CN110070529A - A kind of Endovascular image division method, system and electronic equipment - Google Patents
A kind of Endovascular image division method, system and electronic equipment Download PDFInfo
- Publication number
- CN110070529A CN110070529A CN201910318325.7A CN201910318325A CN110070529A CN 110070529 A CN110070529 A CN 110070529A CN 201910318325 A CN201910318325 A CN 201910318325A CN 110070529 A CN110070529 A CN 110070529A
- Authority
- CN
- China
- Prior art keywords
- image
- endovascular
- primitive medicine
- polar coordinates
- network
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
-
- 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)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiology & Medical Imaging (AREA)
- Quality & Reliability (AREA)
- Geometry (AREA)
- Apparatus For Radiation Diagnosis (AREA)
- Image Analysis (AREA)
Abstract
This application involves a kind of Endovascular image division method, system and electronic equipments.The described method includes: step a: carrying out Polar coordinates processing to primitive medicine image;Step b: inputting deep-neural-network for the Polar coordinates treated image, obtains in image distance of multiple chamber films to center on every ray;Step c: according to the distance of chamber films multiple on every ray to center, the corresponding coordinate of all the points in the primitive medicine image before Polar coordinates are handled is calculated;Step d: primitive medicine Image Segmentation result is calculated according to the coordinate of corresponding points in primitive medicine image.The Endovascular image division method, system and electronic equipment of the embodiment of the present application are by combining image procossing and deep neural network to carry out the automatic segmentation of Endovascular image, error can be greatly reduced, be integrally improved the precision and versatility of Endovascular Image Segmentation.
Description
Technical field
The application belongs to medical image processing technical field, in particular to a kind of Endovascular image division method, system
And electronic equipment.
Background technique
Blood vessel segmentation is the basic and important problem in blood vessel image analysis.Accurately segmentation can be provided to blood vessel structure
The quantitative description of different characteristic, can also observe the Geometrical change for comparing blood vessel, provide important evidence and ginseng for diagnosing and treating
It examines.The result of extraction can be used in other vessels analysis tasks, such as the matching of blood vessel, three-dimensional reconstruction, estimation.Blood vessel
Image is two dimension expression of the three-dimensional blood vessel structure under projection condition, due to by imaging noise, complicated blood vessel structure and at
As the image of the factors such as angle and distance, blood vessel image is usually expressed as low contrast, obscured vessel structure vulnerable to noise and non-
Blood vessel structure interferes and different forms is presented.
The development of blood vessel image cutting techniques is one and is gradually sent out to automatic segmentation again from being manually divided into Interactive Segmentation
The process of exhibition.Automatic division method, which refers to, pursues the segmentation task for independently completing target by computer completely without artificial
Intervene.But current automatic segmentation can't independently complete the segmentation task of target, and accuracy need to be improved.
Summary of the invention
This application provides a kind of Endovascular image division method, system and electronic equipments, it is intended at least in certain journey
One of above-mentioned technical problem in the prior art is solved on degree.
To solve the above-mentioned problems, this application provides following technical solutions:
A kind of Endovascular image division method, comprising:
Step a: Polar coordinates processing is carried out to primitive medicine image;
Step b: the Polar coordinates treated image is inputted into deep-neural-network, is obtained in image on every ray
Distance of multiple chamber films to center;
Step c: according to the distance of every ray epicoele film to center, the original doctor before Polar coordinates processing is calculated
Learn the corresponding coordinate of all the points in image;
Step d: primitive medicine Image Segmentation result is calculated according to the coordinate of corresponding points in primitive medicine image.
The technical solution that the embodiment of the present application is taken further include: described that primitive medicine image is carried out in the step a
Polar coordinates processing specifically: perpendicular to center line, virtual straight line is radiated to all angles, and interpolation calculation obtains on ray
The corresponding mark value of sum of the grayscale values of each point.
