INDIVIDUAL MODELING FOR EMG-CT METHOD TO DETECT ACTIVITY OF EACH MUSCLE IN FOREARM
Naoki Toyota1, Satoshi Yamada1, Masahide Harada2, Harukazu Tohyama1,Norimasa Iwasaki1, Shigeru
1Hokkaido University, Japan;
2Harada Electronics Industry Limited, Japan;
3National Institute of Technology, Hakodate College, Japan
The authors have proposed the electromyography computed tomography (EMG-CT), which provides the activity of individual muscles within a cross section of the forearm by using surface EMG signals and an EMG conduction model. In the previous study, the cross section of the subject’s forearm was approximated by a circle in the EMG conduction model and the bone effects were not considered. Hence, the current study added the subject’s forearm shape and bone effects to the EMG conduction model for further improvement of the method and investigated the muscle activities during finger and thumb loadings. In the experiments, three subjects participated. The outer shape of subject’s forearm was obtained by a handy 3D scanner and bone regions in the cross-section were determined from an MRI image. Loadings were applied to the thumb and middle finger respectively, and the subjects were instructed to maintain the posture during the loadings. The surface EMG signals from the forearm were detected with 40 pairs of bipolar electrodes embedded in a customized EMG-CT band. In the EMG conduction model, it was assumed that EMG signals were not generated from the bone regions and could not pass through the bone. As a result, high muscle activities were detected in FDS and EDC regions during outer loadings to the middle finger at the distal phalanx. During inner loadings to the middle finger at the proximal phalanx, EDC and EDM regions were active. The strength of activity increased with the loadings. These results correspond to the expectation based on the anatomical knowledge. By this method, the distribution of the muscle activities became clearer than that from the previous method. Further, it was also confirmed that high signals were not generated from the bone regions even by the previous methods.