135 results about "Brain mri" patented technology

The invention discloses an MRI medical image correction method and system based on a convolutional neural network and a computer readable storage medium, and the method comprises the steps: obtaining a brain MRI medical image, and carrying out the preprocessing; extracting lesion features of the brain lacuna infarction from the preprocessed MRI medical image through a convolutional neural network, and positioning a lesion area of the brain lacuna infarction; performing detail enhancement on the lesion area of the cerebral lacuna infarction, segmenting the lesion area and optimizing a segmentation result; and carrying out artifact restoration on the segmented image, and generating a three-dimensional focus visualization model based on the restored image. According to the method, through processing in multiple aspects such as quality evaluation, lesion recognition, segmentation, artifact repair and three-dimensional visualization of the MRI image, the lesion area can be recognized more efficiently and accurately, the image quality is improved, and an efficient and accurate brain lacuna infarction diagnosis auxiliary tool is provided.

The invention discloses a post-stroke cognitive impairment prediction method based on multi-modal feature fusion. The method comprises the steps that firstly, multi-modal information of a stroke patient is collected, wherein the multi-modal information comprises a three-dimensional brain MRI image, an EEG electroencephalogram signal and clinical medical record information; secondly, converting MRI into a tensor, and inputting the tensor into a multi-scale spatial-temporal feature extraction backbone network to obtain MRI modal features; the EEG electroencephalogram signals are subjected to electroneurographic signals and are combined with Transform, and EEG modal features are obtained; the clinical medical record information of the stroke patient is converted into semantic sentences, the semantic sentences are input into a two-channel semantic encoder for encoding extraction, and clinical medical record information features are obtained. And finally, inputting the MRI modal features, the EEG modal features and the clinical medical record information features into a three-modal fusion device to obtain fusion features, and outputting probability prediction through a classifier. The post-stroke cognitive impairment prediction method achieves accurate prediction of post-stroke cognitive impairment, and significantly improves robustness and medical interpretation.

The embodiment of the invention discloses an image annotation method and system applied to brain MRI image segmentation, and the method comprises the steps: obtaining a brain MRI image data set of a target object, and the brain MRI image data set comprises original image sequences of a plurality of scanning levels; performing multi-modal feature fusion processing on the original image sequence to generate an enhanced image feature set; calling a multi-layer cascade segmentation network to perform hierarchical feature extraction on the enhanced image feature set to obtain a multi-scale anatomical structure feature map; and performing region boundary optimization processing based on the multi-scale anatomical structure feature map, and generating a marked brain structure segmentation image. Therefore, the boundary of each structure of the brain can be accurately defined, the segmented image is more accurate and clearer, and the image segmentation and marking of the brain MRI image can be accurately and clearer realized.

The invention relates to the technical field of image processing, in particular to a craniocerebral disease area identification and detection method and system based on an MRI (Magnetic Resonance Imaging) image, and the method comprises the steps: obtaining a plurality of sub-images with different scales according to a gray level image of the craniocerebral MRI image; performing multi-scale analysis on the gradient value of any pixel point according to each sub-image to obtain a multi-scale gradient coefficient of any pixel point; obtaining a multi-scale local anomaly degree according to the gray values of the pixel points in different local ranges of any pixel point and the distribution in the gradient direction; optimizing the gradient value of any pixel point according to the multi-scale gradient coefficient and the multi-scale local anomaly degree to obtain a self-adaptive gradient value, and performing image enhancement on the grayscale image by using an anisotropic diffusion filtering algorithm according to the self-adaptive gradient value of each pixel point so as to identify a craniocerebral disease region. And the effect of performing image enhancement on the MRI image by using the anisotropic diffusion filtering algorithm is improved.