The technical solution that the embodiment of the present application is taken further include: the step a further include: to Polar coordinates treated shadow
As carrying out flattening processing;The flattening processing specifically: all rays are successively arranged vertically, and calculate lumen of vessels film to center
Distance.
The technical solution that the embodiment of the present application is taken further include: in the step b, the training of the deep-neural-network
Mode specifically includes:
Step b1: flattening treated training sample is inputted into deep-neural-network, obtains range prediction result;
Step b2: range prediction result and artificial mark are compared, and obtain range prediction error, and by range prediction
Error feeds back to deep-neural-network;
Step b3: according to range prediction error update deep-neural-network and executing iterative operation, until range prediction misses
Difference is no longer obviously reduced or iterative operation reaches the number of setting, obtains trained deep-neural-network.
The technical solution that the embodiment of the present application is taken further include: described according in primitive medicine image in the step d
The coordinate of corresponding points calculates primitive medicine Image Segmentation result and specifically includes:
Step d1: according to the coordinate of all the points in primitive medicine image, the contour curve of interpolation calculation chamber film;
Step d2: according to contour curve, calculating all connected regions, and by the region where primitive medicine image center point
As Endovascular Image Segmentation result;
Step d3: physical area, plaque load, the lumen minimum for calculating chamber film according to Endovascular Image Segmentation result are straight
Diameter, lumen maximum gauge, patch minimum thickness, Angiogenesis, calcification radian, positivity reconstruct index parameters.
A kind of another technical solution that the embodiment of the present application is taken are as follows: Endovascular image dividing system, comprising:
Polar coordinates processing module: for carrying out Polar coordinates processing to primitive medicine image;
Distance calculation module: for the Polar coordinates treated image to be inputted deep-neural-network, image is obtained
In distance of multiple chamber films to center on every ray;
Point coordinate calculation module: it for the distance according to every ray epicoele film to center, calculates at Polar coordinates
The corresponding coordinate of all the points in primitive medicine image before reason;
Image Segmentation module: for calculating primitive medicine Image Segmentation knot according to the coordinate of corresponding points in primitive medicine image
Fruit.
The technical solution that the embodiment of the present application is taken further include: the Polar coordinates processing module to primitive medicine image into
Row Polar coordinates processing specifically: perpendicular to center line, virtual straight line is radiated to all angles, and interpolation calculation obtains ray
The corresponding mark value of sum of the grayscale values of upper each point.
The technical solution that the embodiment of the present application is taken further includes image Flattening Module, and described image Flattening Module is used for pole
Treated that image carries out flattening processing for coordinatograph;The flattening processing specifically: all rays are successively arranged vertically, and are counted
Distance of the calculation lumen of vessels film to center.
The technical solution that the embodiment of the present application is taken further include: the training method of the deep-neural-network specifically: will
Flattening treated training sample inputs deep-neural-network, obtains range prediction result;By range prediction result and artificial mark
Note is compared, and obtains range prediction error, and range prediction error is fed back to deep-neural-network;It is missed according to range prediction
Difference updates deep-neural-network and simultaneously executes iterative operation, until range prediction error is no longer obviously reduced or iterative operation reaches
The number of setting obtains trained deep-neural-network.
The technical solution that the embodiment of the present application is taken further include: the Image Segmentation module is according to right in primitive medicine image
The coordinate that should be put calculates primitive medicine Image Segmentation result specifically:
1, according to the coordinate of all the points in primitive medicine image, the contour curve of interpolation calculation chamber film;
2, all connected regions are calculated according to contour curve, and using the region where primitive medicine image center point as
Endovascular Image Segmentation result;
3, physical area, the plaque load, lumen minimum diameter, pipe of chamber film are calculated according to Endovascular Image Segmentation result
Chamber maximum gauge, patch minimum thickness, Angiogenesis, calcification radian, positivity reconstruct index parameters.
