Myoelectric pattern recognition (PR) based strategies have been well investigated and applied as viable control methods for upper limb prostheses.  

However, transhumeral amputees often lack requisite residual muscles to produce high-quality electromyogram (EMG) signals which is essential for intuitively control their prosthetic device. The limited acquired signal from these amputees often contains noises that make it challenging to accurately decode their limb movement intent, which is essential to realize intuitive and natural control.  

A researcher group led by Prof. CHEN Shixiong and Assoc. Prof.  Oluwarotimi Williams SAMUEL from the Shenzhen Institute of Advanced Technology (SIAT) of the Chinese Academy of Sciences, provides an optimized filtering technique for high-density surface EMG signal quality enhancement towards precise motion intent decoding in the context of upper limb prostheses control.  

Their study are reported in the Biomedical Signal Processing and Control on Jan. 18. 

The study explored the capability of optimized filtering techniques (including Wiener, Hampel, and 1-Dimensional Median filters) in denoising transhusmeral amputees’ EMG signal towards improving its overall quality in a manner that would yield adequate PR-based control for prosthetic devices.  

The schematic diagram of the reported method. (Image by SIAT) 

The performance of the filtering technique was extensively investigated by using myoelectric signals obtained from transhumeral amputees across different features-classifier combination under different experimental conditions. 

Experimental results with statistical test showed that the reported filtering technique led to improved EMG signal quality that resulted in consistently high decoding performance for the amputees’ limb movement intents. 

This study strongly suggests that EMG-PR based multifunctional prostheses has potential for improving the clinical and commercial robustness. 

Media Contact:
ZHANG Xiaomin
Email:xm.zhang@siat.ac.cn