equipment and storage medium | |
Slomka et al. | Cardiac imaging: working towards fully-automated machine analysis & interpretation | |
CN111292314B (en) | Coronary artery segmentation method, device, image processing system and storage medium | |
CN110766730B (en) | Image registration and follow-up evaluation method, storage medium and computer equipment | |
CN111080573B (en) | Rib image detection method, computer device and storage medium | |
Wang et al. | Uncertainty-guided efficient interactive refinement of fetal brain segmentation from stacks of MRI slices | |
CN110189306B (en) | Method and device for determining responsible blood vessels corresponding to abnormal regions in brain image | |
US20230386048A1 (en) | Ultrasound image segmentation method and apparatus, terminal device, and storage medium | |
CN109508787A (en) | Neural network model training method and system for ultrasound displacement estimation | |
CN108846829A (en) | Diseased region recognition methods and device, computer installation and readable storage medium storing program for executing | |
US20220366679A1 (en) | Methods and systems for training learning network for medical image analysis | |
CN108171702A (en) | Vulnerable plaque recognition methods, application server and computer readable storage medium | |
CN110189302A (en) | Brain image analysis method, computer equipment and readable storage medium storing program for executing | |
CN113935943A (en) | Method, device, computer equipment and storage medium for intracranial aneurysm identification detection | |
CN110751187A (en) | Training method of abnormal area image generation network and related product | |
CN114693671B (en) | Lung nodule semi-automatic segmentation method, device, equipment and medium based on deep learning | |
CN113034507A (en) | CCTA image-based coronary artery three-dimensional segmentation method | |
CN111128348B (en) | Medical image processing method, medical image processing device, storage medium and computer equipment | |
Romaszko et al. | Neural network-based left ventricle geometry prediction from CMR images with application in biomechanics | |
CN115409879A (en) | Data processing method and device for image registration, storage medium and electronic equipment | |
CN114299057A (en) | Method for extracting blood vessel center line and storage medium | |
CN116681716B (en) | Method, device, equipment and storage medium for dividing intracranial vascular region of interest | |
CN111724371B (en) | Data processing method and device and electronic equipment | |
CN113962990B (en) | Chest CT image recognition method and device, computer equipment and storage 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 |