The invention discloses a brain MRI image segmentation method based on an RWKV model, and belongs to the technical field of medical image processing and artificial intelligence crossing. Firstly, an RWKV linear self-attention module is introduced into a U-Net network framework, and association between remote pixels is established with relatively low calculation overhead, so that the recognition precision of a brain tumor area is improved, the reasoning time is effectively controlled, and the practicability of a model is enhanced. Secondly, according to the method, multi-scale coding and a feature fusion mechanism are combined, local and global information is extracted in a combined manner, features are mined from different resolution levels, and adaptive fusion is realized, so that the fine-grained segmentation effect is improved. And finally, in order to enhance the generalization ability and lightweight deployment performance of the model, the network structure is further optimized, and the overall computing resource demand is reduced, so that the method can better adapt to cross-patient MRI data in a complex clinical environment, and has good practical value and popularization potential.

The invention discloses a method for determining brain PET standardized reconstruction parameters based on clinical brain MRI and PET data, and relates to the field of medical image processing. The method comprises the following steps: firstly, collecting clinically paired MRI (Magnetic Resonance Imaging) and PET (Positron Emission Tomography) images for preprocessing, carrying out virtual reconstruction by combining a partial volume correction algorithm and system parameters of equipment to estimate'real 'brain activity distribution so as to obtain a first simulated PET image, and calculating difference or similarity between the first simulated PET image and a clinical PET image; through multiple iterations, obtaining an optimal'real 'brain activity distribution diagram, and processing the optimal'real' brain activity distribution diagram into a'standardized PET image 'according to standardized Gaussian filtering; then, constructing different candidate combinations of reconstruction parameters, inputting an optimal'real 'brain activity distribution diagram, traversing a second simulated PET image generated by simulation reconstruction under each combination, and respectively performing difference or similarity calculation with the'standardized PET image', so as to obtain an optimal'real 'brain activity distribution diagram; and selecting the group of candidate reconstruction parameters with the minimum difference or the highest similarity as final standardized reconstruction parameters. The method gets rid of dependence on a physical motif.

The invention discloses a multi-view KAN brain MRI (Magnetic Resonance Imaging) image registration method under the guidance of edge information, and the method comprises the steps: converting a brain moving image and fixed image registration task into a non-iterative coarse-to-fine calculation process through the intensity robustness of the edge information and the powerful representation capability of the KAN in an intensive task; a channel redundancy reduction mechanism is introduced into a convolutional layer of an encoder to extract low-redundancy related features of two images, a decoder of a multi-view KAN structure pays attention to dependency of different distances of the features through three different views, the decoder generates an intermediate deformation field by using a non-iterative coarse-to-fine principle, and the low-redundancy related features of the two images are extracted. And the accuracy of processing the large deformation task by the model is improved. Besides, model training is guided through edge information, a deformed moving image and a deformation field with higher precision can be output through the method, the visual perception quality can be improved, and the method surpasses various current advanced algorithms in the aspect of quantitative evaluation.

The invention provides an automatic segmentation method for a focus of brain injury of a premature infant. The method comprises the following steps: preprocessing a data set; the data set is a brain MRI image of the premature brain injury patient; dividing the preprocessed data set into a training set, a test set and a verification set according to a preset proportion; building an automatic lesion segmentation network model of the brain injury of the premature infant; the network model comprises a network architecture composed of a PAFM module group, a CMSC module group and a 3D-UXNET; inputting the training set into the network model for training, and adjusting hyper-parameters based on the verification set; and inputting the test set into the trained network model to obtain an automatic lesion segmentation result of the brain injury of the premature infant corresponding to the test set. According to the method, features are extracted through multi-scale convolution of the CMSC module, and the capability of segmenting small focuses is enhanced. According to the method, multiple modal data are utilized at the same time, and modal information is fused in parallel through a hierarchical fusion strategy and a PAFM module formed by inter-modal attention and cross-modal attention, so that the segmentation performance of the model is effectively improved.

