CN109741395A - Biventricular quantization method, device, electronic equipment and storage medium - Google Patents
Biventricular quantization method, device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN109741395A CN109741395A CN201811534455.6A CN201811534455A CN109741395A CN 109741395 A CN109741395 A CN 109741395A CN 201811534455 A CN201811534455 A CN 201811534455A CN 109741395 A CN109741395 A CN 109741395A
- Authority
- CN
- China
- Prior art keywords
- target
- image
- preset
- heart
- calculated
- 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
- G06T7/60—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
- 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
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/194—Segmentation; Edge detection involving foreground-background 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
-
- 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
- G06T7/66—Analysis of geometric attributes of image moments or centre of gravity
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10072—Tomographic images
- G06T2207/10088—Magnetic resonance imaging [MRI]
-
- 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
- G06T2207/30048—Heart; Cardiac
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Geometry (AREA)
- Medical Informatics (AREA)
- Quality & Reliability (AREA)
- Radiology & Medical Imaging (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Magnetic Resonance Imaging Apparatus (AREA)
Abstract
The embodiment of the present application provides a kind of biventricular quantization method, device, electronic equipment and storage medium, this method comprises: obtaining target original image, the target original image is the nuclear magnetic resonance image of heart；The target original image is split using preset segmentation network, obtains target multichannel probabilistic image；The target multichannel probabilistic image is input in preset depth Recurrent networks and is calculated, the comprehensive quantized data of target to obtain the heart, the comprehensive quantized data of target includes cardiac chamber diameter, myocardial wall thickness and ventricle area, therefore, comprehensive when the embodiment of the present application is able to ascend to ventricle quantization.
Description
Technical field
This application involves technical field of image processing, and in particular to a kind of biventricular quantization method, device, electronic equipment and
Storage medium.
Background technique
Magnetic resonance imaging is one of the important imaging means for detecting heart disease.Ventricle is carried out from nuclear magnetic resonance image
Quantization facilitates the early diagnosis and therapy of heart disease.In recent years, academia mainly expands in terms of left ventricle quantization
A large amount of research.
It is concentrated mainly on left ventricle due to working at present, and wants to carry out comprehensive and accurate computer-aided diagnosis, is most managed
The case where thinking is can to quantify to biventricular.The existing scheme quantified to biventricular is the calculating of four ventricular volumes,
But in terms of comprehensive quantization, what is quantified is comprehensive lower, and the effect of quantization is caused to reduce.
Summary of the invention
The embodiment of the present application provides a kind of biventricular quantization method, device, electronic equipment and storage medium, is able to ascend pair
It is comprehensive when ventricle quantifies.
The first aspect of the embodiment of the present application provides a kind of biventricular quantization method, wherein this method comprises:
Target original image is obtained, the target original image is the nuclear magnetic resonance image of heart；
The target original image is split using preset segmentation network, obtains target multichannel probabilistic image；
The target multichannel probabilistic image is input in preset depth Recurrent networks and is calculated, it is described to obtain
The comprehensive quantized data of the target of heart, the comprehensive quantized data of target include cardiac chamber diameter, myocardial wall thickness and ventricle area.
Optionally, the preset depth Recurrent networks include N layers of convolutional neural networks, described by the target multichannel
Probabilistic image is input in preset depth Recurrent networks and is calculated, the comprehensive quantized data of the target to obtain the heart,
Include:
The first layer that the target multichannel probabilistic image is input to the N layers of convolutional neural networks is calculated, is obtained
To the first calculated result, wherein N is the positive integer greater than 1；
The second layer that first calculated result is input to the N layers of convolutional neural networks is calculated, obtains second
Calculated result obtains described until the n-th layer that N-1 checkout result is input to the N layers of convolutional neural networks is calculated
The comprehensive quantized data of the target of heart.
Optionally, described to obtain preset depth recurrence net the method also includes obtaining preset depth Recurrent networks
Network, comprising:
Step 1: obtaining sample image, the sample image is the nuclear magnetic resonance image of heart, and to the sample graph
As being pre-processed, pretreatment sample image is obtained；
Step 2: the pretreatment sample image being split using the first segmentation network, obtains sample four-way probability
Image；
Step 3: the sample four-way probabilistic image being input in the N layers of convolutional neural networks and is calculated, is obtained
To predicted value；
Step 4: the predicted value and target true value being subjected to mean square deviation operation, obtain corrected value, the target true value is
The true value determined by preset true value generating function, the target true value are corresponding with the predicted value；
Step 5: the first segmentation network being corrected by the corrected value, the second segmentation network is obtained, by institute
It states the first segmentation network and replaces with the second segmentation network；
M execution step 2 is repeated to step 5, during repeating M execution step 2 to step 5, if parameter preset
When in preset parameter area, then using the N layers of convolutional neural networks as the preset depth Recurrent networks.
Optionally, described to obtain preset true value generation letter the method also includes obtaining preset true value generating function
Number includes:
The original image with cardiac silhouette mark is obtained, the original image with cardiac silhouette mark includes multiple
Profile point；
The original image with cardiac silhouette mark is handled, exposure mask figure is obtained；
The barycentric coodinates of heart are obtained from the exposure mask figure using preset acquisition modes, the barycentric coodinates include a left side
The barycentric coodinates of ventricle and the barycentric coodinates of right ventricle；
Obtain the distance between each profile point in the multiple profile point and corresponding barycentric coodinates and angle；
According to the distance between each profile point and corresponding barycentric coodinates and angle, institute is determined using interpolation method
Preset true value generating function is stated, the preset true value generating function is the function between angle and distance.
Optionally, the barycentric coodinates for obtaining heart from the exposure mask figure using preset acquisition modes, comprising:
It is determined by the first preset function and obtains profile point set from the exposure mask figure；
According to the profile point set, the barycentric coodinates are determined using the second preset function.
Optionally, the target multichannel probabilistic image includes: background area image, left/right chambers of the heart image and left cardiac muscle
Image.
Optionally, the method also includes:
Resampling is carried out to the target original image, obtains resampling image；
The normalized that 0 mean value, 1 variance is carried out to the resampling image, obtains pretreatment image.
The second aspect of the embodiment of the present application provides a kind of biventricular quantization device, wherein the device includes:
First acquisition unit, for obtaining target original image, the target original image is the nuclear magnetic resonance figures of heart
Picture；
It is more to obtain target for being split using preset segmentation network to the target original image for cutting unit
Channel probabilistic image；
Computing unit, based on the target multichannel probabilistic image is input in preset depth Recurrent networks and is carried out
It calculates, the comprehensive quantized data of target to obtain the heart, the comprehensive quantized data of target includes cardiac chamber diameter, myocardium wall thickness
Degree and ventricle area.
