Fractal characteristics-based motor dyskinesia assessment. (July 2021)

• Record Type: Journal Article
• Authors: Zhao, Kunkun; Wen, Haiying; Zhang, Zhisheng; He, Chuan; Wu, Jiankang
• Is Part Of: Biomedical signal processing and control. Volume 68(2021)
• Publication Date: 2021-07
• Subjects: Fractal dimension -- Maximum fractal length -- Motor dyskinesia -- sEMG -- Time-frequency features; Signal processing -- Periodicals; Biomedical engineering -- Periodicals; Signal Processing, Computer-Assisted -- Periodicals; Image Processing, Computer-Assisted -- Periodicals; Biomedical Engineering -- Periodicals; 610.28
• DOI: 10.1016/j.bspc.2021.102707
• Languages: English
• ISSNs: 1746-8094

Abstract

Highlights: Apply fractal characteristics to assess motor dyskinesia. Statistical-based method for feature selection. Fractal dimension and maximum fractal length perform better than time-frequency features. Fractal characteristics capture the disorder of central nervous system. Abstract: Goal: The purpose of the study was to evaluate the feasibility of fractal characteristics (FCs) in the motor dyskinesia assessment. Methods: The study recruited twenty stroke patients and twenty age-matched healthy subjects. Ten upper limb surface electromyography (sEMG) signals were measured when participants performed Fugl-Meyer assessment of upper limb, reaching movement, and hand-to-mouth movement. Two fractal characteristics extracted from sEMG, fractal dimension (FD) and maximum fractal length (MFL) were compared with five time-frequency features (TFFs), root mean square, mean absolute value, variance, waveform length, and median frequency in motor dyskinesia assessment. A statistical-based feature selection method was proposed to select a feature subset. The identified features were then input into the Gaussian process regression (GPR) model to estimate the motor dyskinesia level. Results: A small set of features was selected applying the proposed method. Two fractal characteristics showed consistency among movements, especially MFL, which could capture the disorder of the central nervous system (CNS). Comparing to the TFFs, the FCs are better for motor dyskinesia assessment. Conclusions: The results demonstrate the feasibility of FCs to assess motor dyskinesia. Fractal-based method provides new insight into motor dyskinesia assessment.