287 results about "Functional connectivity" patented technology

The invention provides a multichannel transcranial direct current stimulation cognitive function assessment method and system, and belongs to the field of neuroscience and brain cognitive function assessment. The method comprises the steps that electroencephalogram signal data of a subject under multi-channel transcranial direct current stimulation are collected and preprocessed; respectively performing micro-state analysis, brain function connectivity analysis and oscillation power analysis on the preprocessed data; constructing a linear mixed effect model, and inputting the results of the micro-state analysis, the brain function connectivity analysis and the oscillation power analysis into the model for quantitative evaluation of the cognitive function state; and a time sequence database is constructed, and the time sequence database is used for storing cross-time-point analysis results and analyzing the long-term change trend of the cognitive function based on a statistical model. Dynamic, accurate and high-temporal-spatial-resolution evaluation of the cognitive state of the subject is achieved, the long-term change trend of the cognitive state is dynamically monitored, and technical support is provided for early warning and rehabilitation effect tracking of neurodegenerative diseases.

The invention belongs to the field of electroencephalogram signal processing, and provides a time-frequency space electroencephalogram emotion recognition method based on three-dimensional space position embedding, which comprises the following steps of: firstly, constructing a three-dimensional electrode space position matrix based on an international 10-20 system standard, determining a space adjacency relation between electrodes, and calculating a phase locking value to obtain a functional connection matrix; then, deep feature fusion of an electrode spatial position matrix and a functional connection matrix is realized by adopting a hierarchical cross Transform architecture, the spatial position matrix represents spatial distribution features of a cerebral cortex region, and the functional connection matrix quantifies phase synchronization features of cross-brain region neural oscillation and simulates a brain spatial topological structure; and finally, extracting time, frequency and spatial features of the electroencephalogram signals through combination of a graph attention network and bidirectional long-short-term memory with an attention mechanism for emotion recognition. The method can effectively extract space structure information highly related to the emotional state, and significantly improves the accuracy of emotion recognition.

The invention discloses a Parkinson's dyskinesia individualized SCAN network positioning method based on a multi-modal image and deep learning. The method comprises the steps of obtaining multi-modal medical image data, preprocessing the multi-modal medical image data, obtaining a multi-modal structure image and functional connection data, and calculating a spontaneous neural activity index of a whole-brain voxel level; taking a priori brain region related to the spontaneous neural activity index and dyskinesia as a seed point, constructing a seed point voxel function connection graph representing individual brain function connection, and performing nonlinear feature fusion and extraction through the deep learning network model; the bilinear attention network is adopted to capture the interaction information of the feature data and the individual dyskinesia symptom which is significantly related, an individualized SCAN network positioning result is obtained, the structure-function coupling characteristics of the individual brain are comprehensively described, the cross-modal pathological features related to the dyskinesia can be more sensitively recognized, and the accuracy and accuracy of the diagnosis and treatment of the dyskinesia can be improved. And the accuracy and robustness of abnormal brain region detection are obviously improved.

The invention relates to the technical field of medical care informatics, and discloses an intelligent post-stroke cognitive impairment assessment method based on brain network dynamic reconstruction, which comprises the following steps: acquiring multi-modal brain image data of a patient, dynamically compensating the difference between haemodynamic delay and white matter conduction velocity through a nerve-blood vessel coupling model, and calculating the post-stroke cognitive impairment; precise space-time alignment of function and structure signals is realized; a pulse neural network dynamic entropy change model is constructed based on the alignment data, functional connection strength change is coded into a pulse distribution sequence, and Gamma wave band synchronization enhancement and other compensatory features are extracted; constraining rationality of brain network reconstruction by adopting an energy optimal transmission model, and generating a compensatory thermodynamic diagram; and real-time calculation of the edge end is realized through pulse time sequence compression coding. The limitation of traditional static analysis is avoided, the algorithm achieves microsecond response on embedded equipment through the pulse coding technology, and a reliable stroke rehabilitation evaluation tool is provided for primary hospitals.