Optionally, the preset depth Recurrent networks include N layers of convolutional neural networks, described by the target multi-pass
Road probabilistic image is input in preset depth Recurrent networks and is calculated, the comprehensive quantized data of the target to obtain the heart
Aspect, the computing unit are specifically used for:
The first layer that the target multichannel probabilistic image is input to the N layers of convolutional neural networks is calculated, is obtained
To the first calculated result, wherein N is the positive integer greater than 1；
The second layer that first calculated result is input to the N layers of convolutional neural networks is calculated, obtains second
Calculated result obtains described until the n-th layer that N-1 checkout result is input to the N layers of convolutional neural networks is calculated
The comprehensive quantized data of the target of heart.
Optionally, described device further includes second acquisition unit, and the second acquisition unit is for obtaining preset depth
Recurrent networks, in terms of the preset depth Recurrent networks of acquisition, the second acquisition unit is specifically used for:
Step 1: obtaining sample image, the sample image is the nuclear magnetic resonance image of heart, and to the sample graph
As being pre-processed, pretreatment sample image is obtained；
Step 2: the pretreatment sample image being split using the first segmentation network, obtains sample four-way probability
Image；
Step 3: the sample four-way probabilistic image being input in the N layers of convolutional neural networks and is calculated, is obtained
To predicted value；
Step 4: the predicted value and target true value being subjected to mean square deviation operation, obtain corrected value, the target true value is
The true value determined by preset true value generating function, the target true value are corresponding with the predicted value；
Step 5: the preset segmentation network is corrected by the corrected value, obtains the second segmentation network, it will
The first segmentation network replaces with the second segmentation network；
M execution step 2 is repeated to step 5, during repeating M execution step 2 to step 5, if parameter preset
When in preset parameter area, then using the N layers of convolutional neural networks as the preset depth Recurrent networks.
Optionally, described device further includes third acquiring unit, and the third acquiring unit is for obtaining preset true value
Generating function, in terms of the preset true value generating function of acquisition, the third acquiring unit is specifically used for:
The original image with cardiac silhouette mark is obtained, the original image with cardiac silhouette mark includes multiple
Profile point；
The original image with cardiac silhouette mark is handled, exposure mask figure is obtained；
The barycentric coodinates of heart are obtained from the exposure mask figure using preset acquisition modes, the center of gravity includes left ventricle
Barycentric coodinates and right ventricle barycentric coodinates；
Obtain the distance between each profile point in the multiple profile point and corresponding barycentric coodinates and angle；
According to the distance between each profile point and corresponding barycentric coodinates and angle, institute is determined using interpolation method
Preset true value generating function is stated, the preset true value generating function is the function between angle and distance.
Optionally, in terms of the barycentric coodinates for obtaining heart from the exposure mask figure using preset acquisition modes,
The third acquiring unit is specifically used for:
It is determined by the first preset function and obtains profile point set from the exposure mask figure；
According to the profile point set, the barycentric coodinates are determined using the second preset function.
Optionally, the target multichannel probabilistic image includes: background area image, left/right chambers of the heart image and left cardiac muscle
Image.
Optionally, described device also particularly useful for:
Resampling is carried out to the target original image, obtains resampling image；
The normalized that 0 mean value, 1 variance is carried out to the resampling image, obtains pretreatment image.
The embodiment of the present application third aspect also proposes a kind of electronic equipment, comprising:
Processor；
Memory for storage processor executable instruction；
Wherein, the processor is configured to executing above-mentioned method.
The embodiment of the present application fourth aspect also proposes a kind of computer readable storage medium, is stored thereon with computer program
Instruction, realizes above-mentioned method when computer program instructions are executed by processor.
Implement the application at least to have the following beneficial effects:
By the embodiment of the present application, target original image is obtained, the target original image is the nuclear magnetic resonance figures of heart
Picture takes original image to be split the target, obtains target multichannel probabilistic image, by institute using preset segmentation network
It states target multichannel probabilistic image and is input in preset depth Recurrent networks and calculated, it is complete with the target for obtaining the heart
Face quantized data, the comprehensive quantized data of target include cardiac chamber diameter, myocardial wall thickness and ventricle area, therefore, the application
In by being split to target original image, obtain four-way probabilistic image, can be accurately anti-by image segmentation
The feature of original image is mirrored, then four-way probabilistic image is input in depth Recurrent networks and is calculated, to obtain
Comprehensive quantized data can more comprehensively obtain the amount of heart relative to the volume in existing scheme, being only capable of four ventricles of calculating
Change data, thus comprehensive when improving quantization to a certain extent.
Detailed description of the invention
In order to illustrate the technical solutions in the embodiments of the present application or in the prior art more clearly, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of application for those of ordinary skill in the art without creative efforts, can be with
It obtains other drawings based on these drawings.
Fig. 1 provides a kind of schematic diagram of biventricular quantization system for the embodiment of the present application；
Fig. 2A provides a kind of flow diagram of biventricular quantization method for the embodiment of the present application；
Fig. 2 B provides the schematic diagram that a kind of pair of pretreatment image is split for the embodiment of the present application；
Fig. 2 C provides a kind of structural schematic diagram of preset depth Recurrent networks for the embodiment of the present application；
Fig. 3 provides another biventricular quantization method schematic diagram for the embodiment of the present application；
Fig. 4 the embodiment of the present application provides the flow diagram of another biventricular quantization method；
Fig. 5 is a kind of structural schematic diagram of terminal provided by the embodiments of the present application；
Fig. 6 provides a kind of structural schematic diagram of biventricular quantization device for the embodiment of the present application；
Fig. 7 provides the structural schematic diagram of another biventricular quantization device for the embodiment of the present application；
Fig. 8 provides the structural schematic diagram of another biventricular quantization device for the embodiment of the present application.
Specific embodiment
Below in conjunction with the attached drawing in the embodiment of the present application, technical solutions in the embodiments of the present application carries out clear, complete
Site preparation description, it is clear that described embodiments are only a part of embodiments of the present application, instead of all the embodiments.It is based on
Embodiment in the application, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall in the protection scope of this application.
The description and claims of this application and term " first " in above-mentioned attached drawing, " second " etc. are for distinguishing
Different objects, are not use to describe a particular order.In addition, term " includes " and " having " and their any deformations, it is intended that
It is to cover and non-exclusive includes.Such as the process, method, system, product or equipment for containing a series of steps or units do not have
It is defined in listed step or unit, but optionally further comprising the step of not listing or unit, or optionally also wrap
Include other step or units intrinsic for these process, methods, product or equipment.
" embodiment " mentioned in this application is it is meant that a particular feature, structure, or characteristic described can be in conjunction with the embodiments
Included at least one embodiment of the application.The phrase, which occurs, in each position in the description might not each mean phase
Same embodiment, nor the independent or alternative embodiment with other embodiments mutual exclusion.Those skilled in the art are explicitly
Implicitly understand, embodiments described herein can be combined with other embodiments.
