Estimating Cognitive Decline using P300-based Spelling-Brain-Computer Interface

抄録

Early diagnosis is important in the treatment of dementia; however, many dementia patients are resist seeking medical attention. In our laboratory, we are developing a dementia screening tool using the P300-based Spelling-Brain-Computer Interface (Spelling-BCI) to facilitate early dementia diagnosis. By estimating the results of neuropsychological examinations that must be performed by a specialist with BCI, we consider that an easy cognitive function test with Spelling-BCI can be realized. Multiple regression analysis was performed using the features obtained from the Spelling-BCI and the age of the subjects, and Mini-Mental State Examination (MMSE), Japanese Version of Montreal Cognitive Assessment (MoCA-J), and Frontal Assessment Battery (FAB) scores were estimated. In the multiple regression analysis, variable selection was performed using the forward-backward stepwise selection method, and data exceeding the 95% confidence interval of the estimation error were excluded. As a result, the adjusted R-squared exceeded 0.95 in the estimation model of each neuropsychological examination. Therefore, the experimental results suggest that neuropsychological examinations can be estimated using the Spelling-BCI.