CN109447963A - A kind of method and device of brain phantom identification - Google Patents
A kind of method and device of brain phantom identification Download PDFInfo
- Publication number
- CN109447963A CN109447963A CN201811227491.8A CN201811227491A CN109447963A CN 109447963 A CN109447963 A CN 109447963A CN 201811227491 A CN201811227491 A CN 201811227491A CN 109447963 A CN109447963 A CN 109447963A
- Authority
- CN
- China
- Prior art keywords
- image
- brain
- phantom
- characteristic
- identified
- 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
- 210000004556 brain Anatomy 0.000 title claims abstract description 288
- 238000000034 method Methods 0.000 title claims abstract description 40
- 238000005070 sampling Methods 0.000 claims description 89
- 239000000284 extract Substances 0.000 claims description 29
- 238000009826 distribution Methods 0.000 claims description 24
- 206010051290 Central nervous system lesion Diseases 0.000 claims description 22
- 238000012545 processing Methods 0.000 claims description 19
- 238000004590 computer program Methods 0.000 claims description 11
- 238000003860 storage Methods 0.000 claims description 9
- 238000012549 training Methods 0.000 description 15
- 208000006011 Stroke Diseases 0.000 description 13
- 238000010586 diagram Methods 0.000 description 13
- 230000003902 lesion Effects 0.000 description 11
- 238000013527 convolutional neural network Methods 0.000 description 9
- 206010008190 Cerebrovascular accident Diseases 0.000 description 8
- 238000001514 detection method Methods 0.000 description 8
- 238000000605 extraction Methods 0.000 description 7
- 238000007781 pre-processing Methods 0.000 description 7
- 230000002490 cerebral effect Effects 0.000 description 6
- 230000002708 enhancing effect Effects 0.000 description 6
- 230000006870 function Effects 0.000 description 6
- 230000008569 process Effects 0.000 description 6
- 230000004048 modification Effects 0.000 description 5
- 238000012986 modification Methods 0.000 description 5
- 238000004422 calculation algorithm Methods 0.000 description 4
- 238000010606 normalization Methods 0.000 description 4
- 230000004913 activation Effects 0.000 description 3
- 230000015572 biosynthetic process Effects 0.000 description 3
- 239000008280 blood Substances 0.000 description 3
- 210000004369 blood Anatomy 0.000 description 3
- 230000008447 perception Effects 0.000 description 3
- 238000003786 synthesis reaction Methods 0.000 description 3
- 230000001154 acute effect Effects 0.000 description 2
- 210000005013 brain tissue Anatomy 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000003745 diagnosis Methods 0.000 description 2
- 230000000302 ischemic effect Effects 0.000 description 2
- 238000003475 lamination Methods 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- 208000003174 Brain Neoplasms Diseases 0.000 description 1
- 208000014644 Brain disease Diseases 0.000 description 1
- 210000003323 beak Anatomy 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000000740 bleeding effect Effects 0.000 description 1
- 208000026106 cerebrovascular disease Diseases 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000006735 deficit Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 208000035474 group of disease Diseases 0.000 description 1
- 230000002008 hemorrhagic effect Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000000717 retained effect Effects 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 230000001225 therapeutic effect Effects 0.000 description 1
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
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- 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
- G06T2207/30016—Brain
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiology & Medical Imaging (AREA)
- Quality & Reliability (AREA)
- Image Analysis (AREA)
- Magnetic Resonance Imaging Apparatus (AREA)
Abstract
The embodiment of the invention provides a kind of method and devices of brain phantom identification, which comprises obtains the brain phantom of user to be detected；The brain phantom includes N frame brain image；Brain phantom to be identified is input in characteristic extracting module, the characteristic image of the brain phantom to be identified is exported；The characteristic image is to be determined according to various sizes of characteristic image；The brain phantom to be identified includes brain image and the brain image adjacent with the brain image；According to the characteristic image of the brain phantom to be identified, the focal area of the brain image is determined.
Description
Technical field
The present embodiments relate to machine learning techniques fields more particularly to a kind of brain phantom to know method for distinguishing and dress
It sets.
Background technique
" cerebral apoplexy " (cerebral stroke) is also known as " apoplexy ", is a kind of acute cerebrovascular diseases, is due to brain blood
Rostrum so ruptures or causes one group of disease of brain tissue impairment, including ischemic because blood is caused to cannot flow into brain for angiemphraxis
Property and hemorrhagic apoplexy.Statistical data shows that China's incidence of stroke is just risen with annual 8.7% rate at present, brain blood
Pipe disease has become the primary cause of death of Chinese residents.
For patients with cerebral apoplexy, " time window " from morbidity to treatment is most important to the death rate, disability rate is reduced.
And " time window " of Treatment of Stroke is very short, usually starts within morbidity 3 hours or 4.5 hours, it is therefore desirable to which hospital exhausts
Can be shortened intermediate link strives for that therapeutic time, shooting CT image become the best inspection that efficient and economy is taken into account for patient
Means.However cerebral apoplexy is diagnosed in such a way that artificial diagosis judges, spent time too long, is easy to cause state of an illness quilt
Delay, and need veteran doctor when diagnosis, is easy to lead to state of an illness judgement inaccuracy due to artificial difference.
The state of an illness, subjective factor image of the precision by people are mainly judged by the method for manually checking brain phantom in the prior art
Greatly, efficiency is lower.
Summary of the invention
The state of an illness is mainly judged by the method for manually checking brain phantom in the prior art, precision is by artificial subjective factor shadow
As big, the lower problem of efficiency, the embodiment of the invention provides a kind of method and devices of brain phantom identification.
In a first aspect, the embodiment of the invention provides a kind of brain phantoms to know method for distinguishing, comprising:
Obtain the brain phantom of user to be detected；The brain phantom includes N frame brain image；N is positive integer；
Brain phantom to be identified is input in characteristic extracting module, the feature of the brain phantom to be identified is exported
Image；The characteristic image is to be determined according to various sizes of characteristic image；The brain phantom to be identified includes brain
Image and the brain image adjacent with the brain image；
According to the characteristic image of the brain phantom to be identified, the focal area of the brain image is determined.
