M-003
Automatic Evaluation on Trunk Coordination Ability Using Portable Devices: Feasibility and Preliminary Results

◎Zilu Liang（The Univ. of Tokyo）・Huizhi Elly Liang（Pierre et Marie Curie Univ.）・Takuichi Nishimura・Satoshi Nishimura・Yasuyuki Yoshida（AIST）

In this study we aim to utilize portable devices to evaluate trunk coordination ability. We investigated the feasibility of using accelerometer and Nintendo Wii balance board as alternatives to expensive and bulky motion capture systems for differentiating good and poor coordination ability. We extracted time and frequency domain features from data collected using the two devices, and investigated the differentiating power of these features. We found that the features of accelerometers were more powerful than those of Wii balance board. Further study is required to build a computational model for automatic classification of good and poor coordination ability.