An FFT-based Synchronization Approach to Recognize Human Behaviors using STN-LFP Signal

摘要

Classification of human behavior is key to developing closed-loop Deep BrainStimulation (DBS) systems, which may be able to decrease the power consumptionand side effects of the existing systems. Recent studies have shown that theLocal Field Potential (LFP) signals from both Subthalamic Nuclei (STN) of thebrain can be used to recognize human behavior. Since the DBS leads implanted ineach STN can collect three bipolar signals, the selection of a suitable pair ofLFPs that achieves optimal recognition performance is still an open problem toaddress. Considering the presence of synchronized aggregate activity in thebasal ganglia, this paper presents an FFT-based synchronization approach toautomatically select a relevant pair of LFPs and use the pair together with anSVM-based MKL classifier for behavior recognition purposes. Our experiments onfive subjects show the superiority of the proposed approach compared to othermethods used for behavior classification.