The Modulation Of Methylphenidate And Real-time Feedback On Brain Functional Network

Medical and neuroimaging researches has revealed that both functional separation and functional integration co-exist in brain when specific tasks are performed, thus the majority of mental and neurological diseases are not only related to the activity of single brain region but also influenced significantly by the interregional connection. Regarding brain as an integrated network in which one region interacts and exchanges information with others will provide an effective direction for new insight into the brain, and investigating different technologies of brain functional network regulation will open a new path to explore the pathology and treatment of brain diseases. So far the regulatory mechanism of many psychotropic drugs is still not clear, and the effect of real-time neuro-feedback on specific brain network is yet to be discussed.In recent years, functional Magnetic Resonance Imaging (fMRI) has become one of the most important means of brain network research, due to its advantages of non-invasiveness and high spatial resolution. Therefore, based on fMRI, this work aims to study the regulation mechanism of drugs and the effect of real-time neuro-feedback on brain network. Specifically, we first applied dynamic causal modeling (DCM) to analyze resting state fMRI data of volunteers who underwent methylphenidate (MPH) and placebo scans. As the outcome, we pinpointed the connectivity of default mode network mainly affected by methylphenidate through comparison of scans before and after taking MPH, and explored MPH modulation on the effective connectivity strength and the excitation / inhibition state combining the DCM parameters with the T test results, thereby probed into the specific mechanism of MPH adjusting the DMN and influencing the behavioral expression. Secondly, we set up real-time fMRI protocols to realize online reconstruction and transmission, data processing, statistical analysis and real-time feedback, verified its effectiveness in monitoring data quality and applications of brain-computer interface We further designed four kinds of negative emotion training exercise in order to select the optimal support vector machine classifier via cross-validation, and then obtained the real-time classification result of three times emotional regulation based on this neuro-feedback mechanism. Combining statistical classification accuracy with offline activation maps, we analyzed the change of ability to regulate and control the functional emotion network in volunteers after repeated practice with the real-time feedback, moreover, explored the potential clinic value of neuro-feedback in studies such as post-traumatic stress disorder and depression therapy.