The invention discloses a multi-modal feature combined depression auxiliary diagnosis system. The system comprises a sampling unit which is used for constructing a multi-modal depression data set by acquiring a depression screening scale, an electroencephalogram, a magnetoencephalogram and functional magnetic resonance imaging based on acquisition equipment; the feature extraction unit is used for extracting multi-modal brain features based on the depression data set, and the multi-modal brain features comprise power spectral density obtained by electroencephalogram signals, event-related potential, micro-state, prefrontal lobe gamma frequency band power spectral density obtained by magnetoencephalogram and event-related magnetic field; gray matter volume and resting state functional connection density are obtained through functional magnetic resonance imaging; a data preprocessing unit; the diagnosis model unit is used for constructing a multi-modal depression diagnosis model and training the model on the basis of the multi-modal brain features in combination with a fusion strategy; and an analysis and prediction unit. The extracted features are comprehensive and reasonable, the defect of each mode is overcome by the feature fusion method, and the fused features are advanced.

A juvenile depression cognitive impairment subtype classification and prognosis prediction method relates to the technical field of medical treatment, and mainly comprises the following steps: performing clinical evaluation and therapeutic response evaluation on a subject, performing MRI and magnetoencephalogram data acquisition, constructing a whole brain MSN of the subject, identifying MSN abnormal characteristics, obtaining functional connection change of a frequency band when magnetoencephalogram is abnormal, and determining the cognitive impairment subtype classification and prognosis prediction of the cognitive impairment subtype of the subject. A subtype classification model is established by fusing the MSN and cognitive function evaluation data, and a prognosis prediction model is established by analyzing MSN abnormal features, functional connection changes of frequency bands during abnormality, multi-dimensional treatment reactions and high-risk behaviors. According to the method, different levels of fusion measurement are carried out on the juvenile depression with cognitive function impairment brain mechanism through multi-modal brain images, a subtype classification model with diagnosis and treatment values is established, and a prognosis prediction model with clinical transformation potential is constructed; therefore, a theoretical basis and a technical means are provided for individualized precise diagnosis and treatment of the cognitive impairment of the juvenile depression.

The invention discloses a brain age prediction method based on structure and functional MRI image multi-modal fusion, and belongs to the technical field of medical image processing. The method comprises the following steps: firstly, extracting spatial structure characteristics of a structural magnetic resonance image by using DenseNet121; meanwhile, a function connection matrix is constructed according to the time sequence of the functions, a graph structure is constructed on the basis of the matrix, the characteristic of each node is the connection strength between the node and other nodes, and the edge is converted into sparse graph representation from the absolute value of the connection strength; then extracting functional features by using a graph attention network, and fusing the structure and the functional features by using a cross attention mechanism; and applying a gating mechanism fusion result to a brain age prediction regression task. According to the brain age prediction method, complementary information of multi-modal data is fully utilized, and biological markers of brain aging can be accurately captured.

The invention discloses a time interference electrical stimulation system for suppressing, controlling and adjusting, which comprises an image acquisition module for acquiring functional magnetic resonance images to acquire functional connection data; the stimulation module is used for respectively generating time interference electrical stimulation waveforms in the plurality of stimulation areas so as to stimulate the corresponding stimulation areas and adjusting stimulation parameters; the evaluation module is electrically connected with the image acquisition module and the stimulation module, acquires cognitive emotion evaluation data, and obtains an emotion and cognitive neural circuit regulation result by combining the emotion evaluation data and the function connection data as an inhibition control capability evaluation index; and the control module is electrically connected with the stimulation module and the evaluation module, and controls the stimulation module to adjust the stimulation parameters based on the emotion and cognitive neural circuit adjustment result. According to the system, effective time interference electrical stimulation is carried out on a plurality of inhibition control intervention brain regions, stimulation parameters are improved by implementing stimulation effect evaluation, and a favorable adjustment means is provided for inhibition control of multi-region and complex network modulation.