The another technical solution that the embodiment of the present application is taken are as follows: a kind of electronic equipment, comprising:
At least one processor;And
The memory being connect at least one described processor communication;Wherein,
The memory is stored with the instruction that can be executed by one processor, and described instruction is by described at least one
Device is managed to execute, so that at least one described processor is able to carry out the following operation of above-mentioned Endovascular image division method:
Step a: Polar coordinates processing is carried out to primitive medicine image;
Step b: the Polar coordinates treated image is inputted into deep-neural-network, is obtained in image on every ray
Distance of multiple chamber films to center;
Step c: according to the distance of every ray epicoele film to center, the original doctor before Polar coordinates processing is calculated
Learn the corresponding coordinate of all the points in image;
Step d: primitive medicine Image Segmentation result is calculated according to the coordinate of corresponding points in primitive medicine image.
Compared with the existing technology, the beneficial effect that the embodiment of the present application generates is: the Endovascular of the embodiment of the present application
Image division method, system and electronic equipment are by combining oneself of image procossing and deep neural network progress Endovascular image
Dynamic segmentation, compared with the existing technology, can be greatly reduced error, be integrally improved the precision of Endovascular Image Segmentation and general
Property.
Detailed description of the invention
Fig. 1 is the flow chart of the Endovascular image division method of the embodiment of the present application;
Fig. 2 is the structural schematic diagram of the Endovascular image dividing system of the embodiment of the present application;
Fig. 3 is the hardware device structural schematic diagram of Endovascular image division method provided by the embodiments of the present application.
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, not
For limiting the application.
Referring to Fig. 1, being the flow chart of the Endovascular image division method of the embodiment of the present application.The embodiment of the present application
Endovascular image division method the following steps are included:
Step 100: obtaining primitive medicine image;
In step 100, primitive medicine image include but is not limited to IVUS (intravenous ultrasound, it is intravascular
Ultrasound) image or OCT (Optical CoherenceTomography, optical coherence tomography) image etc..
Step 200: Polar coordinates processing is carried out to primitive medicine image, specifically: perpendicular to center line, to all angles
Virtual straight line is radiated, and interpolation calculation obtains the corresponding mark value of sum of the grayscale values of each point on ray;
Step 300: to Polar coordinates, treated that image carries out flattening processing, specifically: all rays are successively vertical
Arrangement, and calculate lumen of vessels film to center distance;
Step 400: flattening treated image being inputted into trained deep-neural-network, is obtained multiple on every ray
Distance of the chamber film to center;
In step 400, input deep-neural-network can be 1 image, be also possible to the group of the adjacent N image in front and back
It closes.Multiple chamber films include but are not limited to lumen and outer elastic membrane.Deep-neural-network is trained by using great amount of samples data,
By thousands of iteration processes, so that the range prediction result of deep-neural-network output and artificial mark connect
Closely, in the training process, i-th of profile and corresponding i-th of artificial mark are compared.Specifically, deep-neural-network
Training method the following steps are included:
Step 401: flattening treated training sample being inputted into deep-neural-network, obtains range prediction result;
Step 402: range prediction result and artificial mark being compared, obtain range prediction error, and will be apart from pre-
It surveys error and feeds back to deep-neural-network;
Step 403: according to range prediction error update deep-neural-network and iterative operation is executed, until range prediction misses
Difference is no longer obviously reduced or iterative operation reaches the number of setting, obtains trained deep-neural-network.
Step 500: according to the distance of chamber films multiple on every ray to center, calculating the primitive medicine before Polar coordinates
The corresponding coordinate of all the points in image;
Step 600: primitive medicine Image Segmentation result is calculated according to the coordinate of corresponding points in primitive medicine image;
In step 600, Endovascular Image Segmentation result calculation specifically includes the following steps:
Step 601: according to the coordinate of all the points in primitive medicine image, the contour curve of the multiple chamber films of interpolation calculation;
Step 602: according to contour curve, calculating all connected regions, and by the area where primitive medicine image center point
Domain is as Endovascular Image Segmentation result;
Step 603: the physical area S of chamber film is calculated according to Endovascular Image Segmentation resultphy=Spixel*spacing、
Plaque load plaque burden=1-lumen area/eem area*100%, lumen minimum diameter, lumen maximum gauge,
The parameters such as patch minimum thickness, Angiogenesis, calcification radian, positivity reconstruct index.
