Development of a Model for Estimating Cognitive Function in the Elderly Using EEG and Heart Rate Variability

In recent years, the number of people with dementia has been increasing. Treatment of dementia is important. Dementia is a progressive disease that requires treatment and care according to the progress of the disease. As of now, the progress of the disease, such as cognitive decline, is generally diagnosed through various examinations. These methods are time-consuming and are not suitable for simple diagnosis in outpatient settings or welfare facilities. The purpose of this study is to estimate cognitive function using a simpler method. As a proposal for this purpose, a model was constructed using support vector machines (SVM) and logistic regression with data measured by a simple electroencephalograph (EEG) and a heart rate monitor as explanatory variables and neuropsychological examination, which are often used in examination for dementia, as objective variables. The results showed that SVM could estimate cognitive function with an accuracy of up to 71.6%.