Abstract: Gait impairment seriously affects the motor ability and life quality of patients with Parkinson’s disease (PD) with less effective clinical treatment relative to other usual symptoms. Excessive β burst activity in the subthalamic nucleus (STN) has been reported as critical to motor impairment, while auditory cues can improve gait performance. However, whether local field potential (LFP) recordings from the STN can decode the brain state for PD during steps, meanwhile reflecting the effect of rhythmic auditory cues remains unclear. To decode the brain dynamics and examine the effectiveness of auditory cues for gait regulation in PD. Eight PD patients undergoing bilateral STN implantation of deep brain stimulation (DBS) electrodes were included, and a decoder integrating masking empirical mode decomposition (MEMD) and hidden Markov model was proposed to analyze their LFPs during stepping with or without sound cues. β bursts were significantly suppressed during and after cues, which was manifested in the decrease of their fractional occupancy (FO) and the shorter state lifetime (LT). However, α bursts show the opposite, with the FO and LT of α burst activity increased during and after cues. Our results support the validity of MEMD-HMM for brain-state decoding with varying gait and cues. Rhythmic auditory cues promote gait performance, and the fluctuation of temporal features (FO and LT) in multiple frequency bands under sound conditions are reflected by state transitions. These results could provide feedback guidance for closed-loop DBS and facilitate future implementation of a brain-computer interface to treat PD gait problems.

Keywords: brain-states decoding, auditory cues, Parkinson’s disease, masking empirical mode decomposition, hidden Markov model

Acknowledgements: This work is supported by the National Natural Science Foundation of China (Grant No. 62001026, No.62171028), and the BIT High-level Fellow Research Fund Program (No. 3050012222022).

Biography: Chien-Hung Yeh, Ph.D., is a Professor at the Beijing Institute of Technology in China. Prior to this, he worked at the University of Oxford in the UK and Harvard Medical School in the State, respectively. Dr. Yeh is a senior member of IEEE, CIE, and BME, as well as serves on several committees at the Chinese Association for Artificial Intelligence, the Chinese Society of Image and Graphics, and the Chinese Institute of Electronics. His works lie on electrophysiology, nonlinear cross-coupling, and complex systems, acting on topics of movement disorder, emotion or mental illness, sleep, and cardiovascular disease, with application in brain-computer interfaces and wearable devices. His works system elaborated on how the nonlinear oscillations in multiple frequencies entangled, as well as its contribution to providing new insight into human electrophysiology.