Referring to Fig. 2, being the structural schematic diagram of the Endovascular image dividing system of the embodiment of the present application.The application is implemented
The Endovascular image dividing system of example includes image acquiring module, Polar coordinates processing module, image Flattening Module, distance meter
Calculate module, point coordinate calculation module and Image Segmentation module.
Image acquiring module: for obtaining primitive medicine image;Wherein, primitive medicine image includes but is not limited to IVUS
(intravenous ultrasound, intravascular ultrasound) image or OCT (Optical CoherenceTomography, optics
Coherence tomography) image etc..
Polar coordinates processing module: for carrying out Polar coordinates processing to primitive medicine image, specifically: perpendicular to center
Line radiates virtual straight line to all angles, and interpolation calculation obtains the corresponding mark value of sum of the grayscale values of each point on ray;
Image Flattening Module: the image for Polar coordinates processing carries out flattening processing, specifically: successively by all rays
Vertical arrangement, and calculate lumen of vessels film to center distance;
Distance calculation module: for the trained deep-neural-network of image input that will flatten that treated, every is obtained
Distance of multiple chamber films to center on ray;In the embodiment of the present application, input deep-neural-network can be 1 image,
It can be the combination of the adjacent N image in front and back.Multiple chamber films include but are not limited to lumen and outer elastic membrane.By using a large amount of
Sample data trains deep-neural-network, by thousands of iteration processes, so that deep-neural-network exported
Range prediction result and artificial mark are close, and in the training process, i-th of profile and corresponding i-th of artificial mark are carried out
Comparison.Specifically, the training method of deep-neural-network includes: will flattening treated training sample input deep layer nerve net
Network obtains range prediction result;Range prediction result and artificial mark are compared, obtain range prediction error, and will be away from
Deep-neural-network is fed back to from prediction error;According to range prediction error update deep-neural-network and iterative operation is executed,
Until range prediction error is no longer obviously reduced or iterative operation reaches the number of setting, trained deep layer nerve net is obtained
Network.
Point coordinate calculation module: for the distance according to chamber films multiple on every ray to center, calculating Polar coordinates it
The corresponding coordinate of all the points in preceding primitive medicine image;
Image Segmentation module: for calculating primitive medicine Image Segmentation knot according to the coordinate of corresponding points in primitive medicine image
Fruit;Wherein, Endovascular Image Segmentation result calculation includes:
1, according to the coordinate of all the points in primitive medicine image, the contour curve of the multiple chamber films of interpolation calculation;
2, all connected regions are calculated according to contour curve, and using the region where primitive medicine image center point as
Endovascular Image Segmentation result;
3, the physical area S of chamber film is calculated according to Endovascular Image Segmentation resultphy=Spixel* spacing, patch are negative
Lotus plaque burden=1-lumen area/eem area*100%, lumen minimum diameter, lumen maximum gauge, patch are most
The parameters such as small thickness, Angiogenesis, calcification radian, positivity reconstruct index.
Fig. 3 is the hardware device structural schematic diagram of Endovascular image division method provided by the embodiments of the present application.Such as Fig. 3
Shown, which includes one or more processors and memory.It takes a processor as an example, which can also include:
Input system and output system.
Processor, memory, input system and output system can be connected by bus or other modes, in Fig. 3 with
For being connected by bus.
Memory as a kind of non-transient computer readable storage medium, can be used for storing non-transient software program, it is non-temporarily
State computer executable program and module.Processor passes through operation non-transient software program stored in memory, instruction
And module realizes the place of above method embodiment thereby executing the various function application and data processing of electronic equipment
Reason method.
