ブレイン・マシン／コンピュータ・インターフェイス BMI/BCI

P2-2-253 サルの皮質脳波から推定した筋電信号を用いた腕の角度の推定 Prediction of Joint angle from decoded Muscle Activities from Electrocorticograms in Primary Motor Cortex ○辛徳1, 中西康彦1, 陳超1, 神原裕行1, 吉村奈津江1, 渡辺秀典2, 南部篤3, 伊佐正2, 西村幸男2, 小池康晴1 ○Duk Shin1, Yasuhiko Nakanishi1, Chao Chen1, Hiroyuki Kambara1, Natsue Yoshimura1, Hidenori Watanabe2, Atsushi Nambu3, Tadashi Isa2, Yukio Nishimura2, Yasuharu Koike1 東京工業大学精密工学研究所1, 生理学研究所　発達生理学研究系2, 生理学研究所　統合生理研究系3 Precision and Intelligence Laboratory, Tokyo Institute of Technology, Yokohama1, Department of Developmental Physiology, National Institute for Physiological Sciences, Okazaki2, Department of System Integrative Physiology, National Institute for Physiological Sciences, Okazaki3 Electrocorticography (ECoG) is an alternative approach to less invasive brain-machine interfaces (BMI). Previous studies succeeded in classifying the movement direction and predicting hand trajectories from ECoGs. Despite such successful studies, there still remain considerable works for the purpose of realizing an ECoG-based BMI robot. We developed a method to decode multiple muscle activities from ECoG measurements in our previous study. We also verified that ECoG signals could be effective for predicting muscle activities in time varying series for preforming sequential movements. The aim of this study is to predict joint angle using decoded electromyogram (EMG) signals in order to control robot arm. Each ECoG signal was filtered by different bandpass filters for sensorimotor rhythms, normalized by the standard z-score, and smoothed by a Gaussian filter. We used sparse linear regression to find the best fit between frequency bands of ECoG and EMG. We also predicted angle of 4 DOF robot arm from the decoded EMG using 3-layer neural network. Consequently, this study shows that it could derive online prediction of angle of robot arm from ECoG signals.

P2-2-254 患者の大脳皮質におけるECoG信号を用いた腕の三次元軌跡の推定 Prediction of three-dimensional arm trajectory using ECoG signals in patient's cortical cortex ○中西康彦1, 柳澤琢史2,3,4, 平田雅之3, 陳超1, 辛徳1, 吉峰俊樹3, 小池康晴1 ○Yasuhiko Nakanishi1, Takufumi Yanagisawa2,3,4, Masayuki Hirata3, Chao Chen1, Duk Shin1, Toshiki Yoshimine3, Yasuharu Koike1 東工大・精工研1, 国際電気通信基礎技術研究所 脳情報研究所2, 大阪大学大学院医学系研究科脳神経外科3, 大阪大学大学院医学系研究科神経機能診断学4 Precision and Intelligence Lab., Tokyo Tech., Yokohama, Japan1, ATR Computational Neuroscience Laboratories, Kyoto2, Department of Neurosurgery, Osaka University3, Division of Functional Diagnostic Science, Osaka University Medical School, Osaka4 A number of studies have applied brain-machine interface to novel tools for paralyzed patients, such as neuro-prosthesis, neuro-rehabilitation and so forth. The invasive methods using intracortical microelectrodes brought about the good progress of this area. They have, however, limitation with serious burden of damage. Electrocorticogram (ECoG), by contrast, is less invasive and shows high spatial resolution in comparison with EEG. Recently, several ECoG based studies were applied to human beings for the prediction of one- or two-dimensional cursor controls, and finger flexion. However, the prediction study for three-dimensional trajectory could not demonstrate enough yet because most of paralyzed patients are not able to perform stably a long series of repeated trials. The aim of this study is to predict three-dimensional arm trajectory in time series from patient's ECoG data as a basis for a neuro-rehabilitation system. Patients executed tasks rotating their arm among 4 targets on a table. We employed the sparse linear regression to decide weight coefficients for the prediction of elbow angle and wrist trajectory. We also verified with the leave-one-out cross validation. As a result, the mean correlation coefficients between the predicted angles or trajectory and the actual measurements were approximately in the range of 0.4-0.7 in spite of the inadequate number of patient's trials for the precise prediction.

上部に戻る

前に戻る