Electronic device involved by the embodiment of the present application may include the various handheld devices with wireless communication function,
Mobile unit, wearable device calculate equipment or are connected to other processing equipments and various forms of radio modem
User equipment (user equipment, UE), mobile station (mobile station, MS), terminal device (terminal
Device) etc..For convenience of description, apparatus mentioned above is referred to as electronic device.
In order to better understand a kind of biventricular quantization method provided by the embodiments of the present application, first below to biventricular amount
Change method is briefly introduced.Referring to Fig. 1, Fig. 1 provides a kind of signal of biventricular quantization system for the embodiment of the present application
Figure.As shown in Figure 1, biventricular quantization system includes pretreatment network 101, segmentation network 102 and depth Recurrent networks 103, head
First biventricular quantization system obtains original image, and original image is the nuclear magnetic resonance image of heart, and it is right then to pre-process network 101
Original image is pre-processed, and pretreatment image is obtained, and pretreatment image is then input to preset segmentation network 102, right
The pretreatment image is split, and obtains target multichannel probabilistic image, is finally input to target multichannel probabilistic image pre-
If depth Recurrent networks 103 in calculated, obtain the comprehensive quantized data of target, the comprehensive quantized data of the target includes heart
Diameter, myocardial wall thickness and ventricle area, in application scheme, by being pre-processed to original image, to pretreated
Image is split, and obtains four-way probabilistic image, by image segmentation, can accurately reflect the spy of original image
Four-way probabilistic image, is then input in depth Recurrent networks and calculates, to obtain comprehensive quantized data, relatively by sign
In existing scheme, it is only capable of calculating the volume of four ventricles, the quantized data of heart can be more comprehensively obtained, thus certain journey
It is comprehensive when improving quantization on degree.
Fig. 2A is please referred to, Fig. 2A provides a kind of flow diagram of biventricular quantization method for the embodiment of the present application.Such as
Shown in Fig. 2A, biventricular quantization method includes step 201-204, specific as follows:
201, target original image is obtained, the target original image is the nuclear magnetic resonance image of heart.
Wherein, original image can be inputted by user, after can also generating nuclear magnetic resonance image by Nuclear Magnetic Resonance, directly
It is transmitted in biventricular quantization system.The original image may be by the image after silhouette markup, and silhouette markup is pair
The profile of heart in original image is marked, the label by multiple feature point groups at.
It optionally,, can also be to target original image in order to promote the effect of segmentation after getting target original image
It is pre-processed.
Wherein, a kind of possible method for being handled to obtain pretreatment image to original image includes step A1-A2, tool
Body is as follows:
A1, resampling is carried out to the target original image, obtains resampling image；
Wherein, the method for resampling include closest interpolation method (nearest neighbor interpolation), it is double
Linear interpolation (bilinear interpolation) and cubic convolution method interpolation (cubic convolution
interpolation).This programme can carry out resampling using any one in the above method, obtain resampling image.
A2, the normalized that 0 mean value, 1 variance is carried out to the resampling image, obtain the pretreatment image.
Wherein, in the normalized of 0 mean value, 1 variance, 0 mean value can be understood as each pixel in resampling picture
The sum of gray value be 0；1 variance can be understood as the variance of the gray value of each pixel in resampling picture being 1, return
The data (gray value) of resampling picture are mapped to the value between [0,1] after one change processing.
202, the pretreatment image is split using preset segmentation network, obtains target multichannel probability graph
Picture.
Optionally, include M convolution block in preset segmentation network, pretreatment image is input in segmentation network, is held
Row convolution algorithm, sampling operation, the polymerization of iteration depth and level depth aminated polyepichlorohydrin, finally export target four-way probability graph,
Wherein, sampling operation is the sampling rate that promoted.The schematic diagram that a kind of pair of pretreatment image is split is as shown in Figure 2 B, presets
Segmentation network include 26 convolution blocks, 4 iteration depth converging networks, the polymerization of 3 level depth, 5 sampling operations, finally
Export target four-way probability graph, wherein 1 to 26 respectively represents convolution block 1 to convolution block 26.
Wherein, target multichannel probabilistic image includes: background area image, left/right chambers of the heart image and left myocardium image,
In, it is specifically background area image, left myocardium image and left ventricular cavity image or background area image, right myocardium image and the right side
Chambers of the heart image.
203, the target multichannel probabilistic image is input in preset depth Recurrent networks and is calculated, to obtain
The comprehensive quantized data of the target of the heart, the comprehensive quantized data of target includes cardiac chamber diameter, myocardial wall thickness and ventricle
Area.
Fig. 2 C is please referred to, Fig. 2 C provides a kind of structural representation of preset depth Recurrent networks for the embodiment of the present application
Figure.As shown in Figure 2 C, depth Recurrent networks include 11 layers of convolutional neural networks, specifically: conv1, relu layers；4 max,
Pool layers；Conv2, relu layers；Conv3, relu layers；Conv4, relu layers；Conv5, relu layers；Fc, relu layers, fc,
Sigmoid layers.
Optionally, depth Recurrent networks include N layers of convolutional neural networks, a kind of that target four-way probability graph is input to depth
It being calculated in degree Recurrent networks, the method for obtaining comprehensive quantized data includes step B1-B2, specific as follows:
B1, the first layer that the target multichannel probabilistic image is input to the N layers of convolutional neural networks is calculated,
Obtain the first calculated result, wherein N is the positive integer greater than 1；
Optionally, it is illustrated by taking the depth Recurrent networks that Fig. 2 C is provided as an example, target multichannel probabilistic image is inputted
To conv1, relu layers, by conv1, relu layers are calculated, and obtain the first calculated result；First calculated result is input to
The second layer (max, pool layers), by the second calculated result is calculated.
B2, the second layer that first calculated result is input to the N layers of convolutional neural networks is calculated, obtains
Two calculated results obtain institute until the n-th layer that N-1 calculated result is input to the N layers of convolutional neural networks is calculated
State the comprehensive quantized data of target of heart.
Optionally, in the application, the tenth calculated result is input to fc, sigmoid layers are calculated, to obtain target
Comprehensive quantized data.
Optionally, the embodiment of the present application also provides a kind of method for obtaining preset depth Recurrent networks, this method packets
Include following steps:
Step 1: obtaining sample image, the sample image is the nuclear magnetic resonance image of heart, and to the sample graph
As being pre-processed, pretreatment sample image is obtained；
Wherein, sample image is pre-processed, it is right in step A1-A2 that the method for obtaining pretreatment sample image can refer to
Original image carries out pretreated mode and is handled.
Step 2: the pretreatment sample image being split using the first segmentation network, obtains sample four-way probability
Image；
Wherein, the first segmentation network is preset segmentation network, and preset segmentation network is such as the segmentation network in Fig. 2 B.