Memory may include storing program area and storage data area, wherein storing program area can storage program area, extremely
Application program required for a few function;It storage data area can storing data etc..In addition, memory may include that high speed is random
Memory is accessed, can also include non-transient memory, a for example, at least disk memory, flush memory device or other are non-
Transient state solid-state memory.In some embodiments, it includes the memory remotely located relative to processor that memory is optional, this
A little remote memories can pass through network connection to processing system.The example of above-mentioned network includes but is not limited to internet, enterprise
Intranet, local area network, mobile radio communication and combinations thereof.
Input system can receive the number or character information of input, and generate signal input.Output system may include showing
Display screen etc. shows equipment.
One or more of module storages in the memory, are executed when by one or more of processors
When, execute the following operation of any of the above-described embodiment of the method:
Step a: Polar coordinates processing is carried out to primitive medicine image;
Step b: the Polar coordinates treated image is inputted into deep-neural-network, is obtained in image on every ray
Distance of multiple chamber films to center;
Step c: according to the distance of every ray epicoele film to center, the original doctor before Polar coordinates processing is calculated
Learn the corresponding coordinate of all the points in image;
Step d: primitive medicine Image Segmentation result is calculated according to the coordinate of corresponding points in primitive medicine image.
Method provided by the embodiment of the present application can be performed in the said goods, has the corresponding functional module of execution method and has
Beneficial effect.The not technical detail of detailed description in the present embodiment, reference can be made to method provided by the embodiments of the present application.
The embodiment of the present application provides a kind of non-transient (non-volatile) computer storage medium, and the computer storage is situated between
Matter is stored with computer executable instructions, the executable following operation of the computer executable instructions:
Step a: Polar coordinates processing is carried out to primitive medicine image;
Step b: the Polar coordinates treated image is inputted into deep-neural-network, is obtained in image on every ray
Distance of multiple chamber films to center;
Step c: according to the distance of every ray epicoele film to center, the original doctor before Polar coordinates processing is calculated
Learn the corresponding coordinate of all the points in image;
Step d: primitive medicine Image Segmentation result is calculated according to the coordinate of corresponding points in primitive medicine image.
The embodiment of the present application provides a kind of computer program product, and the computer program product is non-temporary including being stored in
Computer program on state computer readable storage medium, the computer program include program instruction, when described program instructs
When being computer-executed, the computer is made to execute following operation:
Step a: Polar coordinates processing is carried out to primitive medicine image;
Step b: the Polar coordinates treated image is inputted into deep-neural-network, is obtained in image on every ray
Distance of multiple chamber films to center;
Step c: according to the distance of every ray epicoele film to center, the original doctor before Polar coordinates processing is calculated
Learn the corresponding coordinate of all the points in image;
Step d: primitive medicine Image Segmentation result is calculated according to the coordinate of corresponding points in primitive medicine image.
The Endovascular image division method, system and electronic equipment of the embodiment of the present application are by combining image procossing and depth
The automatic segmentation that neural network carries out Endovascular image is spent, compared with the existing technology, error can be greatly reduced, it is whole to improve
The precision and versatility of Endovascular Image Segmentation.
The foregoing description of the disclosed embodiments makes professional and technical personnel in the field can be realized or use the application.
Various modifications to these embodiments will be readily apparent to those skilled in the art, defined herein
General Principle can realize in other embodiments without departing from the spirit or scope of the application.Therefore, this Shen
These embodiments shown in the application please be not intended to be limited to, and are to fit to special with principle disclosed in the present application and novelty
The consistent widest scope of point.
