Muscle Activation Patterns of the Forearm: High-Density Electromyography Data of Normally Limbed and Transradial Amputee Subjects

Abstract

In Brief For many years, myoelectric prostheses have been accepted by upper limb amputees. Multifunction pattern recognition has been used successfully to develop myoelectric-controlled limbs; however, much of this research has used normally limbed subjects as they are capable of producing distinguishable muscle activation patterns for different movements. Surprisingly, there has been limited research using clinical populations to test advanced devices and control systems. One reason for the lack of studies with amputees is the limited data regarding their ability to produce repeatable muscle activation patterns. Although it would be expected that these patterns would differ between amputees and normally limbed individuals, it is worthwhile to determine whether amputees can produce repeatable muscle patterns. Therefore, the purpose of this study was to look for distinct and repeatable muscle activation patterns (required for pattern recognition myoelectric control) using amputee subjects. A high-density electromyography system was used to investigate muscle activation patterns during wrist and hand motions. Subjects performed repetitions of different movements, so that intertrial and intermovement electromyography activity comparison could be completed within subject. Topographical (energy maps), classification accuracies, and a separability index were used to determine the discriminate power of the movements performed. It was found that amputee subjects are able to elicit distinct and repeatable patterns for at least a subset of the movements, suggesting promise for amputee control of multifunction myoelectric prostheses. This study aimed to determine if transradial amputees could produce distinct and repeatable muscle activation patterns required for successful pattern recognition myoelectric control. These preliminary results are promising and indicate there is a future for multifunction pattern recognition based control systems in the clinical setting. The results from this work support the hypothesis that amputee subjects are able to elicit distinct and reproducible patterns for a subset of movements.