CN110874614B - Brain image classification method, computer device, and readable storage medium - Google Patents
Brain image classification method, computer device, and readable storage medium Download PDFInfo
- Publication number
- CN110874614B CN110874614B CN201911105414.XA CN201911105414A CN110874614B CN 110874614 B CN110874614 B CN 110874614B CN 201911105414 A CN201911105414 A CN 201911105414A CN 110874614 B CN110874614 B CN 110874614B
- Authority
- CN
- China
- Prior art keywords
- spherical
- brain image
- map
- cerebral cortex
- sample
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/03—Recognition of patterns in medical or anatomical images
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Magnetic Resonance Imaging Apparatus (AREA)
- Image Analysis (AREA)
- Apparatus For Radiation Diagnosis (AREA)
Abstract
The present invention relates to a brain image classification method, a computer device and a readable storage medium, the method comprising: acquiring brain images to be classified; inputting the brain image into a spherical mapping model to obtain a cerebral cortex spherical mapping diagram corresponding to the brain image; and inputting the cerebral cortex spherical map into a spherical convolution network to obtain a classification result of the brain image. According to the method, the computer equipment can accurately map the characteristics of the brain image to be classified into the cerebral cortex spherical map through the spherical map model, the accuracy of the cerebral cortex spherical map corresponding to the obtained brain image to be classified is improved, and the structural change of the cerebral cortex of the brain image to be classified can be accurately displayed due to the obtained cerebral cortex spherical map, so that the characteristics of the cerebral cortex of the brain image to be classified can be more accurately extracted by the spherical convolution network, the brain image to be classified is more accurately classified, and the classification result of the obtained brain image to be classified is improved.
Description
Technical Field
The present invention relates to the field of medical images, and in particular, to a brain image classification method, a computer device, and a readable storage medium.
Background
Neurological and psychiatric disorders may cause alterations in brain structure, for example brain atrophy, a phenomenon in which brain tissue itself undergoes organic lesions resulting in atrophy from various causes, which is manifested by reduced brain tissue volume, reduced cell numbers, and enlarged ventricles and subarachnoid spaces. Atrophy of cerebral cortex, cerebellum, brainstem and the like is one of neuroimaging manifestations of many diseases such as Alzheimer's Disease (AD), mild cognitive impairment (Mild cognitive impairment, MCI), parkinsonism (PD) and the like, and magnetic resonance imaging (Magnetic Resonance Imaging, MRI) can be used for diagnosing AD and related diseases due to the characteristics of non-radioactivity, high imaging quality of brain structures, and visible atrophy of cortex, cerebellum, brainstem and the like.
In the prior art, an MRI image is directly unfolded into a plane image through a conformal mapping based on a neural network of depth convolution, and then the unfolded plane image is subjected to convolution operation, so that characteristic representations of the MRI image at different levels are obtained, corresponding MRI images are classified according to the characteristics of the MRI image at different levels, and auxiliary diagnosis is carried out on diseases.
However, the conventional method of classifying MRI images has a problem in that the classification result is inaccurate.
Disclosure of Invention
Based on this, it is necessary to provide a brain image classification method, a computer device and a readable storage medium for solving the problem that the classification result is inaccurate in the conventional method of classifying MRI images.
In a first aspect, an embodiment of the present invention provides a brain image classification method, including:
acquiring brain images to be classified;
inputting the brain image into a spherical mapping model to obtain a cerebral cortex spherical mapping diagram corresponding to the brain image;
and inputting the cerebral cortex spherical map into a spherical convolution network to obtain a classification result of the brain image.
In one embodiment, the inputting the brain image into a spherical mapping model to obtain a cerebral cortex spherical mapping map corresponding to the brain image includes:
inputting the brain image into the spherical mapping model to obtain a cerebral cortex spherical expansion diagram corresponding to the brain image;
and mapping the cerebral cortex spherical surface unfolded map to a preset standard spherical surface to obtain a cerebral cortex spherical surface mapping map corresponding to the cerebral image.
In one embodiment, the inputting the brain image into a spherical mapping model to obtain a cerebral cortex spherical expansion map corresponding to the brain image includes:
inputting the brain image into the spherical mapping model, and expanding the brain image according to a preset expansion rule to obtain a cerebral cortex spherical expansion diagram corresponding to the brain image.
In one embodiment, the network structure of the spherical convolution network is any one of the following: a three-dimensional dense connection network structure; a spherical residual error network structure; the three-dimensional dense connection network incorporates a feature weighting module network structure.
In one embodiment, the sphere mapping model comprises a sphere mapping algorithm or sphere mapping network.