Step 3: the sample four-way probabilistic image being input in N layers of convolutional neural networks and is calculated, is obtained pre-
Measured value；
Wherein, the N layers of convolutional neural networks are the network not being trained also, and output result is the number that do not classify
According to i.e. predicted value.
Step 4: the predicted value and target true value being subjected to mean square deviation operation, obtain corrected value, the target true value is
The true value determined by preset true value generating function, the target true value are corresponding with the predicted value；
Optionally, the embodiment of the present application also provides a kind of method for obtaining true value generating function, including step C1-C5,
It is specific as follows:
C1, the original image with cardiac silhouette mark is obtained, the original image with cardiac silhouette mark includes
Multiple profile points；
Wherein, the original image with cardiac silhouette mark can also use machine by obtaining after being manually labeled
The methods of the mode of device mark obtains.Cardiac silhouette mark includes: left myocardial contours mark, left ventricular cavity profile mark, right cardiac muscle
Profile mark, right cardiac cavity profile mark etc., this is illustrated for sentencing left cardiac muscle and the mark of left ventricular cavity, that is, has heart wheel
The original image of exterior feature mark is the image with left myocardial contours mark and left ventricular cavity profile mark.
C2, the original image with cardiac silhouette mark is handled, obtains exposure mask figure；
Wherein, original image is handled, obtains the mode of exposure mask figure are as follows: mark the background image in original image
It is 0, is 1 by left myocardial contours, left ventricular cavity silhouette markup, to obtain exposure mask figure, wherein exposure mask figure is 0-1mask figure.
C3, the barycentric coodinates for obtaining heart from the exposure mask figure using preset acquisition modes, the barycentric coodinates packet
Include the barycentric coodinates of left ventricle and the barycentric coodinates of right ventricle；
Optionally, a kind of method of the possible barycentric coodinates that heart is obtained from the exposure mask figure includes step C31-
C32, specific as follows:
C31, the acquisition profile point set from the exposure mask figure is determined by the first preset function；
Wherein, the first preset function is the find_contours function in skimage function packet, find_contours letter
Number can extract profile from two-valued function, and then also available profile point set.
It optionally, further include establishing the first coordinate system, original graph before determining mask image according to the first preset function
As being rectangular image, the origin of the first rectangular coordinate system is located at the top left corner apex of original image, with straight down for x-axis just
Direction is horizontally to the right positive direction of the y-axis.
C32, according to the profile point set, the barycentric coodinates are determined using the second preset function.
Wherein, the second preset function is the center_of_mass function in scipy function packet, passes through center_of_
Mass function can determine barycentric coodinates directly according to profile point set.
Optionally, after determining barycentric coodinates, using the barycentric coodinates as origin, with straight down for positive direction of the x-axis, water
Flat be positive direction of the y-axis to the right, establishes the second coordinate system, and the second coordinate system is polar coordinate system, and by the coordinate of profile point from first
Coordinate system is converted into the coordinate in the second coordinate system.
C4, the distance between each profile point in the multiple profile point and corresponding barycentric coodinates and angle are obtained；
Wherein, barycentric coodinates corresponding with each profile point it is to be understood that left myocardial contours point and left ventricle center of gravity
Coordinate is corresponding, and right myocardial contours point is corresponding with the barycentric coodinates of right ventricle.
Optionally, the distance between profile point and barycentric coodinates can be directly calculated by distance calculation formula, or
Person directly measures and obtains, the angle between profile point and barycentric coodinates it is to be understood that the line of profile point and center of gravity and y-axis just
The angle in direction.
C5, according to the distance between each profile point and corresponding barycentric coodinates and angle, determined using interpolation method
The preset true value generating function out, the preset true value generating function are the function between angle and distance.
Optionally, the method for the preset true value generating function being determined according to interpolation method are as follows: according to profile point and again
The distance between heart and angle establish the discrete function of angle and distance；The discrete function is being connected into using interpolation method
Continuous function, to obtain preset true value generating function.
Step 5: the first segmentation network being corrected by the corrected value, the second segmentation network is obtained, by institute
It states the first segmentation network and replaces with the second segmentation network；
Wherein, corrected value is used to be corrected as using corrected value to the first segmentation network to pre- first segmentation network
It handles generated error when picture is split to be corrected, output result be optimized, thus the output corrected
As a result.
M execution step 2 is repeated to step 5, during repeating M execution step 2 to step 5, if parameter preset
When in preset parameter area, then using the N layers of convolutional neural networks as the preset depth Recurrent networks.
Wherein, parameter preset can difference between loss value, person's true value and index value, preset parameter area can lead to
It crosses empirical value setting or is set by historical data, loss value can be understood as corrected value.
The embodiment of the present application, biventricular quantization system export comprehensive quantized data, can relatively before left ventricle quantization, more
The part of ventricle quantifies, and biventricular quantifies more can react to objective reality the real conditions of heart comprehensively, facilitates heart
Segmentation and Recurrent networks are incorporated in the early diagnosis and therapy and biventricular quantization system of disease, it can be to a certain extent
The accuracy of comprehensive quantized data is promoted, and by devising a kind of new true value generation method, relative to 2D center
The true value such as line generate standard, optimize generating process, improve efficiency when true value generates.
Referring to Fig. 3, Fig. 3 provides another biventricular quantization method schematic diagram for the embodiment of the present application.Such as Fig. 3 institute
Show, original image 301 obtains pretreatment image, pretreatment image is then input to segmentation network after pretreatment 302
Be split in 303, obtain four-way probabilistic image 304, by four-way probabilistic image 304 be input to segmentation network 305 in into
Row calculates, and obtains comprehensive quantized data 306, wherein comprehensive quantized data includes: left ventricle chambers of the heart area, left myocardial area, a left side
Cardiac chamber diameter, left myocardial wall thickness, right ventricle chambers of the heart area, right cardiac cavity diameter specifically indicate in figure are as follows: A1: right cardiac cavity area；
A2: left myocardial area；A3: left ventricular cavity area；D1-d3: left ventricular cavity diameter；Rd1-rd2: right cardiac cavity diameter；A-AS(A,AL,IL,
I, IS, AS): left myocardial wall thickness.