Claims (11)
1. a kind of Endovascular image division method characterized by comprising
Step a: Polar coordinates processing is carried out to primitive medicine image;
Step b: the Polar coordinates treated image is inputted into deep-neural-network, is obtained multiple on every ray in image
Distance of the chamber film to center;
Step c: according to the distance of chamber films multiple on every ray to center, the original doctor before Polar coordinates processing is calculated
Learn the corresponding coordinate of all the points in image;
Step d: primitive medicine Image Segmentation result is calculated according to the coordinate of corresponding points in primitive medicine image.
2. Endovascular image division method according to claim 1, which is characterized in that described right in the step a
Primitive medicine image carries out Polar coordinates processing specifically: perpendicular to center line, virtual straight line is radiated to all angles, and is inserted
The corresponding mark value of sum of the grayscale values of each point on ray is calculated in value.
3. Endovascular image division method according to claim 2, which is characterized in that the step a further include: to pole
Treated that image carries out flattening processing for coordinatograph;The flattening processing specifically: all rays are successively arranged vertically, and are counted
Distance of the calculation lumen of vessels film to center.
4. Endovascular image division method according to any one of claims 1 to 3, which is characterized in that in the step b
In, the training method of the deep-neural-network specifically includes:
Step b1: flattening treated training sample is inputted into deep-neural-network, obtains range prediction result;
Step b2: range prediction result and artificial mark are compared, and obtain range prediction error, and by range prediction error
Feed back to deep-neural-network;
Step b3: according to range prediction error update deep-neural-network and executing iterative operation, until range prediction error not
It is obviously reduced again or iterative operation reaches the number of setting, obtain trained deep-neural-network.
5. Endovascular image division method according to claim 4, which is characterized in that in the step d, described
Primitive medicine Image Segmentation result is calculated according to the coordinate of corresponding points in primitive medicine image to specifically include:
Step d1: according to the coordinate of all the points in primitive medicine image, the contour curve of interpolation calculation chamber film;
Step d2: according to contour curve, calculating all connected regions, and using the region where primitive medicine image center point as
Endovascular Image Segmentation result;
Step d3: according to Endovascular Image Segmentation result calculate the physical area of chamber film, plaque load, lumen minimum diameter,
Lumen maximum gauge, patch minimum thickness, Angiogenesis, calcification radian, positivity reconstruct index parameters.
6. a kind of Endovascular image dividing system characterized by comprising
Polar coordinates processing module: for carrying out Polar coordinates processing to primitive medicine image;
Distance calculation module: it for the Polar coordinates treated image to be inputted deep-neural-network, obtains every in image
The distance of ray epicoele film to center;
Point coordinate calculation module: it for the distance according to chamber films multiple on every ray to center, calculates at Polar coordinates
The corresponding coordinate of all the points in primitive medicine image before reason;
Image Segmentation module: for calculating primitive medicine Image Segmentation result according to the coordinate of corresponding points in primitive medicine image.
7. Endovascular image dividing system according to claim 6, which is characterized in that the Polar coordinates processing module
Polar coordinates processing is carried out to primitive medicine image specifically: perpendicular to center line, virtual straight line is radiated to all angles, and
Interpolation calculation obtains the corresponding mark value of sum of the grayscale values of each point on ray.
8. Endovascular image dividing system according to claim 7, which is characterized in that it further include image Flattening Module,
Described image Flattening Module is used for that treated that image carries out flattening processing to Polar coordinates;The flattening processing specifically: will
All rays are successively arranged vertically, and calculate lumen of vessels film to center distance.
9. according to the described in any item Endovascular image dividing systems of claim 6 to 8, which is characterized in that the deep layer mind
Training method through network specifically: flattening treated training sample is inputted into deep-neural-network, obtains range prediction knot
Fruit;Range prediction result and artificial mark are compared, obtain range prediction error, and range prediction error is fed back into depth
Layer neural network;According to range prediction error update deep-neural-network and execute iterative operation, until range prediction error not
It is obviously reduced again or iterative operation reaches the number of setting, obtain trained deep-neural-network.