A kind of possible implementation, the characteristic extracting module successively include M down-sampling convolution block and M up-sampling
Convolution block；
It is described that brain phantom to be identified is input in characteristic extracting module, export the brain phantom to be identified
Characteristic image, comprising:
Using the brain image and the brain image adjacent with the brain image as multichannel, pass sequentially through under M
Sampling convolution block extracts the fisrt feature image of the brain image, the fisrt feature image that each down-sampling convolution block extracts
Size is different, and M is greater than 0；
Convolution block is up-sampled for k-th, the up-sampling convolution block up-samples the K-1 the second spy of convolution output
Feature after sign image merges with the characteristic image that k-th down-sampling convolution block exports, as the up-sampling convolution block input
Image；The size for the second feature image that each up-sampling convolution block extracts is different；M≥K；K is positive integer；
Using the second feature image of m-th up-sampling convolution block output as the characteristic image of the brain image.
A kind of possible implementation, the characteristic image according to the brain phantom to be identified are determined described
The focal area of brain image, comprising:
The characteristic image of the brain phantom is subjected to convolution, determines the probability distribution graph of the brain image；
It is split according to probability distribution graph of the preset threshold to the brain image；
The block of pixels that probability in the probability distribution graph of the brain image is greater than the preset threshold is determined as the brain
The focal area of portion's image.
A kind of possible implementation, the brain image are the first area figure in the brain phantom to be identified
Picture；The method also includes:
Second area image in the brain phantom to be identified is input in the characteristic extracting module, institute is exported
State the fixed reference feature image of brain phantom to be identified；There is symmetrical close in the second area image and the first area image
System；
According to the characteristic image of the brain phantom to be identified, the focal area of the brain image is determined, comprising:
According to the characteristic image and the fixed reference feature image, the brain lesion area in the first area image is determined
Domain.
Second aspect, the embodiment of the invention provides a kind of devices of brain phantom identification, comprising:
Acquiring unit, for obtaining the brain phantom of user to be detected；The brain phantom includes N frame brain image；N is
Positive integer；
Processing unit exports described to be identified for brain phantom to be identified to be input in characteristic extracting module
The characteristic image of brain phantom；The characteristic image is to be determined according to various sizes of characteristic image；The brain to be identified
Portion's image includes brain image and the brain image adjacent with the brain image；According to the spy of the brain phantom to be identified
Image is levied, determines the focal area of the brain image.
A kind of possible implementation, the characteristic extracting module successively include M down-sampling convolution block and M up-sampling
Convolution block；The processing module, is specifically used for:
Using the brain image and the brain image adjacent with the brain image as multichannel, pass sequentially through under M
Sampling convolution block extracts the fisrt feature image of the brain image, the fisrt feature image that each down-sampling convolution block extracts
Size is different, and M is greater than 0；Convolution block is up-sampled for k-th, the up-sampling convolution block is defeated by the K-1 up-sampling convolution
After second feature image out merges with the characteristic image that k-th down-sampling convolution block exports, as the up-sampling convolution block
The characteristic image of input；The size for the second feature image that each up-sampling convolution block extracts is different；M≥K；K is positive integer；
Using the second feature image of m-th up-sampling convolution block output as the characteristic image of the brain image.
A kind of possible implementation, the processing module are specifically used for:
The characteristic image of the brain phantom is subjected to convolution, determines the probability distribution graph of the brain image；According to pre-
If threshold value is split the probability distribution graph of the brain image；Probability in the probability distribution graph of the brain image is greater than
The block of pixels of the preset threshold is determined as the focal area of the brain image.
A kind of possible implementation, the brain image are the first area figure in the brain phantom to be identified
Picture；The processing module, is also used to:
Second area image in the brain phantom to be identified is input in the characteristic extracting module, institute is exported
State the fixed reference feature image of brain phantom to be identified；There is symmetrical close in the second area image and the first area image
System；According to the characteristic image and the fixed reference feature image, the brain lesion region in the first area image is determined.
The third aspect, the embodiment of the invention provides a kind of brain phantom identification equipment, including at least one processor,
And at least one processor, wherein the storage unit is stored with computer program, when described program is held by the processor
When row, so that the step of processor executes first aspect the method.
Fourth aspect, the embodiment of the invention provides a kind of computer-readable medium, being stored with can be known by brain phantom
The computer program that other equipment executes, when described program is run in the equipment that brain phantom identifies, so that the brain
The step of equipment of image identification executes first aspect the method.
In the embodiment of the present invention, lesion identification is carried out by convolutional neural networks model, determines the brain of user to be detected
The image state of an illness judges the state of an illness according to brain phantom without human subjective, on the one hand improve brain phantom identification
On the other hand precision improves the efficiency of brain phantom identification.
Detailed description of the invention
To describe the technical solutions in the embodiments of the present invention more clearly, make required in being described below to embodiment
Attached drawing is briefly introduced, it should be apparent that, drawings in the following description are only some embodiments of the invention, for this
For the those of ordinary skill in field, without any creative labor, it can also be obtained according to these attached drawings
His attached drawing.
Fig. 1 is the flow diagram that a kind of brain phantom provided in an embodiment of the present invention knows method for distinguishing；
Fig. 2 is a kind of schematic diagram of brain phantom provided in an embodiment of the present invention；
Fig. 3 is a kind of schematic diagram of brain phantom recognition methods provided in an embodiment of the present invention；
Fig. 4 is a kind of structural schematic diagram of the device of brain phantom identification provided in an embodiment of the present invention；
Fig. 5 is a kind of structural schematic diagram of the equipment of brain phantom identification provided in an embodiment of the present invention.
Specific embodiment
In order to which the purpose of the present invention, technical solution and beneficial effect is more clearly understood, below in conjunction with attached drawing and implementation
Example, the present invention will be described in further detail.It should be appreciated that specific embodiment described herein is only used to explain this hair
It is bright, it is not intended to limit the present invention.
In the prior art, characteristics of image in brain sample image, at high cost, effect are extracted using traditional machine learning module
Rate is low, can only predict two kinds of stroke types of bleeding or ischemic, can not adapt to the lesion detection of cerebral apoplexy.Either only for MRI
Input picture does brain tumor segmentation using convolutional neural networks, it is impossible to be used in cerebral apoplexy lesion detection.For the acute of CT images
Cerebral apoplexy lesion region is detected due to being relatively strongly dependent upon expertise, currently without relevant algorithm.