Referring to Fig. 4, Fig. 4 the embodiment of the present application provides the flow diagram of another biventricular quantization method, including
Step 401-410, specific as follows:
401, sample image is obtained, the sample image is the nuclear magnetic resonance image of heart, and to the sample image
It is pre-processed, obtains pretreatment sample image；
402, the pretreatment sample image is split using the first segmentation network, obtains sample four-way probability graph
Picture；
403, the sample four-way probabilistic image is input in the N layers of convolutional neural networks and is calculated, obtained
Predicted value；
404, the predicted value and target true value are subjected to mean square deviation operation, obtain corrected value, the target true value is logical
The true value that preset true value generating function is determined is crossed, the target true value is corresponding with the predicted value；
405, the first segmentation network is corrected by the corrected value, obtains the second segmentation network, it will be described
First segmentation network replaces with the second segmentation network；
406, step 402 is repeated to step 405, until the predicted value of N layers of convolutional neural networks output is to measure comprehensively
When changing data, using the N layers of convolutional neural networks as preset depth Recurrent networks；
407, original image is obtained, the target original image is the nuclear magnetic resonance image of heart；
408, the target original image is pre-processed, obtains pretreatment image；
409, the pretreatment image is split using preset segmentation network, obtains target multichannel probability graph
Picture；
410, the target multichannel probabilistic image is input in preset depth Recurrent networks and is calculated, to obtain
The comprehensive quantized data of the target of the heart, the comprehensive quantized data of target includes cardiac chamber diameter, myocardial wall thickness and ventricle
Area.
In this example, by constructing depth Recurrent networks, and the end pair that segmentation network and depth Recurrent networks will be realized
End connection, to handle original image, obtain comprehensive quantized data, can relative in existing scheme only to left ventricle
Quantization, the part quantization of oversensitive room, can quantify comprehensively biventricular, therefore more can react to objective reality the true of heart
Real situation, thus accuracy when improving data acquisition and comprehensive.
It is consistent with above-described embodiment, referring to Fig. 5, Fig. 5 is that a kind of structure of terminal provided by the embodiments of the present application is shown
It is intended to, as shown, including processor, input equipment, output equipment and memory, the processor, input equipment, output are set
Standby and memory is connected with each other, wherein for the memory for storing computer program, the computer program includes that program refers to
It enables, the processor is configured for calling described program instruction, and above procedure includes the instruction for executing following steps；
Target original image is obtained, the target original image is the nuclear magnetic resonance image of heart；
The target original image is split using preset segmentation network, obtains target multichannel probabilistic image；
The target multichannel probabilistic image is input in preset depth Recurrent networks and is calculated, it is described to obtain
The comprehensive quantized data of the target of heart, the comprehensive quantized data of target include cardiac chamber diameter, myocardial wall thickness and ventricle area.
In this example, target original image is obtained, the target original image is the nuclear magnetic resonance image of heart, using pre-
If segmentation network the target original image is split, target multichannel probabilistic image is obtained, by the target multi-pass
Road probabilistic image is input in preset depth Recurrent networks and is calculated, and quantifies number comprehensively to obtain the target of the heart
Include cardiac chamber diameter, myocardial wall thickness and ventricle area according to, the comprehensive quantized data of target, therefore, in the application by pair
Target original image is split, and obtains four-way probabilistic image, by image segmentation, can accurately be reflected original
Then four-way probabilistic image is input in depth Recurrent networks and calculates by the feature of image, to be quantified comprehensively
Data can more comprehensively obtain the quantized data of heart relative to the volume in existing scheme, being only capable of four ventricles of calculating,
Thus comprehensive when improving quantization to a certain extent.
It is above-mentioned that mainly the scheme of the embodiment of the present application is described from the angle of method side implementation procedure.It is understood that
, in order to realize the above functions, it comprises execute the corresponding hardware configuration of each function and/or software module for terminal.This
Field technical staff should be readily appreciated that, in conjunction with each exemplary unit and algorithm of embodiment description presented herein
Step, the application can be realized with the combining form of hardware or hardware and computer software.Some function actually with hardware also
It is the mode of computer software driving hardware to execute, the specific application and design constraint depending on technical solution.Profession
Technical staff can specifically realize described function to each using distinct methods, but this realization should not be recognized
For beyond scope of the present application.
The embodiment of the present application can carry out the division of functional unit according to above method example to terminal, for example, can be right
The each functional unit of each function division is answered, two or more functions can also be integrated in a processing unit.
Above-mentioned integrated unit both can take the form of hardware realization, can also realize in the form of software functional units.It needs
Illustrate, is schematical, only a kind of logical function partition to the division of unit in the embodiment of the present application, it is practical to realize
When there may be another division manner.
Consistent with the above, referring to Fig. 6, Fig. 6 provides a kind of knot of biventricular quantization device for the embodiment of the present application
Structure schematic diagram.Biventricular quantization device includes first acquisition unit 601, cutting unit 602 and computing unit 603, wherein
First acquisition unit 601, for obtaining target original image, the target original image is the nuclear magnetic resonance of heart
Image；
Cutting unit 602 obtains target for being split using preset segmentation network to the target original image
Multichannel probabilistic image；
Computing unit 603, for by the target multichannel probabilistic image be input in preset depth Recurrent networks into
Row calculates, and the comprehensive quantized data of target to obtain the heart, the comprehensive quantized data of target includes cardiac chamber diameter, cardiac muscle
Wall thickness and ventricle area.
Optionally, the preset depth Recurrent networks include N layers of convolutional neural networks, described by the target multi-pass
Road probabilistic image is input in preset depth Recurrent networks and is calculated, the comprehensive quantized data of the target to obtain the heart
Aspect, the computing unit 603 are specifically used for:
The first layer that the target multichannel probabilistic image is input to the N layers of convolutional neural networks is calculated, is obtained
To the first calculated result, wherein N is the positive integer greater than 1；
The second layer that first calculated result is input to the N layers of convolutional neural networks is calculated, obtains second
Calculated result obtains described until the n-th layer that N-1 checkout result is input to the N layers of convolutional neural networks is calculated
The comprehensive quantized data of the target of heart.
Referring to Fig. 7, Fig. 7 provides the structural schematic diagram of another biventricular quantization device, institute for the embodiment of the present application
Stating device further includes second acquisition unit 604, and the second acquisition unit 604 is used to obtain preset depth Recurrent networks,
Described to obtain preset depth Recurrent networks aspect, the second acquisition unit 604 is specifically used for:
Step 1: obtaining sample image, the sample image is the nuclear magnetic resonance image of heart, and to the sample graph
As being pre-processed, pretreatment sample image is obtained；
Step 2: the pretreatment sample image being split using the first segmentation network, obtains sample four-way probability
Image；
Step 3: the sample four-way probabilistic image being input in the N layers of convolutional neural networks and is calculated, is obtained
To predicted value；
Step 4: the predicted value and target true value being subjected to mean square deviation operation, obtain corrected value, the target true value is
The true value determined by preset true value generating function, the target true value are corresponding with the predicted value；
Step 5: the preset segmentation network is corrected by the corrected value, obtains the second segmentation network, it will
The first segmentation network replaces with the second segmentation network；
Step 2 is repeated to step 5, until when the predicted value of N layer convolutional neural networks output is comprehensive quantized data,
Using the N layers of convolutional neural networks as the preset depth Recurrent networks.
