作者: R. Vinjamuri , D.J. Weber , A.D. Degenhart , J.L. Collinger , G.P. Sudre
DOI: 10.1109/IEMBS.2009.5332746
关键词: Wired glove 、 Computer vision 、 Power (physics) 、 Cerebral cortex 、 Artificial intelligence 、 Computer science 、 Motor cortex 、 Acceleration 、 Electroencephalography 、 Electrode array 、 Fuzzy logic 、 Simulation 、 Neurophysiology
摘要: This paper presents a fuzzy logic model to decode the hand posture from electro-cortico graphic (ECoG) activity of motor cortical areas. One subject was implanted with micro-ECoG electrode array on surface cortex. Neural signals were recorded 14 electrodes this while Subject participated in three reach and grasp sessions. In each session, reached grasped wooden toy hammer for five times. Optimal channels/electrodes which active during task selected. Power spectral densities optimal channels averaged over time period 1/2 second before onset movement 1 after fed into model. decoded whether is open or closed 80% accuracy. Hand postures along by using output two methods (i) velocity based decoding (ii) acceleration decoding. The latter performed better when predicted compared data glove experiment. imported MATLAB®SIMULINK control virtual hand.