10. Endovascular image dividing system according to claim 9, which is characterized in that the Image Segmentation module root
Primitive medicine Image Segmentation result is calculated according to the coordinate of corresponding points in primitive medicine image specifically:
1, according to the coordinate of all the points in primitive medicine image, the contour curve of interpolation calculation chamber film;
2, according to contour curve, all connected regions are calculated, and using the region where primitive medicine image center point as blood vessel
Intracavitary Image Segmentation result;
3, most according to the physical area of Endovascular Image Segmentation result calculating chamber film, plaque load, lumen minimum diameter, lumen
Major diameter, patch minimum thickness, Angiogenesis, calcification radian, positivity reconstruct index parameters.
11. a kind of electronic equipment, comprising:
At least one processor;And
The memory being connect at least one described processor communication;Wherein,
The memory is stored with the instruction that can be executed by one processor, and described instruction is by least one described processor
It executes, so that at least one described processor is able to carry out above-mentioned 1 to 5 described in any item Endovascular image division methods
It operates below:
Step a: Polar coordinates processing is carried out to primitive medicine image;
Step b: the Polar coordinates treated image is inputted into deep-neural-network, is obtained multiple on every ray in image
Distance of the chamber film to center;
Step c: according to the distance of every ray epicoele film to center, the primitive medicine shadow before Polar coordinates processing is calculated
The corresponding coordinate of all the points as in;
Step d: primitive medicine Image Segmentation result is calculated according to the coordinate of corresponding points in primitive medicine image.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910318325.7A CN110070529A (en) | 2019-04-19 | 2019-04-19 | A kind of Endovascular image division method, system and electronic equipment |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910318325.7A CN110070529A (en) | 2019-04-19 | 2019-04-19 | A kind of Endovascular image division method, system and electronic equipment |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110070529A true CN110070529A (en) | 2019-07-30 |
Family
ID=67368043
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910318325.7A Pending CN110070529A (en) | 2019-04-19 | 2019-04-19 | A kind of Endovascular image division method, system and electronic equipment |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110070529A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111920388A (en) * | 2020-08-25 | 2020-11-13 | 南京沃福曼医疗科技有限公司 | Method for evaluating plaque load based on OCT (optical coherence tomography) image |
CN112070778A (en) * | 2020-08-25 | 2020-12-11 | 南京沃福曼医疗科技有限公司 | Multi-parameter extraction method based on intravascular OCT and ultrasound image fusion |
CN112330701A (en) * | 2020-11-26 | 2021-02-05 | 山东师范大学 | Tissue pathology image cell nucleus segmentation method and system based on polar coordinate representation |
WO2021062006A1 (en) * | 2019-09-26 | 2021-04-01 | Boston Scientific Scimed, Inc. | Intravascular ultrasound imaging and calcium detection methods |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104966288A (en) * | 2015-06-11 | 2015-10-07 | 南京沃福曼医疗科技有限公司 | Method for automatic searching of inner wall of blood vessel based on image processing |
WO2016094909A1 (en) * | 2014-12-12 | 2016-06-16 | Lightlab Imaging, Inc. | Systems and methods to detect and display endovascular features |
CN107909585A (en) * | 2017-11-14 | 2018-04-13 | 华南理工大学 | Inner membrance dividing method in a kind of blood vessel of intravascular ultrasound image |
CN108492272A (en) * | 2018-03-26 | 2018-09-04 | 西安交通大学 | Cardiovascular vulnerable plaque recognition methods based on attention model and multitask neural network and system |
-
2019
- 2019-04-19 CN CN201910318325.7A patent/CN110070529A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2016094909A1 (en) * | 2014-12-12 | 2016-06-16 | Lightlab Imaging, Inc. | Systems and methods to detect and display endovascular features |
