(英) |
Noninvasive approach for measurement of brain activity (e.g., magnetoencephalogram (MEG) and functional MRI)
is known as promising in the field of not only medical, but also engineering. MEG is especially appropriate
for application in engineering because it is superior in terms of time and space resolution.
In this study, we propose a new method for classifying emotional responses of examinee which is in being
provided visual stimulation (by presenting pictures of foods), based on the measured MEG signals.
Our method for analysis consists of noise cancellation based on Independent Component Analysis (ICA),
feature extraction by focusing frequency of MEG signals, dimension selection, and
classification with machine learning. |