In one embodiment, when the spherical mapping model is a spherical mapping network, the training process of the spherical mapping model includes:
acquiring a sample brain image;
inputting the sample brain image into a preset spherical mapping network to obtain a sample cerebral cortex spherical mapping diagram corresponding to the sample brain image;
training the preset spherical mapping network according to the sample cerebral cortex spherical mapping diagram and a preset reference cerebral cortex spherical mapping diagram to obtain the spherical mapping network.
In one embodiment, before training the preset spherical mapping network according to the sample cortex spherical mapping map and the preset reference cortex spherical mapping map to obtain the spherical mapping network, the method further includes:
and processing the sample brain image by using a preset brain image segmentation tool to obtain the preset reference brain cortex spherical map.
In one embodiment, the training process of the spherical convolution network includes:
acquiring a sample cerebral cortex spherical map;
inputting the sample cerebral cortex spherical map into a preset spherical convolution network to obtain a sample classification result of a sample cerebral image corresponding to the sample cerebral cortex spherical map;
and training the preset spherical convolution network according to the sample classification result and a label of a sample brain image mark corresponding to the sample cerebral cortex spherical map in advance to obtain the spherical convolution network.
In a second aspect, an embodiment of the present invention provides a brain image classification device, including:
the first acquisition module is used for acquiring brain images to be classified;
the second acquisition module is used for inputting the brain image into a spherical mapping model to obtain a cerebral cortex spherical mapping diagram corresponding to the brain image;
and the classification module is used for inputting the cerebral cortex spherical map into a spherical convolution network to obtain a classification result of the brain image.
In a third aspect, an embodiment of the present invention provides a computer device, including a memory and a processor, the memory storing a computer program, the processor implementing the following steps when executing the computer program:
acquiring brain images to be classified;
inputting the brain image into a spherical mapping model to obtain a cerebral cortex spherical mapping diagram corresponding to the brain image;
and inputting the cerebral cortex spherical map into a spherical convolution network to obtain a classification result of the brain image.
In a fourth aspect, embodiments of the present invention provide a computer readable storage medium having stored thereon a computer program which when executed by a processor performs the steps of:
acquiring brain images to be classified;
inputting the brain image into a spherical mapping model to obtain a cerebral cortex spherical mapping diagram corresponding to the brain image;
and inputting the cerebral cortex spherical map into a spherical convolution network to obtain a classification result of the brain image.
In the brain image classification method, the device, the computer equipment and the readable storage medium provided in the above embodiments, the computer equipment acquires a brain image to be classified, inputs the brain image into a spherical mapping model to obtain a cerebral cortex spherical mapping diagram corresponding to the brain image, and inputs the cerebral cortex spherical mapping diagram into a spherical convolution network to obtain a classification result of the brain image. In the method, the computer equipment can accurately map the characteristics of the brain image to be classified into the cerebral cortex spherical map through the spherical map model, so that the accuracy of the cerebral cortex spherical map corresponding to the obtained brain image to be classified is improved, and the structural change of the cerebral cortex of the brain image to be classified can be accurately displayed due to the obtained cerebral cortex spherical map, so that the characteristics of the cerebral cortex of the brain image to be classified can be more accurately extracted by the spherical convolution network, the brain image to be classified is more accurately classified, and the classification result of the obtained brain image to be classified is improved.
Drawings
FIG. 1 is a schematic diagram of an internal structure of a computer device according to one embodiment;
FIG. 2 is a flow chart of a brain image classification method according to an embodiment;
FIG. 3 is a flow chart of a classification result of obtaining brain images to be classified according to an embodiment;
FIG. 4 is a flow chart of a brain image classification method according to another embodiment;
FIG. 5 is a flow chart of a brain image classification method according to another embodiment;
FIG. 6 is a flow chart of a brain image classification method according to another embodiment;
fig. 7 is a schematic structural diagram of a brain image classification device according to an embodiment.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application will be further described in detail with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the present application.
The brain image classification method provided by the embodiment of the application can be applied to the computer equipment shown in the figure 1. The computer device comprises a processor, a memory, and a computer program stored in the memory, wherein the processor is connected through a system bus, and when executing the computer program, the processor can execute the steps of the method embodiments described below. Optionally, the computer device may further comprise a network interface, a display screen and an input means. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium, which stores an operating system and a computer program, an internal memory. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The network interface of the computer device is used for communicating with an external terminal through a network connection. Optionally, the computer device may be a server, may be a personal computer, may also be a personal digital assistant, may also be other terminal devices, such as a tablet computer, a mobile phone, etc., and may also be a cloud or remote server.