Referring to Fig. 8, Fig. 8 provides the structural schematic diagram of another biventricular quantization device, institute for the embodiment of the present application
Stating device further includes third acquiring unit 605, and the third acquiring unit 605 is used to obtain preset true value generating function,
Described to obtain preset true value generating function aspect, the third acquiring unit 605 is specifically used for:
The original image with cardiac silhouette mark is obtained, the original image with cardiac silhouette mark includes multiple
Profile point；
The original image with cardiac silhouette mark is handled, exposure mask figure is obtained；
The barycentric coodinates of heart are obtained from the exposure mask figure using preset acquisition modes, the center of gravity includes left ventricle
Barycentric coodinates and right ventricle barycentric coodinates；
Obtain the distance between each profile point in the multiple profile point and corresponding barycentric coodinates and angle；
According to the distance between each profile point and corresponding barycentric coodinates and angle, institute is determined using interpolation method
Preset true value generating function is stated, the preset true value generating function is the function between angle and distance.
Optionally, in terms of the barycentric coodinates for obtaining heart from the exposure mask figure using preset acquisition modes,
The third acquiring unit 605 is specifically used for:
It is determined by the first preset function and obtains profile point set from the exposure mask figure；
According to the profile point set, the barycentric coodinates are determined using the second preset function.
Optionally, the target multichannel probabilistic image includes: background area image, left/right chambers of the heart image and left cardiac muscle
Image.
Optionally, described device also particularly useful for:
Resampling is carried out to the target original image, obtains resampling image；
The normalized that 0 mean value, 1 variance is carried out to the resampling image, obtains pretreatment image.
The embodiment of the present application also provides a kind of computer storage medium, wherein computer storage medium storage is for electricity
The computer program of subdata exchange, it is as any in recorded in above method embodiment which execute computer
A kind of some or all of biventricular quantization method step.
The embodiment of the present application also provides a kind of computer program product, and the computer program product includes storing calculating
The non-transient computer readable storage medium of machine program, the computer program make computer execute such as above method embodiment
Some or all of any biventricular quantization method of middle record step.
It should be noted that for the various method embodiments described above, for simple description, therefore, it is stated as a series of
Combination of actions, but those skilled in the art should understand that, the application is not limited by the described action sequence because
According to the application, some steps may be performed in other sequences or simultaneously.Secondly, those skilled in the art should also know
It knows, the embodiments described in the specification are all preferred embodiments, related actions and modules not necessarily the application
It is necessary.
In the above-described embodiments, it all emphasizes particularly on different fields to the description of each embodiment, there is no the portion being described in detail in some embodiment
Point, reference can be made to the related descriptions of other embodiments.
In several embodiments provided herein, it should be understood that disclosed device, it can be by another way
It realizes.For example, the apparatus embodiments described above are merely exemplary, such as the division of the unit, it is only a kind of
Logical function partition, there may be another division manner in actual implementation, such as multiple units or components can combine or can
To be integrated into another system, or some features can be ignored or not executed.Another point, shown or discussed is mutual
Coupling, direct-coupling or communication connection can be through some interfaces, the indirect coupling or communication connection of device or unit,
It can be electrical or other forms.
The unit as illustrated by the separation member may or may not be physically separated, aobvious as unit
The component shown may or may not be physical unit, it can and it is in one place, or may be distributed over multiple
In network unit.It can select some or all of unit therein according to the actual needs to realize the mesh of this embodiment scheme
's.
It, can also be in addition, applying for that each functional unit in bright each embodiment can integrate in one processing unit
It is that each unit physically exists alone, can also be integrated in one unit with two or more units.Above-mentioned integrated list
Member both can take the form of hardware realization, can also be realized in the form of software program module.
If the integrated unit is realized in the form of software program module and sells or use as independent product
When, it can store in a computer-readable access to memory.Based on this understanding, the technical solution of the application substantially or
Person says that all or part of the part that contributes to existing technology or the technical solution can body in the form of software products
Reveal and, which is stored in a memory, including some instructions are used so that a computer equipment
(can be personal computer, server or network equipment etc.) executes all or part of each embodiment the method for the application
Step.And memory above-mentioned includes: USB flash disk, read-only memory (read-only memory, ROM), random access memory
The various media that can store program code such as (random access memory, RAM), mobile hard disk, magnetic or disk.
Those of ordinary skill in the art will appreciate that all or part of the steps in the various methods of above-described embodiment is can
It is completed with instructing relevant hardware by program, which can store in a computer-readable memory, memory
It may include: flash disk, read-only memory, random access device, disk or CD etc..
The embodiment of the present application is described in detail above, specific case used herein to the principle of the application and
Embodiment is expounded, the description of the example is only used to help understand the method for the present application and its core ideas；
At the same time, for those skilled in the art can in specific embodiments and applications according to the thought of the application
There is change place, in conclusion the contents of this specification should not be construed as limiting the present application.
Claims (10)
1. a kind of biventricular quantization method, which is characterized in that the described method includes:
Target original image is obtained, the target original image is the nuclear magnetic resonance image of heart；
The target original image is split using preset segmentation network, obtains target multichannel probabilistic image；
The target multichannel probabilistic image is input in preset depth Recurrent networks and is calculated, to obtain the heart
The comprehensive quantized data of target, the comprehensive quantized data of target includes cardiac chamber diameter, myocardial wall thickness and ventricle area.
2. the method according to claim 1, wherein the preset depth Recurrent networks include N layers of convolution mind
Through network, the described target multichannel probabilistic image is input in preset depth Recurrent networks is calculated, to obtain
The comprehensive quantized data of the target of the heart, comprising:
The first layer that the target multichannel probabilistic image is input to the N layers of convolutional neural networks is calculated, obtains
One calculated result, wherein N is the positive integer greater than 1；
The second layer that first calculated result is input to the N layers of convolutional neural networks is calculated, the second calculating is obtained
As a result, obtaining the heart up to calculating the n-th layer that N-1 checkout result is input to the N layers of convolutional neural networks
The comprehensive quantized data of target.
3. according to the method described in claim 2, it is characterized in that, returning net the method also includes obtaining preset depth
Network, it is described to obtain preset depth Recurrent networks, comprising:
Step 1: obtain sample image, the sample image be heart nuclear magnetic resonance image, and to the sample image into
Row pretreatment obtains pretreatment sample image；
Step 2: the pretreatment sample image being split using the first segmentation network, obtains sample four-way probability graph
Picture；
Step 3: the sample four-way probabilistic image being input in the N layers of convolutional neural networks and is calculated, is obtained pre-
Measured value；
Step 4: the predicted value and target true value being subjected to mean square deviation operation, obtain corrected value, the target true value is to pass through
The true value that preset true value generating function is determined, the target true value are corresponding with the predicted value；
Step 5: the first segmentation network being corrected by the corrected value, the second segmentation network is obtained, by described the
One segmentation network replaces with the second segmentation network；
M execution step 2 is repeated to step 5, during repeating M execution step 2 to step 5, if parameter preset is in
When preset parameter area, then using the N layers of convolutional neural networks as the preset depth Recurrent networks.
