Thinking state monitoring using plural physiological signals

In this paper, we present a method for monitoring the human thinking state, which uses plural physiological signals. First, an experiment was performed over three days to obtain ECG (electrocardiogram), EOG (electrooculogram), and RSP (respiratory activity) data when a subject was solving a mathematics problem. Second, it was confirmed that the three physiological signals had characteristics corresponding to the basic modes of the thinking state. Third, various types of thinking state estimation models were built using one or three physiological signals, and the models were evaluated. As a result, it was found that the model which used ECG, EOG, and RSP together to monitor the human thinking state was the most reliable.