| The human brain is the most complex and sophisticated machine in the world.Although human beings still know little about the brain,people are still exploring the mysteries of the brain,and have achieved some results.At present,the functional magnetic resonance imaging(f MRI)of the brain has become one of the main methods to study the work of the human brain.At the same time in neuroscience and cognitive science and brain science and other fields,exploration is no longer confined to the human brain,motion perception,auditory and visual aspects,but extends to the research on human cognition and thinking,emotional and psychological aspects.At the same time,functional magnetic resonance imaging(functional magnetic resonance imaging)is mainly used in clinical medical diagnosis.In this paper,the method of f MRI brain functional network analysis is studied.The specific work is as follows:First,by studying the Fast ICA algorithm for f MRI data preprocessing,an improved Fast ICA algorithm based on the modified whitening matrix is proposed.The simulation data of the simulated f MRI data produced by the Sim TB toolbox is simulated by Matlab software,which confirms the effectiveness of the two algorithms for separating the original f MRI signal.At the same time,the estimation accuracy compared to the conventional Fast ICA algorithm and the improved Fast ICA,the analysis of signal de-noising accuracy evaluation index is the mean square error(RMSF),the signal-to-noise ratio(SNR)to verify,confirm the source signals are separated by using Fast ICA algorithm improved the accuracy of the value is indeed improved,this is the f MRI data processing foundation.Secondly,in the traditional way of constructing resting state f MRI brain functional connectivity network,SVD matrix decomposition is introduced to further denoise the resting state f MRI data,and the number of singular values of SVD matrix is selected to reconstruct 11 signals.In the construction of brain network function links,the threshold is first set as interval of 0.01 to calculate and analyze the brain network characteristics.Then we restrict the threshold value by restricting two threshold conditions of 15%~45% brain network density and the small world of brain network,so as to ensure the integrity of the constructed brain network.Finally,basedon the selected threshold value of 0.28,we constructed the visualized graph of the adjacency matrix of the brain network,compared and analyzed the difference of the brain network before and after SVD,and the analysis results showed the correctness of the method adopted.Finally,on the basis of the previous research,aiming at the f MRI data of the task state based on the left and right hand imagination movement,we use the currently researched ant colony clustering algorithm to classify the communities.The results of the study showed that the left and right hand task state brain network community is divided into seven communities,different societies are involved in motor imagery tasks,such as,responsible for processing motion information for the back;responsible for processing the image shape feature information consists of two regions: the occipital cortex around the back,which is mainly responsible for the dynamic shape Yan,the anterior cingulate gyrus,cingulate gyrus,and adjacent posterior cingulate board.Receiving and processing external stimulus information,transferring the processed content to the advanced processing area,processing the brain's right occipital gyrus,processing the brain's motor information,processing and processing advanced visual information.The following is the sixth brain regions within the community: the posterior cingulate and adjacent cingulate gyrus and posterior cingulate gyrus,cingulate and adjacent to the anterior cingulate gyrus,its function is almost equivalent to second communities in brain regions.The study lays the foundation for the function of various brain regions and the joint cooperative work mechanism in the task state of the brain... |
PDF Full Text Request