Year: 2008
Type: article
Abstract: The promise of pattern recognition for improved control of upper-extremity powered prostheses has existed for a long time. During the years of offline research and algorithm development, very little experience has been gained with real-time use in clinical and chronic settings. Our group, having the benefit of working with subjects who have undergone targeted muscle reinnervation (TMR) surgery, is at the forefront of real-world application of pattern recognition for upper extremity amputees. Based on our experiences, we highlight a progression of myoelectric control schemes from conventional control to enhanced pattern recognition control, stressing the application of simple pattern recognition schemes to replace more conventional control. These clinically practical pattern recognition systems incorporate a realistic number of electrodes and the ability to control available prosthetic components. Our experience suggests how the impending, and initial deployment of pattern recognition-controlled prostheses for daily use can be more approachable than what is depicted in high-dimension studies common in the literature today. BACKGROUND The use of pattern recognition for robust myoelectric control, and its promise for improved prosthesis function, has been steadily investigated since the 1960s [1-20]. Investigators around the world have developed techniques and demonstrated efficacy for systems involving anywhere
Source: DukeSpace (Duke University)
Authors: Blair A. Lock, Aimee E. Schultz, Todd Kuiken
Language: English
Subfield: Biomedical Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W2097657279