Taking Alzheimer's Disease (AD) as an example, AD is one of the most common neurodegenerative diseases, and because magnetic resonance imaging (Magnetic Resonance Imaging, MRI) has the characteristics of no radioactivity, high imaging quality for brain structures, etc., it is widely applied to diagnosis of AD and related conditions, which makes the need for computer-aided detection (Computer Aided Diagnosis, CAD) urgent. The computer-aided diagnosis of the AD can effectively screen AD patients and early-stage mild cognitive impairment (Mild cognitive impairment, MCI) stages thereof according to medical images such as MRI, so that the workload of doctors is greatly reduced, and meanwhile, the detection accuracy of the doctors can be improved. The main problems faced by the existing CAD are low automation degree, low detection precision and low speed, the CAD is usually processed by using a manually designed feature extractor to extract the features, a lot of useful information is lost in the process, and the learning ability of the model is greatly limited. The depth full convolution neural network based on three dimensions allows processing all information including the whole lesion area and surrounding structures, and the correlation among different layers is considered, so that the representation of multiple abstract levels of data is automatically acquired, useful characteristics can be amplified, and the neuropsychiatric diseases can be diagnosed efficiently and accurately.
The following describes the technical scheme of the present invention and how the technical scheme of the present invention solves the above technical problems in detail with specific embodiments. The following embodiments may be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments.
Fig. 2 is a flow chart of a brain image classification method according to an embodiment. Fig. 3 is a flowchart of a classification result of obtaining a brain image to be classified according to an embodiment. The embodiment relates to a specific implementation process of inputting a brain image to be classified into a spherical mapping model by computer equipment to obtain a cerebral cortex spherical mapping diagram corresponding to the brain image to be classified, inputting the obtained cerebral cortex spherical mapping diagram into a spherical convolution network to obtain a classification result of the brain image to be classified. As shown in fig. 2, the method may include:
s201, acquiring brain images to be classified.
Wherein the brain image to be classified is a medical image of the brain of the patient acquired by the medical imaging device. Alternatively, the brain image to be classified may be a computed tomography (Computed Tomography, CT) image of the brain, such as a T1-weighted image or a T2-weighted image, but also an MRI image of the brain, or other medical images of the brain. Alternatively, the computer device may obtain the brain image to be classified from a PACS (Picture Archiving and Communication Systems, image archiving and communication system) server, or may obtain the brain image to be classified from the medical imaging device in real time. Optionally, after the computer device obtains the brain image to be classified, preprocessing can be performed on the brain image to be classified to remove the influence of the working condition of the medical imaging device on the obtained image quality, and high-quality images are provided for subsequent analysis, wherein the preprocessing comprises resampling processing, size adjustment processing, cephalic bone removal processing, image non-uniformity correction processing, histogram matching processing, gray scale normalization processing and the like. It should be noted that the number of the substrates, brain to be classified after pretreatment the image being the image size 256×256×256mm 3 The directions are all standard Cartesian LPI coordinate systems, and the gray scale ranges are all standard images in the (-1, 1) interval.
S202, inputting the brain image into a spherical mapping model to obtain a cerebral cortex spherical mapping diagram corresponding to the brain image.
Specifically, the computer equipment inputs the acquired brain image to be classified into a spherical mapping model to obtain a cerebral cortex spherical mapping diagram corresponding to the brain image. The cerebral cortex spherical map corresponding to the brain image to be classified contains the sulcus cerebral return information of the brain image to be classified.
S203, inputting the cerebral cortex spherical map into a spherical convolution network to obtain a classification result of the brain image.
Specifically, the computer equipment inputs the obtained cerebral cortex spherical map into a spherical convolution network to obtain a classification result of the brain image to be classified. Optionally, the network structure of the spherical convolution network may be any one of a three-dimensional dense connection network structure, a spherical residual network structure and a three-dimensional dense connection network combined with a characteristic weighting module network structure. Optionally, the obtained classification result of the brain image to be classified may include AD, stable mild cognitive impairment (mci), progressive mild cognitive impairment (pMCI) and normal, and it should be noted that the obtained classification result of the brain image to be classified further includes a likelihood prediction index for each classification category, for example, the obtained classification result of a brain image to be classified may be: normal: 5%, sMCI:7%, pMCI:20%, AD:68%.
In this embodiment, the computer device can accurately map the features of the brain image to be classified into the cortex spherical map through the spherical map model, so that the accuracy of the cortex spherical map corresponding to the obtained brain image to be classified is improved.
Fig. 4 is a flowchart of a brain image classification method according to another embodiment. The embodiment relates to a specific implementation process of inputting brain images to be classified into a spherical mapping model by computer equipment to obtain a cerebral cortex spherical mapping diagram corresponding to the brain images to be classified. As shown in fig. 4, based on the above embodiment, as an alternative implementation manner, the step S202 includes:
s401, inputting the brain image into a spherical mapping model to obtain a cerebral cortex spherical expansion diagram corresponding to the brain image.