Based on the above issues, Fig. 1 illustrates a kind of brain phantom knowledge method for distinguishing provided in an embodiment of the present invention
Flow diagram, the process can by brain phantom identify device execute, specifically includes the following steps:
Step S101 obtains the brain phantom of user to be detected.
Wherein, the brain phantom may include N frame brain image；N is positive integer.N can according to need selection, herein
Without limitation.
Brain phantom refers to the specific image shot using X-ray, for example, CT images.It can be 3-dimensional image, or two
Tie up image.Illustratively, brain phantom can be as shown in the left figure in Fig. 2.
After the brain phantom for obtaining user to be detected, first brain phantom can be pre-processed, preprocessing process master
It may include image normalization.
Image normalization is the following steps are included: be the image of DICOM format by brain phantom, first according to DICOM information, choosing
Fixed window width and window level is taken, for example, window width is W=80, window position is L=40, switchs to the brain phantom image of PNG format.It is near
Few N frame brain image interpolation zooms to fixed size, such as 512*512 pixel.It, can also be by brain figure before scaling
As upside or two sides addition black surround, brain image length-width ratio is adjusted to 1:1, it is not limited here.
Brain phantom to be identified is input in characteristic extracting module by step S102, exports the brain to be identified
The characteristic image of image；The characteristic image is to be determined according to various sizes of characteristic image；The brain shadow to be identified
As including brain image and the brain image adjacent with the brain image；
Wherein, the adjacent brain image can be 1 brain image, or multiple brain images, herein not
It limits.
For the calculation amount for reducing model, the embodiment of the present invention can be using 2 dimension convolutional neural networks, a kind of possible realization
Mode, as shown in figure 3, the characteristic extracting module successively includes M down-sampling convolution block and M up-sampling convolution block；It can be with
Include:
Step S301, using the brain image and the brain image adjacent with the brain image as multichannel, successively
Extract the fisrt feature image of the brain image by M down-sampling convolution block, each down-sampling convolution block extract first
The size of characteristic image is different, and M is greater than 0；
For example, the multichannel can be 3 channels.
Step S302, convolution block is up-sampled for k-th, the up-sampling convolution block is defeated by the K-1 up-sampling convolution
After second feature image out merges with the characteristic image that k-th down-sampling convolution block exports, as the up-sampling convolution block
The characteristic image of input；The size for the second feature image that each up-sampling convolution block extracts is different；M≥K；K is positive integer；
Step S303, using the second feature image of m-th up-sampling convolution block output as the brain shadow to be identified
The characteristic image of picture.
A kind of possible implementation, the characteristic extracting module include 2M convolution module；The 2M convolution module
For down-sampling convolution block and/or up-sampling convolution block；The characteristic image that each down-sampling convolution block or up-sampling convolution block extract
Size it is different, in each convolution module of the 2M convolution module include the first convolutional layer, the second convolutional layer；Described
The number of the characteristic image of one convolutional layer output is less than the number of the characteristic image of first convolutional layer input；The volume Two
The number of the characteristic image of lamination output is greater than the number of the characteristic image of second convolutional layer input；M is greater than 0；
For example, this feature extraction module may include three down-sampling convolution blocks.Each convolution module may include
First convolutional layer and the second convolutional layer, the first convolutional layer include convolutional layer, the normalization (Batch connecting with convolutional layer
Normalization, BN) layer, the activation primitive layer that is connect with BN layers.
For the depth for increasing characteristic extracting module, a kind of possible implementation, step of the characteristic image Jing Guo convolution module
Suddenly may include:
Step 1: the characteristic image that the convolution module inputs is input to first convolutional layer and obtains fisrt feature figure
Picture；The convolution kernel of first convolutional layer can be N1*m*m*N2；N1 is the port number of the characteristic image of convolution module input,
N2 is the port number of fisrt feature image；N1>N2；
Step 2: fisrt feature image is input to second convolutional layer and obtains second feature image；First convolutional layer
Convolution kernel can be N2*m*m*N3；N3 is the port number of second feature image；N3>N2；
Step 3: after the characteristic image that the convolution module is inputted and second feature image merging, it is determined as institute
State the characteristic image of convolution module output.
In a kind of specific embodiment, the number of the characteristic image of the second convolutional layer output can be defeated with the first convolutional layer
The number of the characteristic image entered is equal.That is, N1=N2.
The method of determination of the corresponding characteristic image of brain image as described above is only a kind of possible implementation,
In other possible implementations, the corresponding characteristic image of brain phantom can also be determined otherwise, is not limited specifically
It is fixed.
It should be understood that the activation primitive in the embodiment of the present invention can be a plurality of types of activation primitives, for example, can
Think line rectification function (Rectified Linear Unit, ReLU), specifically without limitation；
Since the image inputted in the embodiment of the present invention is two dimensional image, the feature extraction in the embodiment of the present invention
Module can be the characteristic extracting module in (2Dimensions, 2D) convolutional neural networks, correspondingly, the volume of the first convolutional layer
Product core size can be m*m, the second convolutional layer convolution kernel size can be n*n；M and n can be the same or different, herein
Without limitation；Wherein, m, n are the integer more than or equal to 1.The number of the characteristic image of first convolutional layer output is less than described the
The number of the characteristic image of one convolutional layer input；The number of the characteristic image of the second convolutional layer output is greater than the volume Two
The number of the characteristic image of lamination input.
It further, is optimization characteristic extracting module, a kind of possible implementation, first convolutional layer and described the
It further include third convolutional layer between two convolutional layers；The characteristic image of the third convolutional layer input is first convolutional layer output
Image, the characteristic image of third convolutional layer output is the image of second convolutional layer input.
Wherein, the convolution kernel size of third convolutional layer can be k*k, k and m, and n may be the same or different, herein not
It limits.
In one specific embodiment, the size of the convolution kernel of first convolutional layer is 3*3；Second convolutional layer
The size of convolution kernel is 3*3；The size of the convolution kernel of the third convolutional layer is 1*1.