4. according to the method described in claim 3, it is characterized in that, generating letter the method also includes obtaining preset true value
Number, the preset true value generating function of acquisition include:
The original image with cardiac silhouette mark is obtained, the original image with cardiac silhouette mark includes multiple profiles
Point；
The original image with cardiac silhouette mark is handled, exposure mask figure is obtained；
The barycentric coodinates of heart are obtained from the exposure mask figure using preset acquisition modes, the barycentric coodinates include left ventricle
Barycentric coodinates and right ventricle barycentric coodinates；
Obtain the distance between each profile point in the multiple profile point and corresponding barycentric coodinates and angle；
According to the distance between each profile point and corresponding barycentric coodinates and angle, determined using interpolation method described pre-
If true value generating function, the preset true value generating function be angle and distance between function.
5. according to the method described in claim 4, it is characterized in that, described use preset acquisition modes from the exposure mask figure
Obtain the barycentric coodinates of heart, comprising:
It is determined by the first preset function and obtains profile point set from the exposure mask figure；
According to the profile point set, the barycentric coodinates are determined using the second preset function.
6. a kind of biventricular quantization device, which is characterized in that described device includes:
First acquisition unit, for obtaining target original image, the target original image is the nuclear magnetic resonance image of heart；
Cutting unit obtains target multichannel for being split using preset segmentation network to the target original image
Probabilistic image；
Computing unit is calculated for the target multichannel probabilistic image to be input in preset depth Recurrent networks,
The comprehensive quantized data of target to obtain the heart, the comprehensive quantized data of target includes cardiac chamber diameter, myocardial wall thickness
With ventricle area.
7. device according to claim 6, which is characterized in that the preset depth Recurrent networks include N layers of convolution mind
Through network, calculated in the described target multichannel probabilistic image is input in preset depth Recurrent networks, with
To the comprehensive quantized data aspect of target of the heart, the computing unit is specifically used for:
The first layer that the target multichannel probabilistic image is input to the N layers of convolutional neural networks is calculated, obtains
One calculated result, wherein N is the positive integer greater than 1；
The second layer that first calculated result is input to the N layers of convolutional neural networks is calculated, the second calculating is obtained
As a result, obtaining the heart up to calculating the n-th layer that N-1 checkout result is input to the N layers of convolutional neural networks
The comprehensive quantized data of target.
8. device according to claim 7, which is characterized in that described device further includes second acquisition unit, and described second
Acquiring unit is for obtaining preset depth Recurrent networks, in terms of the preset depth Recurrent networks of acquisition, described second
Acquiring unit is specifically used for:
Step 1: obtain sample image, the sample image be heart nuclear magnetic resonance image, and to the sample image into
Row pretreatment obtains pretreatment sample image；
Step 2: the pretreatment sample image being split using the first segmentation network, obtains sample four-way probability graph
Picture；
Step 3: the sample four-way probabilistic image being input in the N layers of convolutional neural networks and is calculated, is obtained pre-
Measured value；
Step 4: the predicted value and target true value being subjected to mean square deviation operation, obtain corrected value, the target true value is to pass through
The true value that preset true value generating function is determined, the target true value are corresponding with the predicted value；
Step 5: the preset segmentation network is corrected by the corrected value, obtains the second segmentation network, it will be described
First segmentation network replaces with the second segmentation network；
M execution step 2 is repeated to step 5, during repeating M execution step 2 to step 5, if parameter preset is in
When preset parameter area, then using the N layers of convolutional neural networks as the preset depth Recurrent networks.
9. a kind of electronic equipment characterized by comprising
Processor；
Memory for storage processor executable instruction；
Wherein, the processor is configured to: perform claim require any one of 1 to 5 described in method.
10. a kind of computer readable storage medium, is stored thereon with computer program instructions, which is characterized in that the computer
Method described in any one of claim 1 to 5 is realized when program instruction is executed by processor.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811534455.6A CN109741395B (en) | 2018-12-14 | 2018-12-14 | Dual-chamber quantification method and device, electronic equipment and storage medium |
| CN202111063832.4A CN113674338A (en) | 2018-12-14 | 2018-12-14 | Dual-chamber quantification method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811534455.6A CN109741395B (en) | 2018-12-14 | 2018-12-14 | Dual-chamber quantification method and device, electronic equipment and storage medium |
Related Child Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111063832.4A Division CN113674338A (en) | 2018-12-14 | 2018-12-14 | Dual-chamber quantification method and device, electronic equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109741395A true CN109741395A (en) | 2019-05-10 |
| CN109741395B CN109741395B (en) | 2021-07-23 |
Family
ID=66359527
Family Applications (2)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111063832.4A Withdrawn CN113674338A (en) | 2018-12-14 | 2018-12-14 | Dual-chamber quantification method and device, electronic equipment and storage medium |
| CN201811534455.6A Active CN109741395B (en) | 2018-12-14 | 2018-12-14 | Dual-chamber quantification method and device, electronic equipment and storage medium |
Family Applications Before (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111063832.4A Withdrawn CN113674338A (en) | 2018-12-14 | 2018-12-14 | Dual-chamber quantification method and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (2) | CN113674338A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110189310A (en) * | 2019-05-24 | 2019-08-30 | 上海联影智能医疗科技有限公司 | Acquisition methods, computer equipment and the storage medium of image feature value |
| CN110378201A (en) * | 2019-06-05 | 2019-10-25 | 浙江零跑科技有限公司 | A kind of hinged angle measuring method of multiple row vehicle based on side ring view fisheye camera input |
| CN110739050A (en) * | 2019-12-20 | 2020-01-31 | 深圳大学 | left ventricle full parameter and confidence degree quantification method |
| CN113436139A (en) * | 2021-05-10 | 2021-09-24 | 上海大学 | Small intestine nuclear magnetic resonance image identification and physiological information extraction system and method based on deep learning |