Specifically, the computer equipment inputs the brain image to be classified into a spherical mapping model to obtain a cerebral cortex spherical expansion diagram corresponding to the brain image to be classified. Optionally, the computer device may input the brain image to be classified into a spherical mapping model, and expand the brain image to be classified according to a preset expansion rule, so as to obtain a cerebral cortex spherical expansion map corresponding to the brain image to be classified. Optionally, the sphere mapping model comprises a sphere mapping algorithm or a sphere mapping network. Alternatively, the preset expansion rule may be an expansion rule used in a freesurfer segmentation tool.
And S402, mapping the cerebral cortex spherical surface unfolded map to a preset standard spherical surface to obtain a cerebral cortex spherical surface mapped map corresponding to the cerebral image.
Specifically, after obtaining a cerebral cortex spherical expansion map corresponding to a cerebral image to be classified, the computer equipment maps the obtained cerebral cortex spherical expansion map onto a preset standard spherical surface according to the inverse process of the preset expansion rule, so as to obtain a cerebral cortex spherical map corresponding to the cerebral image to be classified. It should be noted that, the process of mapping the cerebral cortex sphere expansion map to the preset standard sphere by the computer device is a process of mapping the sulcus brain return information of the brain image to be classified to the preset standard sphere.
In this embodiment, the computer device inputs the brain image to be classified into the spherical mapping model to obtain a cerebral cortex spherical expansion map corresponding to the brain image to be classified, maps the obtained cerebral cortex spherical expansion map corresponding to the brain image to be classified onto a preset standard spherical surface, so as to obtain a cerebral cortex spherical mapping map corresponding to the brain image to be classified, and because the cerebral cortex surface can be regarded as a highly folded curved surface, the brain image to be classified can be accurately projected onto a standard spherical surface through the process, and the obtained cerebral cortex spherical mapping map corresponding to the brain image to be classified can accurately reflect cerebral cortex information of the brain image to be classified, thereby improving accuracy of the obtained cerebral cortex spherical mapping map corresponding to the brain image to be classified.
Fig. 5 is a flowchart of a brain image classification method according to another embodiment. The embodiment relates to a specific implementation process of training the spherical mapping model by the computer equipment when the spherical mapping model is a spherical mapping network. As shown in fig. 5, based on the above embodiment, as an alternative implementation manner, the training process of the spherical mapping model may include:
s501, acquiring a sample brain image.
Wherein the sample brain image is a sample medical image of the brain of the patient acquired by the medical imaging device. Alternatively, the sample brain image may be a computed tomography (Computed Tomography, CT) image of the brain, such as a T1 weighted image or a T2 weighted image, or may be an MRI image of the brain, or other medical image of the brain. Alternatively, the computer device may obtain the sample brain image from a PACS (Picture Archiving and Communication Systems, image archiving and communication system) server, or may obtain the sample brain image from the medical imaging device in real time. Optionally, after the computer device obtains the sample brain image, the sample brain image may be preprocessed to remove the influence of the working condition of the medical image device on the obtained image quality, and provide a high-quality image for subsequent analysis, where the preprocessing includes resampling, resizing, removing the skull, image non-uniformity correction, histogram matching, and gray scale normalization. It should be noted that the number of the substrates, pretreated sample brain image for an image of size of 256×256×256mm 3 The directions are all standard Cartesian LPI coordinate systems, and the gray scale ranges are all standard images in the (-1, 1) interval.
S502, inputting a sample brain image into a preset spherical mapping network to obtain a sample cerebral cortex spherical mapping diagram corresponding to the sample brain image;
specifically, the computer equipment inputs a sample brain image into a preset spherical mapping network to obtain a sample cerebral cortex spherical expansion map corresponding to the sample brain image, and maps the obtained sample cerebral cortex spherical expansion map onto a preset standard spherical surface to obtain a sample cerebral cortex spherical mapping map corresponding to the sample brain image. Alternatively, the network structure of the preset spherical mapping network may be a V-Net network structure.
S503, training a preset spherical mapping network according to the sample cerebral cortex spherical mapping diagram and a preset reference cerebral cortex spherical mapping diagram to obtain the spherical mapping network.
Specifically, after obtaining a sample cortex spherical map corresponding to a sample brain image, the computer equipment obtains a value of a loss function of a preset spherical map network according to the sample cortex spherical map and a preset reference cortex spherical map, trains the preset spherical map network according to the value of the loss function of the preset spherical map network until the value of the loss function of the preset spherical map network reaches a stable value, and determines the corresponding preset spherical map network as the spherical map network.