By the set-up mode of above-mentioned convolution kernel, the perception that can effectively improve feature extraction is wild, is conducive to improve brain
The accuracy of portion's lesion identification.
Various sizes of characteristic image can be the characteristic image of different pixels, such as the pixel characteristic image that is and pixel
For characteristic image be various sizes of characteristic image.
Optionally, the different rulers of brain phantom to be identified are extracted using brain lesion detection model trained in advance
Very little characteristic image, model are true after being trained using 2D convolutional neural networks to the brain phantom of marked multiple users
Fixed.
Optionally, before the various sizes of characteristic image for extracting brain image, brain image is zoomed into specific ruler
It is very little, keep the scale bar of pixel and physical length in all directions certain.
Alternatively possible implementation, the characteristic extracting module include M down-sampling convolution block and M up-sampling volume
Block；The various sizes of characteristic image for obtaining the brain phantom to be identified, comprising:
The brain phantom to be identified is passed sequentially through into M down-sampling convolution block and extracts the M brains to be identified
The fisrt feature image of image；
The fisrt feature image that m-th down-sampling convolution block exports is passed sequentially through into M up-sampling convolution block and extracts M institute
The second feature image of brain phantom to be identified is stated, the size for the second feature image that each up-sampling convolution block extracts is not
Together；
After the identical fisrt feature image of size and second feature image are merged, N number of brain to be identified is determined
The various sizes of characteristic image of image.
For the perception open country for improving feature extraction, the performance of feature extraction, a kind of possible implementation, the feature are improved
It further include feature preprocessing module before extraction module；The feature preprocessing module include a convolutional layer, one BN layers, one
A ReLU layers and a pond layer；The convolution kernel size of the feature preprocessing module is greater than any in N number of convolution module
The size of the convolution kernel of convolution module.
Preferably, the convolution kernel size of the convolutional layer can be 7*7, be divided into 2 pixels.Pond layer be 2*2 most
Big value pond.By feature preprocessing module, image area can be reduced rapidly, side length becomes original 1/4, effective to improve
The perception of characteristic image is wild, quickly extracts shallow-layer feature, the effective loss for reducing raw information.
A kind of possible implementation, the feature preprocessing module include continuous multiple convolutional layers, and one BN layers, one
A ReLU layers and a pond layer；Maximum in the convolution kernel size of the feature preprocessing module and N number of convolution module
Convolution kernel it is equal in magnitude.
Step S103 determines the focal area of the brain phantom according to the characteristic image of the brain phantom.
As shown in Fig. 2 right figure, for the example of the focal area identified provided in the embodiment of the present invention.
It can be seen that the embodiment of the present invention, by characteristic extracting module, and using adjacent brain phantom as multichannel
It is input to characteristic extracting module and carries out feature extraction, effectively raise the accuracy of the identification to brain phantom.It can reduce
Because of Error Diagnostics rate caused by doctor's level difference, to improve the accuracy of stroke lesions diagnosis.
For the accuracy rate for further increasing brain lesion identification, the embodiment of the present invention provides a kind of brain image lesion identification
Method, comprising:
Step 301: obtaining brain image, the brain image is the first area figure in the brain phantom to be identified
Picture；
For example, first area image can be the brain image in left brain area domain；Second area image can be right brain
The brain image in region；
Step 302: the brain image and the brain image adjacent with brain image are input in characteristic extracting module,
Obtain the characteristic image of the brain image；
Step 303: the second area image in the brain phantom to be identified is input to the characteristic extracting module
In, export the fixed reference feature image of the brain phantom to be identified；
Wherein, there are symmetric relations with the first area image for the second area image；
Step 304: according to the characteristic image and the fixed reference feature image, determining the brain in the first area image
Portion focal area；
It similarly, can also be according to the second area image of brain phantom to be identified and the second area image of brain phantom
Reference picture (for example, first area image), determine the brain lesion region in second area image；According to be identified
All brain lesion regions determined in brain phantom determine the brain disease of the brain phantom all areas to be identified
Stove.
By reference to the identification of characteristic image, the recognition accuracy in characteristic image in focal area is further improved,
The interference for avoiding normal brain tissue improves Lesion Detection rate.
Lower mask body introduces the brain phantom by convolutional neural networks to multiple users of marked brain lesion
It is trained determining brain lesion detection model process, comprising the following steps:
Step 1 obtains the brain phantom of multiple users as training sample.Brain phantom herein can be selected
Multiple brain images, or individual brain image, it is not limited here.
Specifically, several brain phantoms that can be will acquire, can also be to several brains of acquisition directly as training sample
Portion's image carries out enhancing operation, expands the data volume of training sample, enhancing operation includes but is not limited to: translating up and down at random
Set pixel (such as 0~20 pixel), Random-Rotation set angle (such as -20~20 degree), random scaling set multiple (such as
0.8~1.2 times).
Step 2, the brain lesion in handmarking's training sample.
The brain lesion in training sample can be marked by professionals such as doctors, the content of label includes brain
The centre coordinate of portion's lesion and the diameter of brain lesion.Specifically, brain lesion can be labeled by several doctors, and
Final brain lesion and brain lesion parameter are determined in such a way that more people vote synthesis, are as a result protected with the mode of mask figure
It deposits.It should be noted that the enhancing operation of handmarking's training sample deutocerebral region lesion and training sample is in no particular order, Ke Yixian
Then brain lesion in handmarking's training sample will mark the training sample of brain lesion to carry out enhancing operation again, can also
Training sample is first carried out enhancing operation, then manually the training sample after enhancing operation is marked.
Training sample input convolutional neural networks are trained, determine brain lesion identification model by step 3.
The structure of the convolutional neural networks includes input layer, down-sampling convolution block, up-sampling convolution block, target detection network
And output layer.Above-mentioned convolutional neural networks are inputted after training sample is pre-processed, by the probability distribution graph of output and in advance
The mask figure of the training sample first marked carries out loss function calculating, then uses back-propagation algorithm and stochastic gradient descent
(Stochastic Gradient Descent, SGD) optimization algorithm iterates, and determines brain lesion detection model.