The invention discloses an Alzheimer's disease early diagnosis method and device based on a diffusion model. The method comprises the steps that a plurality of brain MRI images of a diagnosis object in different periods are acquired and preprocessed; obtaining an MRI compression feature corresponding to each brain MRI image by using an encoder in the trained auto-encoder network; taking the MRI feature as a conditional variable, and performing prediction by using the trained diffusion model to obtain an MRI prediction feature; and obtaining a brain MRI image obtained by performing MRI scanning on the diagnosis object again, obtaining an MRI real feature through the trained encoder, taking an error between the MRI real feature and the MRI prediction feature as an abnormal score, and if the abnormal score exceeds a predetermined threshold, determining that the Alzheimer's disease is positive. The method combines the generation capability of the diffusion model and the time sequence modeling advantages, has the advantages of high calculation efficiency, high interpretability and the like, and can realize early diagnosis of the Alzheimer's disease only through healthy population data.

The invention discloses a medical image intelligent diagnosis and three-dimensional visualization interaction method and system based on artificial intelligence and Unreal Engine. The system is composed of a data acquisition module, an artificial intelligence analysis module, a three-dimensional reconstruction module, a rendering interaction module and a report generation module. The method comprises the following steps: firstly, carrying out voxel normalization on a DICOM / NIfTI image; the improved 3D U-Net is combined with CBAM attention and WGAN-GP training, and a focus initial density field is output; an AI enhanced implicit voxel reconstruction (AEIVR) algorithm is proposed, an enhanced density field is generated by using position coding + MLP, and grid extraction is completed by using improved Marking Cubes of a dynamic threshold; high-fidelity real-time rendering is realized in combination with a Nand micro polygon and Lumen global illumination in Unreal Engine, and VR / AR interaction and virtual surgery (real-time updating of resection mask driving) are supported. According to the method, the problems of large artifacts and insufficient interactivity and real-time performance of existing two-dimensional output and three-dimensional reconstruction are solved, and the end-to-end segmentation-reconstruction-rendering-interaction-reporting process is realized. On typical CT / MRI data, the segmentation Dice is about 0.90-0.94, the reconstruction voxel resolution is 0.1-0.5 mm, the artifacts are reduced by about 20%-30%, the PSNR is improved by about 3-7 dB, and the rendering interaction delay lt is achieved; the system is suitable for scenes such as lung CT, brain MRI, orthopaedic X-ray and cardiac ultrasound, and can be expanded to preoperative planning, remote consultation and teaching.

The invention discloses a cerebral apoplexy patient postoperative neural function real-time monitoring method. The method comprises the steps that brain MRI image data and clinical information of a to-be-monitored patient and post-operation electroencephalogram data collected in real time are collected and preprocessed; respectively extracting neurophysiological features and radiomics features of the patient to be monitored based on the preprocessed data; performing significant feature screening on the clinical information, the neurophysiological features and the radiomics features, performing preprocessing on the screened data, and combining minimum absolute contraction and selection operator regression analysis to obtain quantitative electroencephalogram data feature indexes and radiomics scores; inputting the screened clinical information, the quantitative electroencephalogram data characteristic index and the radiomics score of the to-be-monitored patient into a trained prediction model, and predicting a risk index of early neurological deterioration of the to-be-monitored patient; the problem that the evaluation result is inaccurate due to the fact that a single clinical feature is adopted to evaluate the neural function in a traditional method is solved.

The invention discloses an MC I conversion classification method for an incomplete multi-modal hierarchical feature fusion network. The method comprises the following steps: firstly, acquiring brain MRI, PET images and clinical table data of a patient; constructing a pyramid structure module composed of an MR pyramid, a PET pyramid and a common pyramid, capturing image features of different scales by using the MRI pyramid and the PET pyramid, and inputting the image features of different scales and clinical table features into a layered Mamba-cross attention module to obtain fusion features; splicing the fusion features with the output of the last layer of the MR I pyramid, the PET pyramid and the common pyramid to obtain three groups of feature images with different scales; generating multi-modal fusion features from the three groups of feature images with different scales through a triple hybrid fusion module; four classification results are obtained through an MLKAN classification head on the three sets of feature images of different scales and the multi-modal fusion features, and the four classification results are spliced to obtain a final classification result.

The invention discloses a multi-modal medical image processing method and device, a storage medium and computer equipment. Comprising the following steps: performing three-dimensional discrete wavelet transform on a brain MRI image of a target patient to generate an MRI wavelet coefficient; inputting the MRI wavelet coefficient into a diffusion model to obtain a reference PET wavelet coefficient of the brain of the target patient in a healthy state; performing inverse wavelet transform on the reference PET wavelet coefficient to generate a reference PET image; and comparing the brain PET image of the target patient with the reference PET image, and determining the metabolic deviation index of the brain of the target patient. Therefore, each patient can take the condition without the neurodegenerative change as a contrast, space standardization does not need to be carried out on a group template, anatomical structure distortion caused by the space standardization is greatly reduced, voxel-level accurate analysis of the neurodegenerative disease is realized, tiny pathological change aiming at the patient can be identified, and the accuracy of voxel-level accurate analysis of the neurodegenerative disease is improved. And clinical doctors are assisted in early diagnosis.

The invention discloses a brain MRI super-resolution reconstruction method based on a learnable weight linear accumulation strategy, and relates to the technical field of magnetic resonance imaging image super-resolution reconstruction. According to the method, the Swin-WKV model is built, the model is based on a Swin Transform and RWKV architecture Transform-like model, high-quality reconstruction of the brain MRI image can be achieved with quite low memory consumption, a mixed loss function is provided, image domain difference and frequency domain difference are combined, and the model is guided to restore frequency spectrum and pixel-level features of a reference image as much as possible.

The invention discloses a tumor diagnosis method and system for generating a molecular spatial distribution diagram based on MRI (Magnetic Resonance Imaging), and belongs to the technical field of image data processing. The method comprises the following steps: constructing an MRI-pathological local matching data set; the Transform and the GAN are combined to construct a local MRI (Magnetic Resonance Imaging) to generate a small pathological graph model; and constructing and verifying a molecular spatial distribution diagram. According to the method, the histopathology is taken as a bridge, a two-layer mapping relation is constructed between the MRI and the molecular features, the global molecular features are estimated through local matching data, existing clinical resources are effectively utilized, and three-dimensional space distribution evaluation of the molecular features is indirectly achieved at low cost. And the constructed model can be used for pathological diagnosis in any region on the brain MRI sequence, so that the potential invasion region of the tumor can be explored, and reference is provided for the surgical resection range definition and treatment strategy planning of the glioma.

The invention discloses an MRI (Magnetic Resonance Imaging) accelerated reconstruction method and system based on a K-space enhanced reversible diffusion model, and the method comprises the steps: firstly collecting a brain MRI scanning image, carrying out the undersampling processing, normalizing a gray value, and randomly cutting an image block with a specified size; inputting the image blocks into an end-to-end adaptive K-space enhanced reversible diffusion model to generate a reconstructed image; a progressive mask strategy and cross-domain feature injection are adopted, image edges and detail textures are recovered in a targeted mode, adaptive optimization of different anatomical structures is achieved through a learnable mechanism, and reconstruction quality and robustness are remarkably improved. According to the method, the performance is outstanding on the core indexes such as the peak signal-to-noise ratio (PSNR) and the structural similarity (SSIM), the precise recovery capability of the image edge structure is verified through Canny and Sobel operators, key breakthrough is achieved on the reconstruction quality, the training efficiency and the clinical applicability, and a solution with theoretical innovation and engineering application value is provided for the MRI accelerated imaging technology.

