2A1-A09 表面筋電位を利用したリアルタイム舌運動推定法(感覚・運動・計測(1))

DOI Web Site

書誌事項

タイトル別名
  • 2A1-A09 Real-Time Tongue Movement Classification by Using Surface Electromyogram Signals(Sense, Motion and Measurement(1))

抄録

Present study explores an approach for tongue movement recognition by real time analysis of the surface electromyogram (EMG) signals of the suprahyoid muscles. The EMG signals from the suprahyoid muscles were captured by an electrode array of nine surface electrodes mounted on an elastic base that was affixed to the underside of the jaw. The signals from each electrode were processed by using monopolar leads. In order to utilize the EMG signals from neighboring muscles (crosstalk) positively, we calculated the EMG signals between any two electrode channels and used the resulting 36 (=_9C_2) signals as inputs to a neural network that estimated the tongue movements from these signals. This approach allowed us to design a highly robust real-time algorithm for detection of intentional tongue movements. The optimal parameters of the algorithm for real-time estimation were clarified by analysis of the discrimination rate and the time-delay of the electrode signals.

収録刊行物

キーワード

詳細情報 詳細情報について

問題の指摘

ページトップへ