CN104966288A (en) * | 2015-06-11 | 2015-10-07 | 南京沃福曼医疗科技有限公司 | Method for automatic searching of inner wall of blood vessel based on image processing |
CN107909585A (en) * | 2017-11-14 | 2018-04-13 | 华南理工大学 | Inner membrance dividing method in a kind of blood vessel of intravascular ultrasound image |
CN108492272A (en) * | 2018-03-26 | 2018-09-04 | 西安交通大学 | Cardiovascular vulnerable plaque recognition methods based on attention model and multitask neural network and system |
Non-Patent Citations (1)
Title |
---|
吴宇鹏: "《基于深度学习网络的血管内超声影像斑块识别与评估方法研究》", 《中国优秀硕士学位论文全文数据库(医药卫生科技辑)》 * |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2021062006A1 (en) * | 2019-09-26 | 2021-04-01 | Boston Scientific Scimed, Inc. | Intravascular ultrasound imaging and calcium detection methods |
JP2022549877A (en) * | 2019-09-26 | 2022-11-29 | ボストン サイエンティフィック サイムド,インコーポレイテッド | Intravascular ultrasound imaging and calcium detection method |
JP7369865B2 (en) | 2019-09-26 | 2023-10-26 | ボストン サイエンティフィック サイムド,インコーポレイテッド | System for processing intravascular ultrasound images |
CN111920388A (en) * | 2020-08-25 | 2020-11-13 | 南京沃福曼医疗科技有限公司 | Method for evaluating plaque load based on OCT (optical coherence tomography) image |
CN112070778A (en) * | 2020-08-25 | 2020-12-11 | 南京沃福曼医疗科技有限公司 | Multi-parameter extraction method based on intravascular OCT and ultrasound image fusion |
CN112330701A (en) * | 2020-11-26 | 2021-02-05 | 山东师范大学 | Tissue pathology image cell nucleus segmentation method and system based on polar coordinate representation |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP6878509B2 (en) | Systems and methods for image reconstruction and correction using data and models | |
AU2019238323B2 (en) | Deep encoder-decoder models for reconstructing biomedical images | |
CN110070529A (en) | A kind of Endovascular image division method, system and electronic equipment | |
US9754390B2 (en) | Reconstruction of time-varying data | |
JP2021521982A (en) | Systems and methods for feature extraction and classification on ultrasound tomographic images | |
US11642171B2 (en) | Systems and methods for an interactive tool for determining and visualizing a functional relationship between a vascular network and perfused tissue | |
JP2021521993A (en) | Image enhancement using a hostile generation network | |
AU2015221826B2 (en) | System and method for auto-contouring in adaptive radiotherapy | |
US11127138B2 (en) | Automatic detection and quantification of the aorta from medical images | |
CN112424835A (en) | System and method for image reconstruction | |
CN111145160B (en) | Method, device, server and medium for determining coronary artery branches where calcified regions are located | |
US20210386389A1 (en) | Deep spectral bolus tracking | |
US11410308B2 (en) | 3D vessel centerline reconstruction from 2D medical images | |
CN109658470B (en) | System and method for generating images | |
JP2024515635A (en) | System and method for reconstructing 3D images from ultrasound and camera images - Patents.com | |
IL262027A (en) | System and methods for diagnostic image analysis and image quality assessment | |
CN113888566A (en) | Target contour curve determining method and device, electronic equipment and storage medium | |
Xiao et al. | Automatic vasculature identification in coronary angiograms by adaptive geometrical tracking | |
US20230060613A1 (en) | Blood parameter assessment systems and methods | |
WO2022143835A1 (en) | Systems and methods for image processing | |
US11406343B2 (en) | CT scan results | |
US20220215601A1 (en) | Image Reconstruction by Modeling Image Formation as One or More Neural Networks | |
US11151732B2 (en) | Motion correction of angiography images for 3D reconstruction of coronary arteries | |
CN112419462A (en) | Rendering synthesis method, system and storage medium for three-dimensional blood vessel | |
CN110136804A (en) | Myocardial mass calculation method and system and electronic equipment |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190730 |
|
RJ01 | Rejection of invention patent application after publication |