A training strategy for learning pattern recognition control for myoelectric prostheses

Michael A. Powell, Nitish V. Thakor

Research output: Contribution to journalArticle

Abstract

Pattern recognition-based control of myoelectric prostheses offers amputees a natural, intuitive way of controlling the increasing functionality of modern myoelectric prostheses. Although this approach to prosthesis control is certainly attractive, it is a significant departure from existing control methods. The transition from the more traditional methods of direct or proportional control to pattern recognition-based control presents a training challenge that will be unique to each amputee. In this article, we describe specific ways that a transradial amputee, prosthetist, and occupational therapist team can overcome these challenges by developing consistent and distinguishable muscle patterns. A central part of this process is the use of a computer-based pattern recognition training system with which an amputee can learn and improve pattern recognition skills throughout the process of prosthesis fitting and testing. We describe in detail the manner in which four transradial amputees trained to improve their pattern recognition-based control of a virtual prosthesis by focusing on building consistent, distinguishable muscle patterns. We also describe a three-phase framework for instruction and training: 1) initial demonstration and conceptual instruction, 2) in-clinic testing and initial training, and 3) at-home training.

Original languageEnglish (US)
Pages (from-to)30-41
Number of pages12
JournalJournal of Prosthetics and Orthotics
Volume25
Issue number1
DOIs
StatePublished - Jan 1 2013

Keywords

  • Pattern recognition
  • motor learning
  • myoelectric prosthesis

ASJC Scopus subject areas

  • Biomedical Engineering
  • Orthopedics and Sports Medicine
  • Rehabilitation

Fingerprint Dive into the research topics of 'A training strategy for learning pattern recognition control for myoelectric prostheses'. Together they form a unique fingerprint.

  • Cite this