Further, the difference of brain phantom to be identified is extracted using the brain lesion detection model that above-mentioned training determines
The process of the characteristic image of size, the characteristic extracting module successively include M down-sampling convolution block and M up-sampling convolution
Block；The following steps are included:
Step 1 is successively led to using the brain image and the brain image adjacent with the brain image as multichannel
Cross the fisrt feature image that M down-sampling convolution block extracts the brain image, the first spy that each down-sampling convolution block extracts
The size for levying image is different, and M is greater than 0；
Step 2 up-samples convolution block for k-th, and the up-sampling convolution block exports the K-1 up-sampling convolution
Second feature image merge with the characteristic image that k-th down-sampling convolution block exports after, it is defeated as the up-sampling convolution block
The characteristic image entered；The size for the second feature image that each up-sampling convolution block extracts is different；M≥K；K is positive integer；
Step 3, using the second feature image of m-th up-sampling convolution block output as the brain phantom to be identified
Characteristic image.
In step s 103, a kind of possible implementation, the characteristic pattern according to the brain phantom to be identified
Picture determines the focal area of the brain phantom, comprising:
The characteristic image is subjected to convolution, determines the probability distribution graph of the brain image；
The focal area of the brain image is determined according to the probability distribution graph of the brain image.
A kind of possible implementation, the focal area that the brain image is determined according to the probability distribution graph,
Include:
It is split according to probability distribution graph of the preset threshold to the brain image；
The block of pixels that probability in the probability distribution graph of the brain image is greater than the preset threshold is determined as the brain
The focal area of portion's image.
For example, probability results can be split by preset threshold, the pixel greater than the preset threshold is protected
It stays.Several connection blocks maximum in withed a hook at the end pixel are remained, the result as output focal area.
It can specifically include following steps: the probability distribution graph of whole brain image will be filtered using gaussian filtering first
The result binaryzation of wave, the threshold value of binaryzation is by asking the maximum kind distance method of image grey level histogram to obtain.Then by two
Independent region unit one by one is obtained by " seed filling " method (flood fill) after the result expansion of value, statistics is each
The area of region unit.The maximum region unit of area is retained, the focal area split is determined as.
Alternatively possible implementation can determine the characteristic pattern of brain phantom according to various sizes of characteristic image
Picture.Specifically, the method for determining various sizes of characteristic image may include:
Brain phantom to be identified is passed sequentially through M down-sampling convolution block and extracts n*M brains to be identified by step 1
The fisrt feature image of image.N is the number of the brain image in brain phantom to be identified.
The size for the fisrt feature image that each down-sampling convolution block extracts is different, and M is greater than 0.
Optionally, down-sampling convolution block include the first convolutional layer and the second convolutional layer, group articulamentum, front and back articulamentum, under
Sample level.
The fisrt feature image that m-th down-sampling convolution block exports is passed sequentially through M up-sampling convolution block and mentioned by step 2
Take the second feature image of n*M brain phantom.
The size for the second feature image that each up-sampling convolution block extracts is different.
Optionally, up-sampling convolution block includes that convolutional layer, group articulamentum, front and back articulamentum, up-sampling layer and synthesis connect
Connect layer.Convolutional layer includes convolution algorithm, and BN layers and ReLU layers.
Step 3 determines n*M brain phantom after merging the identical fisrt feature image of size and second feature image
Various sizes of characteristic image.
By up-sampling the synthesis articulamentum in convolution block for the identical fisrt feature image of size and second feature image
Merge and determines various sizes of characteristic image.It optionally, is by the logical of fisrt feature image and second feature image when merging
Road number merges, the size and the size phase of fisrt feature image and second feature image of the characteristic image obtained after merging
Together.
In step s 103, a kind of possible implementation, the characteristic pattern according to the brain phantom to be identified
Picture determines the focal area of the brain phantom, comprising:
The various sizes of characteristic image is subjected to convolution respectively, determines the corresponding probability of various sizes of characteristic image
Distribution map；
The focal area of the brain phantom is determined according to the probability distribution graph.
By various sizes of characteristic image, more information in brain phantom to be identified can be obtained, it can be further
Improve lesion discrimination.
Based on the same technical idea, the embodiment of the invention provides a kind of devices of brain phantom identification, such as Fig. 4 institute
Show, which includes:
Module 401 is obtained, for obtaining the brain phantom of user to be detected；The brain phantom includes N frame brain image；
N is positive integer；
Processing module 402 exports described to be identified for brain phantom to be identified to be input in characteristic extracting module
Brain phantom characteristic image；The characteristic image is to be determined according to various sizes of characteristic image；It is described to be identified
Brain phantom includes brain image and the brain image adjacent with the brain image；According to the brain phantom to be identified
Characteristic image determines the focal area of the brain image.
A kind of possible implementation, the characteristic extracting module successively include M down-sampling convolution block and M up-sampling
Convolution block；The processing module 402, is specifically used for:
Using the brain image and the brain image adjacent with the brain image as multichannel, pass sequentially through under M
Sampling convolution block extracts the fisrt feature image of the brain image, the fisrt feature image that each down-sampling convolution block extracts
Size is different, and M is greater than 0；Convolution block is up-sampled for k-th, the up-sampling convolution block is defeated by the K-1 up-sampling convolution
After second feature image out merges with the characteristic image that k-th down-sampling convolution block exports, as the up-sampling convolution block
The characteristic image of input；The size for the second feature image that each up-sampling convolution block extracts is different；M≥K；K is positive integer；
Using the second feature image of m-th up-sampling convolution block output as the characteristic image of the brain image.
A kind of possible implementation, the processing module 402, is specifically used for:
The characteristic image of the brain phantom is subjected to convolution, determines the probability distribution graph of the brain image；According to pre-
If threshold value is split the probability distribution graph of the brain image；Probability in the probability distribution graph of the brain image is greater than
The block of pixels of the preset threshold is determined as the focal area of the brain image.
A kind of possible implementation, the brain image are the first area figure in the brain phantom to be identified
Picture；The processing module 402, is also used to:
Second area image in the brain phantom to be identified is input in the characteristic extracting module, institute is exported
State the fixed reference feature image of brain phantom to be identified；There is symmetrical close in the second area image and the first area image
System；According to the characteristic image and the fixed reference feature image, the brain lesion region in the first area image is determined.
