独立成分分析を用いた表面筋電信号から運動単位活動電位への分離: 模擬信号を用いた分離可能性の検討

摘要

A motor unit (MU) is a medical term of the basic element for muscle activity, and it is composed of a single motoneuron and several muscle fibers. Because the motoneuron governs all fibers in the same MU, the fibers are activated simultaneously by a motor command from the brain through a motoneuron. By the way, each muscle in our body consists of several MUs, whose properties are different in maximum force, recruitment, and so on, therefore a recording of muscle activity called a surface electromyogram (sEMG) represents a superimposed waveform of motor unit action potentials (MUAPs). From the viewpoint of motion analysis, decomposing the recorded sEMG signals into each MUAP is one of the attractive topics, and its results must give us various kinds of suggestions also applicable in the field of welfare engineering. Then, in order to carry out this challenging project, an independent component analysis (ICA) approach is adopted, and some computer simulations with biomimetic signals are tried under the simplified condition. As a result, it is found that the sEMG signals seem to be decomposed successfully according to the correlation coefficient, an evaluation index used in this study, but the actual observed waveforms have not been so sufficient for the final goal which we are aiming at.%筋は複数の筋線稚から成り，遠心性の運動神経を介して中枢神経からの運動指令を受け取って収縮する。通常，同一の神経支配を受ける筋線椎は複数あり，筋収縮時には，それらが同時にまとまって活動するため，このような神経と筋線椎をまとめて運動単位と呼んでいる）。また，その活動時に発せられる電気信号を運動単位活動電位という。