The invention discloses a transcranial magnetic stimulation positioning system based on a deep effect brain region, and the system comprises the steps: obtaining the data of a tested magnetic resonance image, which comprises a resting state functional image and a high-resolution T1 structure image; performing data preprocessing on the function image and the structure image, and establishing an individual space; the deep effect brain region template is converted to an individual space through nonlinear registration, and a target region of interest is defined; calculating functional connections between voxels of the deep effect brain region and the cortical region by adopting a Granger causality analysis method based on wavelet transformation; and screening Top block masses according to the functional connection strength, calculating the gravity centers of the Top block masses, and determining individualized stimulation targets in combination with the skull surface distance. According to the method, spontaneous brain activity function variation of the brain of an individual and the brain signal propagation direction are fully considered, accurate positioning of different individual levels is achieved, relatively stable cortex stimulation targets are obtained, and therefore the clinical curative effect of TMS is optimized.

The invention discloses a disturbance of consciousness stimulation regulation and control system and method fused with electroencephalogram connection recognition. The system comprises a simulated electroencephalogram signal data acquisition stage, a connection recognition analysis stage, a stimulation parameter optimization stage and an executable stimulation instruction conversion stage. The method has the following advantages and effects that a whole-electroencephalogram activity distribution diagram is generated by simulating an electroencephalogram signal data acquisition stage, a key connection area is identified, and a brain function connection map is generated by utilizing a function connection analysis network and a phase synchronization algorithm in a connection identification analysis stage; in the stimulation parameter optimization stage, space-time correlation between a whole electroencephalogram activity distribution map and a brain function connection map is established, a multi-objective optimization algorithm is adopted to generate an optimized stimulation parameter set through a fusion network, and finally, in the executable stimulation instruction conversion stage, the optimized stimulation parameter set is converted into an executable stimulation instruction based on a self-adaptive control model. Therefore, the accuracy and the self-adaptive capability of electroencephalogram signal stimulation regulation and control are remarkably improved.

The invention belongs to the field of brain-computer interfaces, and particularly relates to a decision-making brain-computer method and device based on virtual reality induction, which combines two psychological decision-making tasks of auditory stimulation and visual stimulation and utilizes virtual reality equipment to create a decision-making brain-computer interface normal form of panoramic interaction, so that a subject can fit a scene facing a decision in reality to the greatest extent, and the accuracy of decision making is improved. The sensory motor cortex is effectively activated, and cooperative activation of the cognitive-motor neural network is induced; a decision interaction feedback link is added to enhance a decision stimulation effect and a cranial nerve feedback mechanism by analyzing related characteristics P300, power spectral density and brain network function connection quantity of the acquired electroencephalogram signals during normal form execution decision reaction. Compared with the prior art, the method has the advantages that immersive audio-visual stimulation and task interaction feedback are brought into a decision-making brain-computer interface for the first time, and a new scheme with neural rehabilitation and human-computer interaction functions is provided for neural feedback training of cognitive functions.

The invention discloses a brain region correlation analysis system and method for autistic children. The system comprises a brain region division module used for dividing the cerebral cortex into a plurality of brain regions; the training data set construction module is used for constructing a training data set, and each training sample comprises a sensor space function connection matrix and a source space function connection matrix corresponding to the sensor space function connection matrix; the signal preprocessing module is used for acquiring a real electroencephalogram signal and calculating a real value of a corresponding sensor space function connection matrix; the deep learning mapping module is used for learning a mapping relation from the sensor space function connection matrix to the source space function connection matrix and outputting a predicted value of the source space function connection matrix; and the brain region correlation analysis module is used for calculating the correlation between the brain regions. The method can be used for accurately analyzing the correlation between the brain areas of the autism children.

The invention is applicable to the technical field of brain discipline, and provides a brain function connection modeling and coding method, device and equipment and a storage medium, the method comprises the following steps: constructing a first brain function connection graph according to rs-fMRI data, performing graph enhancement processing on the first brain function connection graph by adopting a preset adaptive graph enhancement strategy to obtain a second brain function connection graph, performing feature coding on the second brain function connection graph by adopting a preset topological attention coding strategy to obtain corresponding graph features, and performing feature projection on the graph features by adopting a preset feature projection strategy to obtain brain function discrimination features for brain function connection classification, therefore, the generated brain function discrimination features can accurately identify subtle differences of different types of brain function connections, and high-accuracy discrimination can be realized whether normal and abnormal brain function states or brain connection modes corresponding to different diseases, so that powerful support is provided for clinical aid decision making, and the accuracy of brain function discrimination is improved. And the accuracy and scientificity of brain disease diagnosis and treatment can be improved.