In this embodiment, the computer device obtains the sample cortex spherical map corresponding to the sample brain image by inputting the sample brain image into the preset spherical map network, trains the preset spherical map network according to the sample cortex spherical map and the preset reference cortex spherical map, and can train the preset spherical map network more accurately through a large number of sample brain images, thereby improving the accuracy of the obtained spherical map network.
On the basis of the foregoing embodiment, as an optional implementation manner, before training the preset spherical mapping network according to the sample cortex spherical mapping map and the preset reference cortex spherical mapping map to obtain the spherical mapping network, the foregoing method further includes: and processing the sample brain image by using a preset brain image segmentation tool to obtain a preset reference brain cortex spherical map.
Specifically, the computer device processes the sample brain image by using a preset brain image segmentation tool to obtain a preset reference brain cortex spherical map. Alternatively, the preset brain image segmentation tool may be a freesurfer segmentation tool, or may be another segmentation tool for processing brain images. In this embodiment, the computer device processes the sample brain image by using a preset brain image segmentation tool, so that a preset reference cortex spherical map corresponding to the sample brain image can be obtained quickly, the efficiency of obtaining the preset reference cortex spherical map is improved, and further the efficiency of training a preset spherical map network according to the obtained sample cortex spherical map and the preset reference cortex spherical map is improved, so as to obtain the spherical map network.
Fig. 6 is a flowchart of a brain image classification method according to another embodiment. The embodiment relates to a specific implementation process of training a spherical surface convolution network by computer equipment. As shown in fig. 6, based on the above embodiment, as an alternative implementation manner, the training process of the spherical convolution network may include:
s601, acquiring a sample cerebral cortex spherical map.
Specifically, the computer device first obtains a sample cerebral cortex spherical map corresponding to a sample brain image. Optionally, the computer device may input the sample brain image into the above-mentioned spherical mapping model to obtain a sample cerebral cortex spherical mapping map corresponding to the sample brain image.
S602, inputting the sample cerebral cortex spherical map into a preset spherical convolution network to obtain a sample classification result of the sample cerebral image corresponding to the sample cerebral cortex spherical map.
Specifically, the computer equipment inputs the obtained sample cerebral cortex spherical map into a preset spherical convolution network to obtain a sample classification result of a sample cerebral image corresponding to the sample cerebral cortex spherical map. The sample classification result of the sample brain image corresponding to the sample cortex spherical map may include AD, stable mild cognitive impairment (sMCI), progressive mild cognitive impairment (pMCI), and normal, and it should be noted that the obtained sample classification result of the sample brain image corresponding to the sample cortex spherical map may further include a probability prediction index for each classification category, for example, the obtained sample classification result of the sample brain image corresponding to a sample cortex spherical map may be: normal: 70%, sMCI:20%, pMCI:7%, AD:3%.
S603, training a preset spherical convolution network according to a sample classification result and a label of a sample brain image mark corresponding to a sample cerebral cortex spherical map in advance to obtain the spherical convolution network.
Specifically, the computer equipment acquires a value of a loss function of a preset spherical convolution network according to the obtained sample classification result of the sample brain image corresponding to the sample brain cortex spherical map and a label of a sample brain image mark corresponding to the sample brain cortex spherical map in advance, trains the preset spherical convolution network according to the value of the loss function of the preset spherical convolution network, and determines the corresponding preset spherical convolution network as the spherical convolution network until the value of the loss function of the preset spherical convolution network reaches a stable value. Optionally, the label of the sample brain image label corresponding to the sample cerebral cortex spherical map may be normal, mci, pMCI, or both. For example, when the label of the sample brain image label corresponding to the sample brain cortex spherical map is normal, the sample brain cortex spherical map is input into a preset spherical convolution network, and when the obtained sample classification result is: normal: 96%, sMCI:5%, pMCI:2%, AD: and 3%, when the value of the loss function of the preset spherical convolution network obtained at the moment is 0.05, and when the result obtained when the parameters of the preset spherical convolution network are changed no longer changes, the value of the loss function of the preset spherical convolution network reaches a stable value, and the computer equipment determines the corresponding preset spherical convolution network as the spherical convolution network when the value of the loss function is 0.05.
In this embodiment, the computer device inputs the sample cerebral cortex spherical map into the preset spherical convolution network to obtain a sample classification result of the sample cerebral image corresponding to the sample cerebral cortex spherical map, trains the preset spherical convolution network according to the obtained sample classification result and the label of the sample cerebral image label corresponding to the sample cerebral cortex spherical map in advance, and can train the preset spherical convolution network more accurately through a large number of sample cerebral cortex spherical maps, thereby improving the accuracy of the obtained spherical convolution network.