Based on the same technical idea, the embodiment of the invention provides a kind of equipment of brain phantom identification, such as Fig. 5 institute
Show, including at least one processor 501, and the memory 502 being connect at least one processor, in the embodiment of the present invention not
The specific connection medium between processor 501 and memory 502 is limited, is passed through between processor 501 and memory 502 in Fig. 5
For bus connection.Bus can be divided into address bus, data/address bus, control bus etc..
In embodiments of the present invention, memory 502 is stored with the instruction that can be executed by least one processor 501, at least
The instruction that one processor 501 is stored by executing memory 502 can execute brain phantom above-mentioned and know institute in method for distinguishing
Include the steps that.
Wherein, processor 501 is the control centre of the equipment of brain phantom identification, can use various interfaces and route connects
The various pieces for connecing the equipment of brain phantom identification, by running or executing the instruction and calling that are stored in memory 502
The data being stored in memory 502, to realize that brain phantom identifies.Optionally, processor 501 may include one or more
Processing unit, processor 501 can integrate application processor and modem processor, wherein application processor mainly handles behaviour
Make system, user interface and application program etc., modem processor mainly handles wireless communication.It is understood that above-mentioned
Modem processor can not also be integrated into processor 501.In some embodiments, processor 501 and memory 502 can
To realize on the same chip, in some embodiments, they can also be realized respectively on independent chip.
Processor 501 can be general processor, such as central processing unit (CPU), digital signal processor, dedicated integrated
Circuit (Application Specific Integrated Circuit, ASIC), field programmable gate array or other can
Perhaps transistor logic, discrete hardware components may be implemented or execute present invention implementation for programmed logic device, discrete gate
Each method, step and logic diagram disclosed in example.General processor can be microprocessor or any conventional processor
Deng.The step of method in conjunction with disclosed in the embodiment of the present invention, can be embodied directly in hardware processor and execute completion, Huo Zheyong
Hardware and software module combination in processor execute completion.
Memory 502 is used as a kind of non-volatile computer readable storage medium storing program for executing, can be used for storing non-volatile software journey
Sequence, non-volatile computer executable program and module.Memory 502 may include the storage medium of at least one type,
It such as may include flash memory, hard disk, multimedia card, card-type memory, random access storage device (Random Access
Memory, RAM), static random-access memory (Static Random Access Memory, SRAM), may be programmed read-only deposit
Reservoir (Programmable Read Only Memory, PROM), read-only memory (Read Only Memory, ROM), band
Electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory,
EEPROM), magnetic storage, disk, CD etc..Memory 502 can be used for carrying or storing have instruction or data
The desired program code of structure type and can by any other medium of computer access, but not limited to this.The present invention is real
Applying the memory 502 in example can also be circuit or other devices that arbitrarily can be realized store function, for storing program
Instruction and/or data.
Based on the same inventive concept, the embodiment of the invention also provides a kind of computer-readable medium, being stored with can be by
The computer program that the equipment of brain phantom identification executes makes when described program is run in the equipment that brain phantom identifies
The equipment for obtaining the brain phantom identification executes the step of brain phantom knows method for distinguishing.
It should be understood by those skilled in the art that, the embodiment of the present invention can provide as method or computer program product.
Therefore, complete hardware embodiment, complete software embodiment or embodiment combining software and hardware aspects can be used in the present invention
Form.It is deposited moreover, the present invention can be used to can be used in the computer that one or more wherein includes computer usable program code
The shape for the computer program product implemented on storage media (including but not limited to magnetic disk storage, CD-ROM, optical memory etc.)
Formula.
The present invention be referring to according to the method for the embodiment of the present invention, the process of equipment (system) and computer program product
Figure and/or block diagram describe.It should be understood that every one stream in flowchart and/or the block diagram can be realized by computer program instructions
The combination of process and/or box in journey and/or box and flowchart and/or the block diagram.It can provide these computer programs
Instruct the processor of general purpose computer, special purpose computer, Embedded Processor or other programmable data processing devices to produce
A raw machine, so that being generated by the instruction that computer or the processor of other programmable data processing devices execute for real
The device for the function of being specified in present one or more flows of the flowchart and/or one or more blocks of the block diagram.
These computer program instructions, which may also be stored in, is able to guide computer or other programmable data processing devices with spy
Determine in the computer-readable memory that mode works, so that it includes referring to that instruction stored in the computer readable memory, which generates,
Enable the manufacture of device, the command device realize in one box of one or more flows of the flowchart and/or block diagram or
The function of being specified in multiple boxes.
These computer program instructions also can be loaded onto a computer or other programmable data processing device, so that counting
Series of operation steps are executed on calculation machine or other programmable devices to generate computer implemented processing, thus in computer or
The instruction executed on other programmable devices is provided for realizing in one or more flows of the flowchart and/or block diagram one
The step of function of being specified in a box or multiple boxes.
Although preferred embodiments of the present invention have been described, it is created once a person skilled in the art knows basic
Property concept, then additional changes and modifications may be made to these embodiments.So it includes excellent that the following claims are intended to be interpreted as
It selects embodiment and falls into all change and modification of the scope of the invention.
Obviously, various changes and modifications can be made to the invention without departing from essence of the invention by those skilled in the art
Mind and range.In this way, if these modifications and changes of the present invention belongs to the range of the claims in the present invention and its equivalent technologies
Within, then the present invention is also intended to include these modifications and variations.
Claims (10)
1. a kind of brain phantom knows method for distinguishing characterized by comprising
Obtain the brain phantom of user to be detected；The brain phantom includes N frame brain image；N is positive integer；
Brain phantom to be identified is input in characteristic extracting module, the characteristic pattern of the brain phantom to be identified is exported
Picture；The characteristic image is to be determined according to various sizes of characteristic image；The brain phantom to be identified includes brain figure
Picture and the brain image adjacent with the brain image；
According to the characteristic image of the brain phantom to be identified, the focal area of the brain image is determined.