The invention discloses a multi-mode autism diagnosis method and system based on resting state fMRI and phenotypic text information, a medium and a product. FMRI and phenotypic text information are processed through a pre-trained heterogeneous graph neural network model to output a diagnosis result; the resting state fMRI is utilized to construct a brain function connection graph, and the phenotypic text information is coded into a two-dimensional vector; respectively extracting brain connection feature embedding of the fMRI mode and phenotype information feature embedding of the text mode; fusing the feature embedding in the two modes through a gating fusion network; and constructing a heterogeneous group diagram based on the fusion features and phenotypic information similarity, and obtaining group diagram features on the heterogeneous group diagram by using dual-channel information aggregation and adaptive feature fusion to output a diagnosis result. The invention aims to improve the accuracy of autism diagnosis by using multi-modal information, and can be applied to the fields of neuroimaging analysis, intelligent medical treatment and the like.

The invention provides a network graph generation method based on brain image data, and relates to the technical field of brain connection.The method comprises the steps that brain image data (including fMRI data and sMRI data) of a target object are obtained, the brain image data are preprocessed and registered, and registered brain graph data are obtained; dividing the brain map data into a plurality of brain regions based on a set brain map, and determining time sequence data of each brain region; based on the time sequence data of each brain region, a function connection matrix is determined, and non-diagonal elements in the function connection matrix reveal the connection strength between the brain regions; and based on the brain region and the functional connection matrix of the brain map data, generating a brain network map (which can be visualized) of the target object. According to the scheme, the sMRI data (the precision is millimeter level) is introduced as a registration basis, so that the registration precision can be improved, and a brain network diagram with higher precision can be generated subsequently.

The invention belongs to the technical field of artificial intelligence and physiological signal processing, and discloses an electroencephalogram emotion recognition method and system based on a multi-task self-supervision and dynamic graph fusion network, and the method comprises the steps: obtaining and preprocessing electroencephalogram and other physiological signals, and extracting multi-band energy features; constructing a dynamic graph structure, taking electrodes as nodes, taking frequency band energy as characteristics, and fusing spatial distance and functional connectivity to generate a dynamic adjacency matrix; designing multi-task self-supervised pre-training, including spatial jigsaw, frequency jigsaw and cross-modal contrast learning tasks, to learn general characterization; a dynamic graph fusion network is adopted to carry out end-to-end training, and a shared feature extraction module of the dynamic graph fusion network realizes adaptive fusion of multi-modal features by utilizing Chebyshev graph convolution and embedding a cross-modal attention mechanism; the classification module sets an independent classification head for each task, and optimization is carried out through a joint loss function. According to the method, the accuracy and generalization ability of electroencephalogram emotion recognition are remarkably improved.

The invention discloses a brain function magnetic resonance imaging data analysis method based on a contrast graph neural network, and the method comprises the steps: firstly carrying out the data enhancement of a brain function connection graph, and simulating the heterogeneity of brain function magnetic resonance data, so as to improve the diversity of a data set; secondly, the hidden space embedding features of the brain function connection diagram are efficiently learned by fusing a double-Hough-Laplacian diagram convolutional network and a contraction incentive mechanism; the embedded features are mapped to a group of prototype vectors, and prototype allocation codes corresponding to the embedded features are calculated by adopting a Sinkhorn-Knopp algorithm; performing exchange optimization on prototype codes between different enhanced brain connection diagrams of the same subject through a contrast learning strategy, and compelling codes of homologous subjects to be aligned at the minimum cost in combination with a cross entropy loss function; and finally, applying the pre-training model to a functional magnetic resonance imaging data set of the Alzheimer's disease, and carrying out interpretability analysis on learning features to improve the classification efficiency and pathological analysis of the Alzheimer's disease under a limited tag condition.