| WO2021258955A1 (en) * | 2020-06-24 | 2021-12-30 | 中兴通讯股份有限公司 | Method and apparatus for marking object outline in target image, and storage medium and electronic apparatus |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102397070A (en) * | 2011-01-26 | 2012-04-04 | 无锡宙示医学软件开发有限公司 | Method for fully-automatically segmenting and quantifying left ventricle of cardiac magnetic resonance image |
| US20150379724A1 (en) * | 2014-06-26 | 2015-12-31 | The Governors Of The University Of Alberta | Cardiac segmentation with point correspondence |
| CN106600596A (en) * | 2016-12-21 | 2017-04-26 | 南昌航空大学 | Dynamic display method and system for heart cavity |
| CN106780467A (en) * | 2016-12-20 | 2017-05-31 | 深圳先进技术研究院 | Cine cardiac imaging quantization method, EEP image determine method and device |
| CN107330888A (en) * | 2017-07-11 | 2017-11-07 | 中国人民解放军第三军医大学 | Each chamber dividing method of dynamic heart based on CTA images |
| CN108013904A (en) * | 2017-12-15 | 2018-05-11 | 无锡祥生医疗科技股份有限公司 | Cardiac ultrasonic imaging method |
| CN108230342A (en) * | 2017-12-29 | 2018-06-29 | 电子科技大学 | A kind of left and right ventricles level-set segmentation methods based on cardiac anatomy |
Family Cites Families (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP3074949A2 (en) * | 2013-11-27 | 2016-10-05 | Universidad Politecnica De Madrid | Method and system for determining the prognosis of a patient suffering from pulmonary embolism |
| CN108603922A (en) * | 2015-11-29 | 2018-09-28 | 阿特瑞斯公司 | Automatic cardiac volume is divided |
| CN106022221B (en) * | 2016-05-09 | 2021-11-30 | 腾讯科技（深圳）有限公司 | Image processing method and system |
| CN106296692A (en) * | 2016-08-11 | 2017-01-04 | 深圳市未来媒体技术研究院 | Image significance detection method based on antagonism network |
| US10582907B2 (en) * | 2016-10-31 | 2020-03-10 | Siemens Healthcare Gmbh | Deep learning based bone removal in computed tomography angiography |
| US10528848B2 (en) * | 2016-11-04 | 2020-01-07 | Case Western Reserve University | Histomorphometric classifier to predict cardiac failure from whole-slide hematoxylin and eosin stained images |
| WO2018222755A1 (en) * | 2017-05-30 | 2018-12-06 | Arterys Inc. | Automated lesion detection, segmentation, and longitudinal identification |
| GB201716890D0 (en) * | 2017-10-13 | 2017-11-29 | Optellum Ltd | System, method and apparatus for assisting a determination of medical images |
-
2018
- 2018-12-14 CN CN202111063832.4A patent/CN113674338A/en not_active Withdrawn
- 2018-12-14 CN CN201811534455.6A patent/CN109741395B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102397070A (en) * | 2011-01-26 | 2012-04-04 | 无锡宙示医学软件开发有限公司 | Method for fully-automatically segmenting and quantifying left ventricle of cardiac magnetic resonance image |
| US20150379724A1 (en) * | 2014-06-26 | 2015-12-31 | The Governors Of The University Of Alberta | Cardiac segmentation with point correspondence |
| CN106780467A (en) * | 2016-12-20 | 2017-05-31 | 深圳先进技术研究院 | Cine cardiac imaging quantization method, EEP image determine method and device |
| CN106600596A (en) * | 2016-12-21 | 2017-04-26 | 南昌航空大学 | Dynamic display method and system for heart cavity |
| CN107330888A (en) * | 2017-07-11 | 2017-11-07 | 中国人民解放军第三军医大学 | Each chamber dividing method of dynamic heart based on CTA images |
| CN108013904A (en) * | 2017-12-15 | 2018-05-11 | 无锡祥生医疗科技股份有限公司 | Cardiac ultrasonic imaging method |
| CN108230342A (en) * | 2017-12-29 | 2018-06-29 | 电子科技大学 | A kind of left and right ventricles level-set segmentation methods based on cardiac anatomy |
Non-Patent Citations (1)
| Title |
|---|
| 萧毅 等: "人工智能技术在医学影像中的应用讨论", 《第二军医大学学报》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110189310A (en) * | 2019-05-24 | 2019-08-30 | 上海联影智能医疗科技有限公司 | Acquisition methods, computer equipment and the storage medium of image feature value |
| CN110378201A (en) * | 2019-06-05 | 2019-10-25 | 浙江零跑科技有限公司 | A kind of hinged angle measuring method of multiple row vehicle based on side ring view fisheye camera input |
| CN110739050A (en) * | 2019-12-20 | 2020-01-31 | 深圳大学 | left ventricle full parameter and confidence degree quantification method |
| CN110739050B (en) * | 2019-12-20 | 2020-07-28 | 深圳大学 | Left ventricle full-parameter and confidence coefficient quantification method |
| WO2021258955A1 (en) * | 2020-06-24 | 2021-12-30 | 中兴通讯股份有限公司 | Method and apparatus for marking object outline in target image, and storage medium and electronic apparatus |
| CN113436139A (en) * | 2021-05-10 | 2021-09-24 | 上海大学 | Small intestine nuclear magnetic resonance image identification and physiological information extraction system and method based on deep learning |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113674338A (en) | 2021-11-19 |
| CN109741395B (en) | 2021-07-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109741395A (en) | Biventricular quantization method, device, electronic equipment and storage medium | |
| CN110742653B (en) | Cardiac cycle determination method and ultrasonic equipment | |
| CN110047095A (en) | Tracking, device and terminal device based on target detection | |
| Alessandrini et al. | Simulation of realistic echocardiographic sequences for ground-truth validation of motion estimation | |
| CN108721792B (en) | Method, program memory and system for radiation therapy treatment planning | |
| CN109828302A (en) | A kind of seismic source location method and device based on more vibrating sensors | |
| CN108664860A (en) | The recognition methods of room floor plan and device | |
| CN109102509B (en) | Segmentation model training method and device and computer readable storage medium | |
| US11763468B2 (en) | Structured landmark detection via topology-adapting deep graph learning | |
| CN112164447B (en) | Image processing method, device, equipment and storage medium | |
| CN115511703B (en) | Method, device, equipment and medium for generating two-dimensional heart ultrasonic section image | |
| CN117116451A (en) | Hospital examination queuing method and system based on digital twin technology | |
| CN108597589A (en) | Model generating method, object detection method and medical image system | |
| Rahman et al. | Deep learning-based left ventricular ejection fraction estimation from echocardiographic videos | |
| CN110739050A (en) | left ventricle full parameter and confidence degree quantification method | |
| CN116130090A (en) | Ejection fraction measuring method and device, electronic device, and storage medium | |
| CN110288646A (en) | A kind of human dimension calculation method and device based on image | |
| CN114010227B (en) | Right ventricle characteristic information identification method and device | |
| CN115880358A (en) | Construction method of positioning model, positioning method of image mark points and electronic equipment | |
| EP4036930A1 (en) | Image reconstruction method for dielectric anatomical mapping | |
| EP3270308A1 (en) | Method for providing a secondary parameter, decision support system, computer-readable medium and computer program product | |
| CN112426170A (en) | Placenta thickness determination method, device, equipment and storage medium | |
| CN101926648B (en) | Cardiac wall stress-strain measuring method based on four-dimensional medical image | |
| CN113631941A (en) | Slice alignment for short axis cardiac MR cine playback slice stacking | |
| CN114972242B (en) | Training method and device for myocardial bridge detection model 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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |