“Machine learning enabled to predict human movements from fMRI signals: contribution of the posterior parietal cortex for kinematic transformation.”
Dr. Izawa, an associate professor of EMP, predicted the wrist movement by fMRI decoding technique in a collaboration among ATR and Tokyo Metropolitan Institute of Medical Science. This research team measured the BOLD signal from the brain while conducting repetitive wrist movements and decoded the multi-voxel activities to predict the ongoing movement by using a machine learning technology (SVM). Using this method, they estimated the coordinate frames of each region in the brain that encodes movement-related information and suggested that PPC is responsible for transforming the coordinate frames. While this result is along with the previously reported results in monkey studies, this is first to report in human studies. This result was published in SCIENTIFIC REPORTS.