The present invention belongs to the field of rehabilitation therapy technology and discloses a longitudinal analysis method and system for magnetic resonance imaging data of mild brain injury. The method extracts BOLD signals from the magnetic resonance imaging data of subjects; constructs a symmetric positive definite sparse brain functional connectivity network set of subjects based on sparse inverse covariance matrix estimation; determines the brain functional connectivity network dictionary and sparse coefficient matrix in kernel space based on Riemannian manifold sparse coding; performs spatial distribution analysis of brain functional connectivity atomic networks; and performs longitudinal analysis of magnetic resonance imaging data of mild brain injury. By analyzing the differences in the spatial distribution of these highly present brain functional connectivity atomic networks in the brain, the present invention digs out brain functional connectivity imaging markers for distinguishing the three mild brain injury rehabilitation treatment stages: acute phase, subacute phase, and complete recovery, thereby realizing longitudinal analysis of the mild brain injury rehabilitation process.

The invention discloses a nerve regulation and control method, device and equipment based on electrical stimulation and a storage medium. The method comprises the steps that metabolism image data, functional image data and anatomical image data of a detected person are acquired; based on the metabolism image data and the functional image data, respectively calculating the standardized uptake value ratio of each brain region of the detected person and the functional connection strength between each brain region and all other brain regions; on the basis of the standardized uptake value ratio of each brain region of the detected person and the functional connection strength between each brain region and all other brain regions, determining the abnormal characteristics of the brain of the detected person and the position data of a target spot; and time domain interference electrical stimulation parameters are set based on the position data of the target spot, the brain abnormal characteristics of the detected person and the anatomical image data. According to the method, the PET image data and the rs-fMRI image data can be combined and analyzed, the metabolism and function state of the brain of the detected person can be evaluated more comprehensively, and corresponding nerve regulation and control can be carried out based on the time domain interference electrical stimulation parameters.

The invention discloses a method for detecting interpersonal nerve synchronization under audio-visual stimulation, which comprises the following steps of: designing a double-person super-scanning experiment normal form, collecting double-person electroencephalogram signals, and constructing a database of visual stimulation and auditory stimulation; preprocessing the data in the database to obtain electroencephalogram signals of four different frequency bands delta, theta, alpha and beta, extracting the electroencephalogram signals of the alpha frequency band, segmenting the electroencephalogram signals, and calculating a correlation ISC value between subjects of each segment; an intra-brain network and an inter-brain network are constructed by constructing a functional connection matrix, and the similarity of the intra-brain network and the global efficiency of the inter-brain network are calculated, so that the cooperation and synchronization degree between neural activities of subjects under visual stimulation and auditory stimulation is effectively evaluated. The invention further discloses a system for detecting interpersonal nerve synchronization under audiovisual stimulation. According to the method, the difference of subjects is reduced by standardizing experimental conditions, and the influence of single sensory stimulation on an intracerebral network activation mode is studied.

The invention discloses a three-dimensional brain network dynamic segmentation method based on deep learning. The method comprises the following steps: S1, constructing fused image volume data with consistent time and consistent space; s2, a multi-scale sparse Transform coding feature pyramid is generated; s3, obtaining a same-scale brain region map structure; s4, taking the cross-scale node alignment fusion graph structure as initial output of a cross-scale node alignment fusion mechanism; s5, obtaining a first round of fusion brain region graph node embedding feature; s6, an updated multi-scale sparse Transform coding feature pyramid is obtained, and the step S3 to the step S5 are repeated until interaction updating of the multi-scale sparse Transform-GNN is completed; and S7, obtaining a three-dimensional brain region dynamic segmentation result. According to the method, collaborative extraction of local fine granularity and global coarse granularity features is realized, redundant information can be effectively inhibited, and the modeling capability of spatial structure and functional connection features in a cross-modal fusion image can be enhanced.

