Central Pattern Generator Based Band-Limited Multiple Fourier Linear Combiner for Tremor Estimation in Functional Electrical Stimulation System

摘要

Central pattern generator (CPG) is a useful mathematic model to describe the periodic motion of human. As the pathological tremor can be approximated as roughly periodic and sinusoidal motion signal, the CPG based algorithm for tremor signal processing exhibits great potential. However, CPG for pathological tremor suppression has rarely been applied in real time implementation. In this work, a new algorithm named central pattern generator based band limited multiple Fourier linear combiner (CPG-BMLC) for tremor suppression in functional stimulation system (FES) is proposed. The proposed algorithm is employed to process the signal collected from the patient with Parkinson tremor. The stimulation results have shown that CPG-BMLC can predict the motion signal and filter off the high frequency tremor accurately. In addition, the algorithm is programed in FES for tremor suppression in real time. The experimental results verify that the CPG-BMFLC in FES can suppress frequency tremor signal in real-time implementation. The CPG-BMLC in FES for tremor suppression can extract the tremor part from motion and suppress the tremor accurately.