It should be understood that, although the steps in the flowcharts of fig. 2-6 are shown in order as indicated by the arrows, these steps are not necessarily performed in order as indicated by the arrows. The steps are not strictly limited to the order of execution unless explicitly recited herein, and the steps may be executed in other orders. Moreover, at least some of the steps in fig. 2-6 may include multiple sub-steps or stages that are not necessarily performed at the same time, but may be performed at different times, nor does the order in which the sub-steps or stages are performed necessarily occur in sequence, but may be performed alternately or alternately with at least a portion of the sub-steps or stages of other steps or other steps.
Fig. 7 is a schematic structural diagram of a brain image classification device according to an embodiment. As shown in fig. 7, the apparatus may include: a first acquisition module 10, a second acquisition module 11 and a classification module 12.
Specifically, the first acquiring module 10 is configured to acquire a brain image to be classified;
the second obtaining module 11 is configured to input a brain image into the spherical mapping model, and obtain a cerebral cortex spherical mapping map corresponding to the brain image;
the classification module 12 is configured to input the cortex spherical map into a spherical convolution network to obtain a classification result of the brain image.
Optionally, the network structure of the spherical convolution network is any one of the following: a three-dimensional dense connection network structure; a spherical residual error network structure; the three-dimensional dense connection network incorporates a feature weighting module network structure.
Optionally, the sphere mapping model comprises a sphere mapping algorithm or a sphere mapping network.
The brain image classification device provided in this embodiment may execute the above method embodiment, and its implementation principle and technical effects are similar, and will not be described herein.
On the basis of the above embodiment, optionally, the above second obtaining module 11 includes: a first acquisition unit and a second acquisition unit.
Specifically, the first acquisition unit is used for inputting the brain image into the spherical mapping model to obtain a cerebral cortex spherical expansion diagram corresponding to the brain image;
the second acquisition unit is used for mapping the cerebral cortex spherical surface unfolded map to a preset standard spherical surface to obtain a cerebral cortex spherical surface map corresponding to the cerebral image.
The brain image classification device provided in this embodiment may execute the above method embodiment, and its implementation principle and technical effects are similar, and will not be described herein.
On the basis of the foregoing embodiment, optionally, the first obtaining unit is specifically configured to input the brain image into a spherical mapping model, and expand the brain image according to a preset expansion rule, so as to obtain a cerebral cortex spherical expansion map corresponding to the brain image.
The brain image classification device provided in this embodiment may execute the above method embodiment, and its implementation principle and technical effects are similar, and will not be described herein.
On the basis of the above embodiment, when the spherical mapping model is a spherical mapping network, optionally, the apparatus further includes: the system comprises a third acquisition module, a fourth acquisition module and a first training module.
Specifically, the third acquisition module is used for acquiring a sample brain image;
a fourth acquisition module, configured to input a sample brain image into a preset spherical mapping network, to obtain a sample cerebral cortex spherical mapping map corresponding to the sample brain image;
the first training module is used for training the preset spherical mapping network according to the sample cerebral cortex spherical mapping diagram and the preset reference cerebral cortex spherical mapping diagram to obtain the spherical mapping network.
The brain image classification device provided in this embodiment may execute the above method embodiment, and its implementation principle and technical effects are similar, and will not be described herein.
On the basis of the above embodiment, optionally, the above apparatus further includes: and a fifth acquisition module.
Specifically, the fifth acquisition module is configured to process the sample brain image by using a preset brain image segmentation tool, so as to obtain a preset reference cerebral cortex spherical map.
The brain image classification device provided in this embodiment may execute the above method embodiment, and its implementation principle and technical effects are similar, and will not be described herein.
On the basis of the above embodiment, optionally, the above apparatus further includes: a sixth acquisition module, a seventh acquisition module, and a second training module.
Specifically, the sixth acquisition module is used for acquiring a sample cerebral cortex spherical map;
a seventh obtaining module, configured to input the sample cortex spherical map into a preset spherical convolution network, to obtain a sample classification result of a sample brain image corresponding to the sample cortex spherical map;
and the second training module is used for training a preset spherical convolution network according to the sample classification result and a label of a sample brain image mark corresponding to the sample cerebral cortex spherical map in advance to obtain the spherical convolution network.
The brain image classification device provided in this embodiment may execute the above method embodiment, and its implementation principle and technical effects are similar, and will not be described herein.
For specific limitations of the brain image classification device, reference may be made to the above limitation of the brain image classification method, and no further description is given here. The respective modules in the above brain image classification device may be implemented in whole or in part by software, hardware, and a combination thereof. The above modules may be embedded in hardware or may be independent of a processor in the computer device, or may be stored in software in a memory in the computer device, so that the processor may call and execute operations corresponding to the above modules.