The invention provides a working memory ability assessment method and device for dynamic brain network attention fusion, and belongs to the technical field of intelligent medical treatment, and the method comprises the steps: segmenting a BOLD signal time sequence in functional magnetic resonance imaging data, extracting time sequence data of a region of interest, and constructing a functional connection matrix; extracting and fusing spatial dependence and time fluctuation characteristics in the functional connection matrix based on an attention mechanism to obtain a full connection graph; clustering the full connection graph by using multi-head attention to obtain a functional magnetic resonance imaging embedded representation of each testee; training the prediction network by using the functional magnetic resonance imaging embedding representation of the testee; and using the trained prediction network to complete work memory evaluation of the target subject. According to the method, dynamic multi-graph fusion and neuroscience priori knowledge are combined, individual differences and specific noise are eliminated through comparative learning, accurate evaluation of the working memory ability is achieved, and an objective basis is provided for evaluation of the cognitive function.

The embodiment of the invention discloses a brain deep part electrical stimulation operation planning system and method based on a human brain network and a Lead-DBS technology, and the system and method are applied to operation planning of neuropsychiatric diseases. The data preprocessing module is used for denoising, correcting and standardizing the patient data to generate preprocessed data; the brain network model construction module is used for determining a functional connection matrix and a structural connection matrix based on the preprocessed data so as to construct an individualized brain network model; the electrode positioning module is used for generating an implantation target spot of a DBS electrode based on an individualized brain network model; the stimulation parameter optimization module optimizes stimulation parameters of the DBS based on the individualized brain network model and the electrophysiological data; the postoperative evaluation module compares the postoperative image with the preoperative plan to verify the actual implantation position and stimulation effect of the DBS electrode; the dynamic adjustment module is used for dynamically adjusting stimulation parameters.

The invention discloses a fatigue state prediction method, device and system based on multi-modal data and a storage medium, and relates to the technical field of electroencephalogram signal processing, and the method comprises the following steps: constructing a time sequence connectivity filtering and dynamic brain region enhancement model; wherein the time sequence connectivity filtering and dynamic brain region enhancement model comprises a frequency band division band attention module, a time sequence connectivity filtering module and a dynamic region enhancement module; on the basis of the input electroencephalogram data, channel enhancement features are obtained by using a frequency-band-division band attention module; processing the channel enhancement feature by using a time sequence connectivity filtering module to obtain a first feature value; processing the channel enhancement feature by using a dynamic region enhancement module to obtain a second feature value; and fusing the first characteristic value and the second characteristic value, and outputting a prediction result through a full connection layer to obtain a fatigue state prediction result based on the multi-modal data. According to the invention, the dynamic topology mode of the brain area in the fatigue state and the time sequence change of the functional connection strength can be revealed.

The invention relates to the field of medical image analysis and neuropsychiatric disease diagnosis, and particularly discloses a brain network typing diagnosis system and method for attention deficit hyperactivity disorder. Comprising a data acquisition and preprocessing module, a topological feature extraction module, a supervised manifold learning module, a network reconstruction and diagnosis module, a parameter optimization module and a result verification module, continuous homology analysis is performed on a brain region function connection matrix based on an algebraic topology theory, and a brain region topological feature matrix is generated; utilizing supervised manifold learning to map the brain region topological feature matrix to a Riemannian manifold, and calculating a brain region importance weight map based on ADHD phenotypic features as supervised signals; according to the method, the brain network is subjected to subtype specific reconstruction through the curvature flow theory of Riemannian geometry, the topological difference between different subtypes is calculated, accurate typing diagnosis of ADHD is achieved, and the diagnosis accuracy is improved by 15%-20%.

The invention relates to the technical field of brain anomaly detection and artificial intelligence auxiliary diagnosis, in particular to a graph convolutional network brain disease diagnosis method based on sub-graph sampling and multi-feature fusion, and aims to improve the accuracy of brain disease diagnosis. The method comprises the following steps: obtaining resting state functional magnetic resonance imaging data, preprocessing the data, and constructing a brain function connection diagram; and the brain function connection graph represents a brain interval collaborative activation relationship in a graph structure. Subgraph sampling is carried out based on function module division and node degree sorting, and an initial subgraph set is generated; and performing optimization selection on the initial sub-graph set by utilizing reinforcement learning to obtain an optimal sub-graph, introducing a node attention mechanism into the optimal sub-graph, screening key nodes based on attention scores, and generating a discriminant sub-graph. And extracting and fusing position features, neighborhood features and structural features of the discriminant subgraphs, and performing brain disease diagnosis based on the fused features.