2. the method as described in claim 1, which is characterized in that the characteristic extracting module successively includes M down-sampling convolution
Block and M up-sampling convolution block；
It is described that brain phantom to be identified is input in characteristic extracting module, export the feature of the brain phantom to be identified
Image, comprising:
Using the brain image and the brain image adjacent with the brain image as multichannel, M down-sampling is passed sequentially through
Convolution block extracts the fisrt feature image of the brain image, the size for the fisrt feature image that each down-sampling convolution block extracts
Different, M is greater than 0；
Convolution block is up-sampled for k-th, the up-sampling convolution block up-samples the K-1 in the second feature figure of convolution output
Characteristic image after merging with the characteristic image of k-th down-sampling convolution block output, as the up-sampling convolution block input；
The size for the second feature image that each up-sampling convolution block extracts is different；M≥K；K is positive integer；
Using the second feature image of m-th up-sampling convolution block output as the characteristic image of the brain phantom to be identified.
3. the method as described in claim 1, which is characterized in that the characteristic pattern according to the brain phantom to be identified
Picture determines the focal area of the brain image, comprising:
The characteristic image of the brain phantom is subjected to convolution, determines the probability distribution graph of the brain image；
It is split according to probability distribution graph of the preset threshold to the brain image；
The block of pixels that probability in the probability distribution graph of the brain image is greater than the preset threshold is determined as the brain figure
The focal area of picture.
4. the method as described in claim 1, which is characterized in that the brain image is in the brain phantom to be identified
First area image；The method also includes:
Second area image in the brain phantom to be identified is input in the characteristic extracting module, output it is described to
The fixed reference feature image of the brain phantom of identification；There are symmetric relations with the first area image for the second area image；
According to the characteristic image of the brain phantom to be identified, the focal area of the brain image is determined, comprising:
According to the characteristic image and the fixed reference feature image, the brain lesion region in the first area image is determined.
5. a kind of device of brain phantom identification characterized by comprising
Acquiring unit, for obtaining the brain phantom of user to be detected；The brain phantom includes N frame brain image；N is positive whole
Number；
Processing unit exports the brain to be identified for brain phantom to be identified to be input in characteristic extracting module
The characteristic image of image；The characteristic image is to be determined according to various sizes of characteristic image；The brain shadow to be identified
As including brain image and the brain image adjacent with the brain image；According to the characteristic pattern of the brain phantom to be identified
Picture determines the focal area of the brain image.
6. device as claimed in claim 5, which is characterized in that the characteristic extracting module successively includes M down-sampling convolution
Block and M up-sampling convolution block；The processing module, is specifically used for:
Using the brain image and the brain image adjacent with the brain image as multichannel, M down-sampling is passed sequentially through
Convolution block extracts the fisrt feature image of the brain image, the size for the fisrt feature image that each down-sampling convolution block extracts
Different, M is greater than 0；Convolution block is up-sampled for k-th, the up-sampling convolution block up-samples convolution output for the K-1
After second feature image merges with the characteristic image that k-th down-sampling convolution block exports, inputted as the up-sampling convolution block
Characteristic image；The size for the second feature image that each up-sampling convolution block extracts is different；M≥K；K is positive integer；By
Characteristic image of the second feature image of M up-sampling convolution block output as the brain image.
7. device as claimed in claim 5, which is characterized in that the processing module is specifically used for:
The characteristic image of the brain phantom is subjected to convolution, determines the probability distribution graph of the brain image；According to default threshold
Value is split the probability distribution graph of the brain image；Probability in the probability distribution graph of the brain image is greater than described
The block of pixels of preset threshold is determined as the focal area of the brain image.
8. device as claimed in claim 5, which is characterized in that the brain image is in the brain phantom to be identified
First area image；The processing module, is also used to:
Second area image in the brain phantom to be identified is input in the characteristic extracting module, output it is described to
The fixed reference feature image of the brain phantom of identification；There are symmetric relations with the first area image for the second area image；
According to the characteristic image and the fixed reference feature image, the brain lesion region in the first area image is determined.
9. a kind of equipment of brain phantom identification, which is characterized in that including at least one processor and at least one storage
Device, wherein the storage unit is stored with computer program, when described program is executed by the processor, so that the place
Manage the step of device perform claim requires 1~4 any claim the method.
10. a kind of computer-readable medium, which is characterized in that it is stored with the calculating that the equipment that can be identified by brain phantom executes
Machine program, when described program is run in the equipment that brain phantom identifies, so that the equipment of brain phantom identification executes
The step of Claims 1 to 4 any the method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811227491.8A CN109447963B (en) | 2018-10-22 | 2018-10-22 | Brain image recognition method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811227491.8A CN109447963B (en) | 2018-10-22 | 2018-10-22 | Brain image recognition method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109447963A true CN109447963A (en) | 2019-03-08 |
| CN109447963B CN109447963B (en) | 2024-09-13 |
Family
ID=65546831
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811227491.8A Active CN109447963B (en) | 2018-10-22 | 2018-10-22 | Brain image recognition method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109447963B (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110533639A (en) * | 2019-08-02 | 2019-12-03 | 杭州依图医疗技术有限公司 | A kind of key independent positioning method and device |
| CN110782446A (en) * | 2019-10-25 | 2020-02-11 | 杭州依图医疗技术有限公司 | Method and device for determining volume of lung nodule |
| CN111242926A (en) * | 2020-01-13 | 2020-06-05 | 北京妙医佳健康科技集团有限公司 | Focus detection method and device and electronic equipment |
| CN111275558A (en) * | 2020-01-13 | 2020-06-12 | 上海维跃信息科技有限公司 | Method and device for determining insurance data |
| CN113076893A (en) * | 2021-04-09 | 2021-07-06 | 太原理工大学 | Highway drain pipe blocking situation sensing method based on deep learning |
| CN113269711A (en) * | 2021-04-06 | 2021-08-17 | 东北大学 | Method and device for processing brain image, electronic equipment and storage medium |
| CN113780421A (en) * | 2021-06-07 | 2021-12-10 | 广州天鹏计算机科技有限公司 | Brain PET image identification method based on artificial intelligence |
| CN116051745A (en) * | 2023-01-10 | 2023-05-02 | 北京柏惠维康科技股份有限公司 | Cerebral vascular model construction method and device, storage medium and electronic equipment |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050020945A1 (en) * | 2002-07-02 | 2005-01-27 | Tosaya Carol A. | Acoustically-aided cerebrospinal-fluid manipulation for neurodegenerative disease therapy |
| CN1628611A (en) * | 2003-10-29 | 2005-06-22 | 株式会社东芝 | Apparatus for aiding diagnosis and treatment of cerebral disease and x ray computer tomography apparatus |
| CN107665491A (en) * | 2017-10-10 | 2018-02-06 | 清华大学 | The recognition methods of pathological image and system |
| CN108648192A (en) * | 2018-05-17 | 2018-10-12 | 杭州依图医疗技术有限公司 | A kind of method and device of detection tubercle |
-
2018
- 2018-10-22 CN CN201811227491.8A patent/CN109447963B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050020945A1 (en) * | 2002-07-02 | 2005-01-27 | Tosaya Carol A. | Acoustically-aided cerebrospinal-fluid manipulation for neurodegenerative disease therapy |
| CN1628611A (en) * | 2003-10-29 | 2005-06-22 | 株式会社东芝 | Apparatus for aiding diagnosis and treatment of cerebral disease and x ray computer tomography apparatus |
| CN107665491A (en) * | 2017-10-10 | 2018-02-06 | 清华大学 | The recognition methods of pathological image and system |
| CN108648192A (en) * | 2018-05-17 | 2018-10-12 | 杭州依图医疗技术有限公司 | A kind of method and device of detection tubercle |
Non-Patent Citations (1)
| Title |
|---|
| 王腾: ""早期微小病变识别系统的研究和实现"", 《中国优秀博硕士论文全文数据库（硕士） 信息科技缉》 * |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110533639A (en) * | 2019-08-02 | 2019-12-03 | 杭州依图医疗技术有限公司 | A kind of key independent positioning method and device |
| CN110782446A (en) * | 2019-10-25 | 2020-02-11 | 杭州依图医疗技术有限公司 | Method and device for determining volume of lung nodule |
| CN110782446B (en) * | 2019-10-25 | 2022-04-15 | 杭州依图医疗技术有限公司 | Method and device for determining volume of lung nodule |
| CN111242926A (en) * | 2020-01-13 | 2020-06-05 | 北京妙医佳健康科技集团有限公司 | Focus detection method and device and electronic equipment |
| CN111275558A (en) * | 2020-01-13 | 2020-06-12 | 上海维跃信息科技有限公司 | Method and device for determining insurance data |
| CN111275558B (en) * | 2020-01-13 | 2024-02-27 | 上海维跃信息科技有限公司 | Method and device for determining insurance data |
| CN113269711A (en) * | 2021-04-06 | 2021-08-17 | 东北大学 | Method and device for processing brain image, electronic equipment and storage medium |
| CN113269711B (en) * | 2021-04-06 | 2024-01-09 | 东北大学 | Brain image processing method and device, electronic equipment and storage medium |
| CN113076893B (en) * | 2021-04-09 | 2022-08-02 | 太原理工大学 | Highway drain pipe blocking situation sensing method based on deep learning |
| CN113076893A (en) * | 2021-04-09 | 2021-07-06 | 太原理工大学 | Highway drain pipe blocking situation sensing method based on deep learning |
| CN113780421B (en) * | 2021-06-07 | 2022-06-07 | 广州天鹏计算机科技有限公司 | Brain PET image identification method based on artificial intelligence |
| CN113780421A (en) * | 2021-06-07 | 2021-12-10 | 广州天鹏计算机科技有限公司 | Brain PET image identification method based on artificial intelligence |
| CN116051745A (en) * | 2023-01-10 | 2023-05-02 | 北京柏惠维康科技股份有限公司 | Cerebral vascular model construction method and device, storage medium and electronic equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109447963B (en) | 2024-09-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109447963A (en) | A kind of method and device of brain phantom identification | |
| CN111798462B (en) | Automatic delineation method of nasopharyngeal carcinoma radiotherapy target area based on CT image | |
| CN110934606B (en) | Cerebral apoplexy early-stage flat-scan CT image evaluation system and method and readable storage medium | |
| CN107203999B (en) | Dermatoscope image automatic segmentation method based on full convolution neural network | |
| CN109063710B (en) | 3D CNN nasopharyngeal carcinoma segmentation method based on multi-scale feature pyramid | |
| CN110969626B (en) | Method for extracting hippocampus of human brain nuclear magnetic resonance image based on 3D neural network | |
| CN109389584A (en) | Multiple dimensioned rhinopharyngeal neoplasm dividing method based on CNN | |
| CN109886986A (en) | A kind of skin lens image dividing method based on multiple-limb convolutional neural networks | |
| CN113808146B (en) | Multi-organ segmentation method and system for medical image | |
| CN115496771A (en) | Brain tumor segmentation method based on brain three-dimensional MRI image design | |
| CN110910408A (en) | Image segmentation method and device, electronic equipment and readable storage medium | |
| CN110533029A (en) | Determine the method and device of target area in image | |
| CN109523546A (en) | A kind of method and device of Lung neoplasm analysis | |
| CN113781488A (en) | Tongue picture image segmentation method, apparatus and medium | |
| CN116452618A (en) | Three-input spine CT image segmentation method | |
| CN113269799A (en) | Cervical cell segmentation method based on deep learning | |
| CN113764101B (en) | Novel auxiliary chemotherapy multi-mode ultrasonic diagnosis system for breast cancer based on CNN | |
| CN110992310A (en) | Method and device for determining partition where mediastinal lymph node is located | |
| CN115311193A (en) | Abnormal brain image segmentation method and system based on double attention mechanism | |
| CN109711315A (en) | A kind of method and device of Lung neoplasm analysis | |
| CN112990367A (en) | Image processing method, device, equipment and storage medium | |
| CN116740081A (en) | Method, device, terminal equipment and medium for segmenting pulmonary vessels in CT image | |
| CN113379770B (en) | Construction method of nasopharyngeal carcinoma MR image segmentation network, image segmentation method and device | |
| CN112102327A (en) | Image processing method and device and computer readable storage medium | |
| CN113129297B (en) | Diameter automatic measurement method and system based on multi-phase tumor image |
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 |