The application discloses an auxiliary discrimination method for a nasopharyngeal carcinoma radiation-induced brain injury image, a device and a medium, relating to the technical field of medical information. The method comprises: preprocessing an MRI image, segmenting temporal lobe areas, using the MRI image temporal lobe areas and corresponding focus of infection labels to train a slice-level prediction model, and predicting the probability of a radiation-induced brain injury focus existing in a single temporal lobe image area; taking the probability of all slices in each subject containing the focus of infection as the input, wherein the probability is outputted by the slice-level prediction model, and taking whether the subject has radiation-induced brain injury as the output, training a patient-level prediction model; finally, using the trained slice-level prediction model and the patient-level prediction model to obtain the radiation-induced brain injury disease probability of a single nasopharyngeal carcinoma subject, and when a plurality of modes coexist, taking a plurality of modal prediction results to output a mean value as a final prediction result. The present application can determine, according to the input nasopharyngeal carcinoma brain MRI image, whether radiation-induced brain injury exist, so as to reduce missed diagnosis of nasopharyngeal carcinoma radiation-induced brain injury.

The invention relates to the technical field of psychiatric disease diagnosis, and discloses a defective schizophrenia prediction and analysis system based on multi-modal brain network characteristics, which comprises an image data acquisition module, an image data processing module, a deep learning model module and a prediction module which are connected in sequence, the image data acquisition module acquires a brain MRI image; the image data processing module carries out cortical and subcortical reconstruction on the brain MRI image to obtain a model input index; the deep learning model module performs training by using a training data set composed of a plurality of brain MRI images and defective schizophrenia diagnosis results thereof to obtain a target prediction model; and the prediction module inputs the real-time brain MRI image into the target prediction model to obtain a prediction result. According to the method, the multi-dimensional brain structure image features are integrated, diagnosis information contained in brain region changes is fully mined, prediction results of defective schizophrenia and non-defective schizophrenia are improved, and a more comprehensive biological basis is provided for clinical diagnosis.

The invention relates to the technical field of medical image recognition, in particular to an Alzheimer's disease MRI diagnosis method based on width neural network learning, and the method comprises the steps: 1, carrying out the adaptive ROI cutting of a brain MRI image; step 2, performing dynamic contrast compensation binaryzation; step 3, morphological optimization; step 4, gray scale returning; 5, density peak value guided K-means + + feature extraction is carried out; and step 6, obtaining a width neural network, inputting the final enhanced feature map into the width neural network for final classification prediction, and realizing diagnosis of the Alzheimer's disease MRI. According to the method, a binary segmentation method of dynamic contrast compensation and multi-scale feature enhancement of K-means + + driving are combined, the method is devoted to the hippocampus segmentation problem in image recognition for the Alzheimer's disease and the problem that subtle variation of an early-stage AD patient cannot be captured, and convenience is provided for doctors.

The present invention discloses a convolutional neural network-based MRI medical image correction method, system, and computer-readable storage medium. The method comprises: obtaining a brain MRI medical image and performing preprocessing; extracting lesion features of lacunar infarction from the preprocessed MRI medical image using a convolutional neural network, and locating the lesion area of the lacunar infarction; performing detail enhancement on the lacunar infarction lesion area, segmenting the lesion area, and optimizing the segmentation result; performing artifact repair on the segmented image, and generating a three-dimensional lesion visualization model based on the repaired image. The present invention can more efficiently and accurately identify the lesion area, improve image quality, and provide an efficient and accurate diagnostic aid for lacunar infarction through multiple processing such as MRI image quality assessment, lesion identification, segmentation, artifact repair, and three-dimensional visualization.

The invention discloses a whole brain positioning and health state evaluation method and system based on an adaptive attention mechanism, electronic equipment and a computer readable storage medium. The whole-brain positioning and health state evaluation method based on the adaptive attention mechanism comprises the following steps: collecting whole-brain MRI data, and marking the structure of each part of the region of the whole brain; performing data preprocessing on the whole brain MRI data; on the basis of the whole-brain MRI data after data preprocessing, detecting each partial region of the whole brain by using a whole-brain positioning network based on an adaptive weight and residual error reconstruction mechanism; and according to the regional form of each part of the whole brain in the detection result, performing comprehensive health form evaluation. The method can solve the problems that the edge detection precision of each region of the whole brain is low and each partial region of the whole brain is difficult to accurately distinguish at present; the generalization ability is improved; the health state is accurately evaluated; manual intervention is reduced, one-key rapid whole-brain positioning and accurate health state evaluation are achieved, and simplicity and high efficiency are achieved.

