Analysis of Cerebral Blood Flow in Imagination of Moving Object

We proposed operating HMD by BMI to speed up operation and to enhance the convenience in wearable device equipping HMD. But it is difficult to reflect user's thought direct on operation at this stage. In order to realize it, we need to identify a kind of human's brain signal when human thinks about moving object. So our study verifies whether it is possible to identify by imaging direction in data that we investigate cerebral blood flow changes in using NIRS when subject thinks about moving object. Identification was using neural network. The average of identification ratio of 4 direction in Total-Hb, the average of identification ratio of 4 direction in Oxy-Hb, the average of identification ratio of 4 direction in Deoxy-Hb are 32.25%, 36.0% and 37.0%, respectively. The average of identification ratio of 2 direction, up and down, in Total-Hb, the average of identification ratio of 2 direction, up and down, in Oxy-Hb, the average of identification ratio of 2 direction, up and down, in Deoxy-Hb are 63.5%, 62.0% and 62.5%, respectively. The average of identification ratio of 2 direction, left and right, in Total-Hb, the average of identification ratio of 2 direction, left and right, in Oxy-Hb, the average of identification ratio of 2 direction, left and right, in Deoxy-Hb are 58.0%, 55.5% and 69.0%, respectively. From the results, we show the possibility of identification because identification ratio are higher than chance level