FMRI dynamic brain function sub-network construction and parallel connection SVM weighted recognition method

Abstract

The invention discloses an fMRI dynamic brain function sub-network construction and parallel connection SVM weighted recognition method which comprises the steps that (1) data are preprocessed; (2) the time series of each brain area is extracted; (3)interested brain areas are selected; (4) dynamic brain function sub-networks of all the brain areas are constructed; (5) all the sub-network classifiers are trained; (6) values are assigned to all sub-classifiers to form parallel connection SVM classifiers; (7) unknown samples are classified. Compared with a traditional static function network, information on the time dimension is added on the constructed dynamic brain function networks; prior knowledge is combined for constructing dynamic sub-networks on different interested brain areas, and the feature dimensions are reduced while useful information is reserved; SVM classifiers of all the sub-networks are trained, the parallel connection SVM classifiers is formed by determining the weight of the sub-classifiers through the recognition rate, the brain areas are integrally weighed and classified, and the classifiers have better robustness.

Description

technical field [0001] The present invention relates to a method for processing and identifying brain functional magnetic resonance images, in particular to fMRI dynamic brain function sub-network construction and parallel SVM weighted identification methods; through digital image processing, pattern recognition, machine learning and other fields of knowledge, and Using processing methods such as dynamic network, principal component analysis and support vector machine, a method for identifying and extracting features of functional magnetic resonance image data is designed, and the processing results can be used to classify functional magnetic resonance image data. technical background [0002] In recent years, functional Magnetic Resonance Imaging (fMRI) technology has attracted extensive attention of researchers. The technology provides a non-invasive way to obtain images of brain activity by detecting changes in blood oxygen saturation levels in the brain to observe change...

AI Technical Summary

Problems solved by technology

[0006] Aiming at the problems that the fMRI brain function network only contains the activity information in the spatial dimension of the brain area in the prior art field, ignoring its non-stationary characteristics in the time dimension, the data dimension of the brain function network is too high, and the recognition accuracy of the classifier is not high. The invention provides a fMRI dynamic brain function sub-network construction and parallel SVM weighted identification method, using the fMRI dynamic network to obtain more abundant brain image information, and classifying the fMRI data through the parallel SVM classifier to solve the problem of reducing data dimension, While effectively extracting features, it also improves the robustness of the classifier

Images