Research On Parkinson’s Tremor Extraction Algorithm Based On Fourier Linear Combiner

Parkinson’s tremor is the main clinical manifestation of Parkinson’s disease,yet physiological tremor severely limits the extraction of Parkinson’s tremor,which in turn can affect the treatment of Parkinson’s disease patients.In this regard,suitable algorithms are needed to predict physiological tremor and filter it out from the detected tremor as the main noise,in order to improve the accuracy of Parkinson’s tremor extraction.With the development of adaptive filters,the Fourier Linear Combiner(FLC)based approach to physiological tremor prediction based on the Least Mean Square(LMS)algorithm has taken a prominent place.Since adaptive filters cannot simultaneously combine faster convergence with lower steady-state errors,there is still much room for improvement in the predictive estimation of physiological tremor using this class of algorithms.In this paper,relevant improvements are made to the existing algorithms in terms of both variable step size ideas and dual system combinations,and the main research work is summarized as follows:First,starting from the research background and significance of Parkinson’s tremor,the current status of research on predictive estimation of physiological tremor signals is introduced,different characteristics of tremor signals and extraction strategies of Parkinson’s tremor signals are described,and the implementation principles of three common predictive estimation algorithms for physiological tremor are illustrated.Then,a detailed introduction was given to the reason why the LMS algorithm cannot simultaneously consider the faster convergence speed and lower steady-state error when predicting physiological tremors based on FLC.Optimization and improvement were made to address the problems of the single system LMS algorithm: a third-order weight coefficient iteration formula was used instead of the traditional weight coefficient iteration formula to accelerate the convergence speed of the filter;the idea of variable step size is introduced instead of the traditional fixed The proposed improvement is demonstrated through comparative experiments.The proposed improved algorithm has been shown to reconcile the conflict between fast convergence and low steady-state error to a certain extent,and to improve the accuracy of Parkinson’s tremor extraction through comparative experiments.Next,the paper then addresses the problems of poor tracking and computational complexity of the dual system algorithm by improving it in two ways.The first is to address the poor tracking performance of the fixed step size used in both systems by replacing the large step size factor with a dynamically adjustable step size factor;the other is to address the large computational complexity of the Sigmoid function by introducing a new function and modifying it to obtain a new expression for the combination factor.The simulation results show that the improved algorithm has better overall performance for extracting Parkinson’s tremor.Finally,the research work in the full text is summarized and trends in research on extracting Parkinson’s tremor are looked forward to.