The invention relates to a lightweight brain tumor segmentation method based on MSLQA-Net. The method is applied to tumor segmentation of brain tumor magnetic resonance imaging (MRI). The method comprises the following steps: S1, preprocessing brain MRI data, and dividing the data into a training set, a verification set and a test set; s2, constructing a multi-scale decomposition residual convolution module (MSDRC) to perform feature extraction, and forming a multi-level secondary feature map; s3, constructing a cross-layer feature aggregation module (CLFA), and performing efficient aggregation on the multi-level secondary feature map to obtain an aggregated feature map; s4, constructing a quadruple-dimension attention module (QDA), and further refining the aggregated feature map to obtain a refined feature map; s5, fusing multi-scale decomposition residual convolution, cross-layer feature aggregation and a quadruple-dimension attention module, and constructing an MSLQA-Net segmentation model; and S6, performing training, verification optimization and testing on the segmentation model by using the training set, the verification set and the test set. According to the method, in a brain tumor segmentation task, a high-precision result can be still kept while the reasoning speed is increased.

The present application relates to the field of data analysis, and in particular to a fetal brain MRI tissue analysis method, apparatus, and electronic device, comprising obtaining multiple post-treatment brain MRI images, each corresponding to a different gestational age; performing brain tissue segmentation on each post-treatment brain MRI image to obtain multiple post-treatment local brain tissue segmentation maps corresponding to multiple local brain tissues; obtaining, for each post-treatment brain MRI image, a pre-treatment local brain tissue segmentation map corresponding to each post-treatment local brain tissue segmentation map; determining, for each post-treatment brain MRI image, a deformation coefficient for each local brain tissue; generating, for each local brain tissue, a linear analysis map corresponding to each local brain tissue based on the deformation coefficients of the local brain tissue corresponding to all gestational ages; and determining final brain tissue analysis information based on the linear analysis maps corresponding to all local brain tissues. The present application facilitates a holistic analysis of the impact of changes before and after treatment on different gestational ages and different brain tissues.

The present invention relates to a method, device and medium for classifying brain small vessel disease images based on magnetic resonance imaging. The method first constructs and trains an image classification model based on ensemble learning, then obtains the brain MRI image to be classified, and applies the image classification model to obtain the lesion category corresponding to the brain MRI image to be classified. Specifically, training the image classification model based on ensemble learning includes the following steps: S101, obtaining a brain MRI data set, preprocessing the images in the brain MRI data set to obtain preprocessed images; S102, performing computational analysis on the preprocessed images to obtain corresponding multiple functional metrics, and screening a number of metric features for classification based on the multiple functional metrics; S103, training the image classification model based on the metric features. Compared with the prior art, the present invention has the advantages of high accuracy and easy operation.

The invention provides a brain longitudinal image generation method and device based on a novel diffusion model and a deformation field technology, a terminal, a medium and a product. The method comprises the following steps: acquiring a brain MRI image of a target detected object at a target time point and determining a time point to be predicted; obtaining a deformation field from the target time point to the to-be-predicted time point based on a trained novel diffusion model according to the preprocessed brain MRI image of the target time point and the to-be-predicted time point; and according to the brain MRI image of the target detected object at the target time point and the deformation field from the target time point to the to-be-predicted time point, based on the spatial deformation layer, obtaining a brain longitudinal image of the target detected object at the to-be-predicted time point. According to the method, the accuracy of generating the longitudinal brain image is improved, and early intervention and disease management of high-risk groups can be realized.

The invention discloses a three-dimensional surgical risk model reconstruction method based on brain anatomy and a functional atlas, and the method comprises the steps: S1, carrying out the image processing and fine segmentation of a brain MRI image or a DICOM sequence image of a patient, and recognizing a plurality of anatomical structures; s2, establishing a multi-dimensional risk assessment model, and performing quantitative scoring on each brain region from four dimensions of functional importance, structural vulnerability, recovery potential and clinical significance; s3, constructing a continuous three-dimensional risk field based on the risk score, and realizing the propagation of the risk value in the brain tissue based on a Gaussian diffusion model; and S4, based on an improved path planning algorithm, searching an optimal operation path in the three-dimensional risk field to realize risk minimization. The invention further discloses a corresponding system, electronic equipment and a computer readable storage medium.

The invention relates to the technical field of image processing, in particular to a brain age prediction method, system and device based on federal learning. According to the method, a three-dimensional brain MRI image is taken as input data, and multi-center collaborative modeling is realized by adopting a federated learning framework aiming at the problems of medical data privacy protection and data islands; each medical institution does not need to share original image data, and only completes feature extraction, model training and local parameter updating locally based on private data; after the central server receives the local parameters uploaded by the mechanisms, data distribution differences are considered, and a unified global model is generated through fusion of a preset aggregation rule. According to the method, privacy leakage and compliance risks of cross-mechanism transmission of original image data are effectively avoided, common characteristics of multi-center data can be integrated, the brain age prediction precision and generalization ability of a global model are improved, and the method is suitable for brain age evaluation research and clinical auxiliary diagnosis scenes jointly developed by multiple mechanisms.

The application provides a child neuropsychological development monitoring and management system. The system uses brain physiological parameters, brain MRI images, development scale characteristics and corresponding child neuropsychological development suffering grades to construct a child neuropsychological development evaluation model. The constructed child neuropsychological development evaluation model generates a current child neuropsychological development suffering grade of a user according to real-time brain physiological parameters collected by a system terminal, real-time brain MRI images monitored by a hospital and corresponding real-time development scale characteristics, provides a basis for whether the corresponding child needs further monitoring according to the current child neuropsychological development suffering grade, and improves the probability of finding potential abnormal development children. The application uses three kinds of fusion parameters related to child brain nerves, such as brain physiological parameters, brain image characteristics and subjective measurement table characteristics, to construct a high-precision real-time processing model, realizes intelligent monitoring of the morbidity of child neuropsychological development disorders by doctors, and reduces the burden on families.

The present invention discloses a method for segmenting brain tumors in MRI images based on an improved 3D-UNet, which specifically includes: 1) preprocessing the brain MRI images in the dataset to obtain a training set; 2) constructing an improved 3D-UNet framework by combining dilated convolution, channel attention mechanism, and residual convolution; 3) training the improved 3D-UNet, the brain tumor segmentation network, with the training set data; 4) importing the test data of the brain MRI image to be segmented into the brain tumor segmentation network to obtain the segmented result. The present invention provides an automatic, accurate, and repeatable tumor segmentation algorithm, which can effectively solve the problems that the traditional manual segmentation method is time-consuming and overly dependent on the subjective experience of experts, and has a higher segmentation accuracy than the method for segmenting brain MRI images based on UNet. At the same time, the method proposed by the present invention can be further applied to other medical MRI image segmentation tasks.

The application provides a medical image classification and feature visualization method based on a deep learning model and a feature visualization tool, and comprises the following steps: A, removing non-key information frames from a brain MRI image; B, performing slice enhancement on the data to obtain a slice data set; C, transversely splicing the slices in sequence to obtain a spliced picture data set; D, putting the spliced picture data set into SE-ResNet50 for training and testing to obtain a classification result; E, putting the trained spliced picture network and a sample spliced picture into Grad-CAM to generate a heat map; F, selecting key frames from the heat map, putting the key frame slice set into SE-ResNet50 to extract features and classify; G, generating a key frame heat map through Grad-CAM; and H, using the obtained classification result and heat map to assist in diagnosis. The application discloses a method for classifying and visualizing features of brain images based on a deep learning method. An optimized SE-ResNet50 network is used to extract features and classify patient brain images. Abnormal regions are marked and visualized through Grad-CAM, so that the efficiency and diagnosis rate are improved.