Adaptive myoelectric control applied to video game. (April 2015)

• Record Type: Journal Article
• Authors: Oskoei, Mohammadreza Asghari; Hu, Huosheng
• Is Part Of: Biomedical signal processing and control. Volume 18(2015)
• Page Start: 153
• Page End: 160
• Publication Date: 2015-04
• Subjects: Myoelectric control -- Adaptive schemes -- Video game -- Rehabilitation -- Support vector machine; 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.2014.11.002
• Languages: English
• ISSNs: 1746-8094

Abstract

Highlights: We proposed adaptive schemes for myoelectric control applied to a video game. Both unsupervised and supervised adaptive schemes improve the performance. Both cope with time-related steady state deviations in myoelectric signal patterns. They keep the performance stable in long-term applications. Adaptive schemes improve game scores by 10% and reduce the high-risk errors by 40%. Abstract: This paper proposes adaptive schemes to cope with time-related changes in muscle activities during playing video game. A myoelectric control with the core of support vector machine is applied to manipulate a car in a computer-based video game. The proposed adaptive schemes model fatigue-based changes in myoelectric signals and modify the classification criteria to keep stable performance in long-term operations. Both unsupervised and supervised methods were applied to detect time-related steady state deviations in myoelectric signal patterns. Both methods improve the performance of myoelectric control and keep it stable in long-term applications.