In one embodiment, a computer device is provided comprising a memory and a processor, the memory having stored therein a computer program, the processor when executing the computer program performing the steps of:
acquiring brain images to be classified;
inputting the brain image into a spherical mapping model to obtain a cerebral cortex spherical mapping diagram corresponding to the brain image;
and inputting the cerebral cortex spherical map into a spherical convolution network to obtain a classification result of the brain image.
The computer device provided in the foregoing embodiments has similar implementation principles and technical effects to those of the foregoing method embodiments, and will not be described herein in detail.
In one embodiment, a computer readable storage medium is provided having a computer program stored thereon, which when executed by a processor, performs the steps of:
acquiring brain images to be classified;
inputting the brain image into a spherical mapping model to obtain a cerebral cortex spherical mapping diagram corresponding to the brain image;
and inputting the cerebral cortex spherical map into a spherical convolution network to obtain a classification result of the brain image.
The computer readable storage medium provided in the above embodiment has similar principle and technical effects to those of the above method embodiment, and will not be described herein.
Those skilled in the art will appreciate that implementing all or part of the above described methods may be accomplished by way of a computer program stored on a non-transitory computer readable storage medium, which when executed, may comprise the steps of the embodiments of the methods described above. Any reference to memory, storage, database, or other medium used in the various embodiments provided herein may include non-volatile and/or volatile memory. The nonvolatile memory can include Read Only Memory (ROM), programmable ROM (PROM), electrically Programmable ROM (EPROM), electrically Erasable Programmable ROM (EEPROM), or flash memory. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms such as Static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double Data Rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous Link DRAM (SLDRAM), memory bus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), among others.
The technical features of the above-described embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above-described embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.
The above examples illustrate only a few embodiments of the invention, which are described in detail and are not to be construed as limiting the scope of the invention. It should be noted that it will be apparent to those skilled in the art that several variations and modifications can be made without departing from the spirit of the invention, which are all within the scope of the invention. Accordingly, the scope of protection of the present invention is to be determined by the appended claims.
Claims (10)
1. A method of classifying brain images, the method comprising:
acquiring brain images to be classified;
inputting the brain image into a spherical mapping model to obtain a cerebral cortex spherical expansion diagram corresponding to the brain image;
mapping the cerebral cortex spherical surface unfolded map to a preset standard spherical surface to obtain a cerebral cortex spherical surface mapping map corresponding to the cerebral image;
and inputting the cerebral cortex spherical map into a spherical convolution network to obtain a classification result of the brain image.
2. The method according to claim 1, wherein inputting the brain image into a spherical mapping model, to obtain a cerebral cortex spherical expansion map corresponding to the brain image, comprises:
inputting the brain image into the spherical mapping model, and expanding the brain image according to a preset expansion rule to obtain a cerebral cortex spherical expansion diagram corresponding to the brain image.
3. The method according to claim 1, wherein the network structure of the spherical convolution network is any one of the following: a three-dimensional dense connection network structure; a spherical residual error network structure; the three-dimensional dense connection network incorporates a feature weighting module network structure.
4. The method of claim 1, wherein the sphere mapping model comprises a sphere mapping algorithm or a sphere mapping network.
5. The method of claim 4, wherein when the spherical mapping model is a spherical mapping network, the training process of the spherical mapping model comprises:
acquiring a sample brain image;
inputting the sample brain image into a preset spherical mapping network to obtain a sample cerebral cortex spherical mapping diagram corresponding to the sample brain image;
training the preset spherical mapping network according to the sample cerebral cortex spherical mapping diagram and a preset reference cerebral cortex spherical mapping diagram to obtain the spherical mapping network.
6. The method of claim 5, wherein the training the predetermined spherical map network based on the sample cortical sphere map and a predetermined reference cortical sphere map, further comprises, prior to obtaining the spherical map network:
and processing the sample brain image by using a preset brain image segmentation tool to obtain the preset reference brain cortex spherical map.
7. The method of claim 1, wherein the training process of the spherical convolution network comprises:
acquiring a sample cerebral cortex spherical map;
inputting the sample cerebral cortex spherical map into a preset spherical convolution network to obtain a sample classification result of a sample cerebral image corresponding to the sample cerebral cortex spherical map;
and training the preset spherical convolution network according to the sample classification result and a label of a sample brain image mark corresponding to the sample cerebral cortex spherical map in advance to obtain the spherical convolution network.
8. A brain image classification device, the device comprising:
the first acquisition module is used for acquiring brain images to be classified;
the second acquisition module is used for inputting the brain image into a spherical mapping model to obtain a cerebral cortex spherical expansion diagram corresponding to the brain image; mapping the cerebral cortex spherical surface unfolded map to a preset standard spherical surface to obtain a cerebral cortex spherical surface mapping map corresponding to the cerebral image;
and the classification module is used for inputting the cerebral cortex spherical map into a spherical convolution network to obtain a classification result of the brain image.
9. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor implements the steps of the method of any of claims 1-7 when the computer program is executed.
10. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the method according to any of claims 1-7.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911105414.XA CN110874614B (en) | 2019-11-13 | 2019-11-13 | Brain image classification method, computer device, and readable storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911105414.XA CN110874614B (en) | 2019-11-13 | 2019-11-13 | Brain image classification method, computer device, and readable storage medium |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110874614A CN110874614A (en) | 2020-03-10 |
CN110874614B true CN110874614B (en) | 2023-04-28 |
Family
ID=69717949
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201911105414.XA Active CN110874614B (en) | 2019-11-13 | 2019-11-13 | Brain image classification method, computer device, and readable storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110874614B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111899217A (en) * | 2020-06-19 | 2020-11-06 | 上海联影智能医疗科技有限公司 | Brain image processing method and readable storage medium |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109215104A (en) * | 2018-09-04 | 2019-01-15 | 江文涛 | It is a kind of for brain structural images display methods and device through cranium stimulation therapy |
CN109816632A (en) * | 2018-12-25 | 2019-05-28 | 东软集团股份有限公司 | Brain image processing method, device, readable storage medium storing program for executing and electronic equipment |
CN110163271A (en) * | 2019-05-13 | 2019-08-23 | 武汉大学 | A kind of full-view image object detection method based on spherical projection grid and spherical convolution |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US10580159B2 (en) * | 2017-01-31 | 2020-03-03 | Siemens Healthcare Gmbh | Coarse orientation detection in image data |
-
2019
- 2019-11-13 CN CN201911105414.XA patent/CN110874614B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109215104A (en) * | 2018-09-04 | 2019-01-15 | 江文涛 | It is a kind of for brain structural images display methods and device through cranium stimulation therapy |
CN109816632A (en) * | 2018-12-25 | 2019-05-28 | 东软集团股份有限公司 | Brain image processing method, device, readable storage medium storing program for executing and electronic equipment |
CN110163271A (en) * | 2019-05-13 | 2019-08-23 | 武汉大学 | A kind of full-view image object detection method based on spherical projection grid and spherical convolution |
Non-Patent Citations (1)
Title |
---|
基于立体全景视觉的移动机器人3D SLAM研究;熊斯睿;《中国优秀博硕士学位论文全文数据库(硕士)信息科技辑》;20160215(第2期);第I138-1412页 * |
Also Published As
Publication number | Publication date |
---|---|
CN110874614A (en) | 2020-03-10 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109993726B (en) | Medical image detection method, device, equipment and storage medium | |
US20220351370A1 (en) | Auxiliary pathological diagnosis method | |
CN111161269B (en) | Image segmentation method, computer device, and readable storage medium | |
CN110717905B (en) | Brain image detection method, computer device, and storage medium | |
CN110223279B (en) | Image processing method and device and electronic equipment | |
CN110738643B (en) | Analysis method for cerebral hemorrhage, computer device and storage medium | |
CN110210519B (en) | Classification method, computer device, and storage medium | |
CN111210465B (en) | Image registration method, image registration device, computer equipment and readable storage medium | |
CN110210543B (en) | Image classification system, method, apparatus and storage medium | |
CN110751187B (en) | Training method of abnormal area image generation network and related product | |
CN110866909A (en) | Training method of image generation network, image prediction method and computer equipment | |
CN112102235B (en) | Human body part recognition method, computer device, and storage medium | |
CN111223158B (en) | Artifact correction method for heart coronary image and readable storage medium | |
CN110874614B (en) | Brain image classification method, computer device, and readable storage medium | |
CN111723836A (en) | Image similarity calculation method and device, electronic equipment and storage medium | |
CN111160441B (en) | Classification method, computer device, and storage medium | |
CN111160442B (en) | Image classification method, computer device, and storage medium | |
CN110766653B (en) | Image segmentation method and device, computer equipment and storage medium | |
CN111091504B (en) | Image offset field correction method, computer device, and storage medium | |
CN110738639B (en) | Medical image detection result display method, device, equipment and storage medium | |
CN114066798A (en) | Brain tumor nuclear magnetic resonance image data synthesis method based on deep learning | |
CN111210414B (en) | Medical image analysis method, computer device, and readable storage medium | |
CN111027469B (en) | Human body part recognition method, computer device, and readable storage medium | |
CN107799166B (en) | Medical imaging system | |
CN112102244A (en) | Fetus ultrasonic standard section image detection method, computer equipment and storage medium |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |