Study On Image Processing And Applications Of Resting State FMRI

Abstract:Functional Magnetic Resonance Imaging (fMRI) is a safe, non-invasive brain research tool. Resting state fMRI is suitable for the study of clinical diseases, especially neurological diseases, since it can observe the brain spontaneous activity status with simple experimental design, simple operation and without specific tasks. In this paper, SCA-3and SPG-4which are both genetic neurological diseases were used as the research objects, since the patients with them have the similar clinical symptoms. We study the brain functional effects of these diseases by resting state fMRI in order to provide clues for these diseases in Neurology.Firstly, we study the preprocessing of resting state fMRI analysis including automatic segmentation of structure images and elimination of physiological noise. A new automatic segmentation method is proposed in order to solve the problems about edge leakage and noise. An image set from a simulated brain database and a real brain MR image was tested to validate the accuracy of the proposed method. The experimental results show that the proposed method can detect the edge accurately, and reduce the effect of noise. We also analyze the effect by the physiological noise caused by cardiac and respiration. Physiological noise was modeled by RETROICOR method and tested through the simulation experiment. The experiment result shows the method is able to reduce the effect of physiological noise.Then, the brain fMRI data under the resting state of healthy people and the patients with different genetic neurological diseases are acquired, and a statistical analysis of the data with frequency between0.01-0.08Hz is performed by using Regional Homogeneity (ReHo) method and Amplitude of low frequency fluctuation (ALFF) method. The statistical result shows that, comparing with the control group, there is a significant difference in ALFF and ReHo in both groups of patients’brain:left Insula, left middle occipital and other regions; however, there is also a significant difference in left hippocampus, left posterior cingulate and other regions of brain in the statistic of these two diseases. Meanwhile, there is a consistency in the results got by both ReHo and ALFF methods, such as in right cerebellum, left putamen and other regions, which implies some significant regions of brain. What’s more, the broad distribution of abnormal brain region in the analysis result implies that the neurophysiological mechanism of these two types of disease results from the abnormal of the Brain function network most probably, instead of the abnormal of a single brain region. Hence, it demonstrates the validity of the study of SCA-3and SPG-4using the method of resting state fMRI.