The invention discloses a personalized electroencephalogram emotion recognition method and system based on dynamic brain region division, and belongs to the technical field of electroencephalogram emotion recognition. The method comprises the steps that electroencephalogram signals are preprocessed, and an electroencephalogram signal matrix is constructed; on the basis of the preprocessed electroencephalogram signals, five-frequency-band differential entropy feature values are extracted, multi-head self-attention calculation is carried out in combination with rotation position coding, and channel feature representation with strong emotion intensity is output; on the basis of the constructed electroencephalogram signal matrix, dynamically dividing a brain function region by using a CNM algorithm, optimizing the structure of the brain function region by using a modularity value, and outputting brain region characteristics; and fusing the channel feature representation and the brain region features, predicting an emotion tag through a graph neural network, and outputting an emotion recognition result. According to the method, the individual brain function region can be dynamically divided, the function integration information of the brain region is effectively captured, and deep fusion of brain region function connection and channel emotion information is realized.

The invention relates to an autism classification method based on a multi-scale residual image neural network, and aims to cope with the challenge of crowd autism classification in multi-modal medical data. The method comprises the following steps: firstly, providing a new function connection feature construction method, extracting second-order function connection features by using tangent Pearson embedding to capture a high-order interaction relationship between brain intervals, and then adopting a maximum independent domain to adaptively minimize statistical dependence between the features and acquisition sites, and combining F-score to screen the features with the most discriminative ability, so as to obtain the feature with the most discriminative ability. And redundancy is effectively removed. Secondly, a multi-modal edge weight calculation method fusing imaging information and non-imaging information is provided, so that noise interference is effectively suppressed while a key discriminant relation is reserved. And finally, expanding a node receptive field layer by layer by stacking multiple layers of Chebyshev convolutions with residual errors on the subject graph so as to capture multi-level relation characteristics, and performing weighted modeling and adaptive fusion on convolution output of each layer by using a multi-head self-attention mechanism, so that effective integration of multi-scale information is realized, and accurate classification of autism is realized. The method has excellent performance in the aspect of autism classification, and an innovative, feasible and effective solution is provided for solving the autism classification task in the multi-modal medical data.

The invention relates to an epilepsy electroencephalogram detection method adopting multi-modal feature fusion and an attention mechanism, and belongs to the technical field of medical signal processing and artificial intelligence. According to the method, automatic detection of epilepsy is realized through three core modules: firstly, preprocessing and time-frequency transformation are performed on electroencephalogram signals, a time-frequency graph is generated, and power spectrum characteristics of different frequency bands are extracted; the method comprises the steps that firstly, a time-frequency diagram and a function connection diagram are processed at the same time through a multi-modal feature combined extraction network, and finally fusion enhancement and classification decision making of multi-modal features are achieved through a self-adaptive feature fusion and classification network in combination with a learnable frequency band attention mechanism and a Transform encoder. According to the method, the time-frequency characteristics, the brain region connection relation and the long-range dependency relation of the electroencephalogram signals can be captured at the same time, the limitation of a traditional method in the aspects of feature expression and mode recognition is solved, and the accuracy of epilepsy detection is remarkably improved.

The invention discloses a federated learning-based functional magnetic resonance imaging data classification method. The method comprises the following steps of obtaining multi-site resting state functional magnetic resonance imaging data and performing preprocessing; dividing a brain region and extracting a time sequence; constructing a brain function connection network by adopting a Pearson correlation coefficient; a graph sampling aggregation neural network fusing a residual connection structure and a multi-head attention mechanism is trained at each site, and shallow brain region features are reserved; adopting a linear kernel maximum mean value difference loss function to align the brain region node feature distribution of each site, and minimizing the data distribution difference between the sites; and carrying out classification training by adopting cross validation, aggregating parameters of each station through federal weighting, and evaluating classification performance indexes to obtain a final classification prediction result. According to the method, the graph sampling aggregation neural network and the cross-network layer feature alignment method are combined, the heterogeneity problem of multi-site functional magnetic resonance imaging data can be solved, and therefore generalization and classification performance of a global model are improved.