Estimation of mild cognitive impairment and frailty using foot pressure, acceleration, and angular velocity information during walking

Abstract

The number of elderly people with dementia is increasing every year and becoming a social problem. As symptoms of dementia progress, physical function declines and physical activity decreases, creating a vicious cycle. However, in the case of mild cognitive impairment (MCI), half of all dementia patients can recover to a normal state through treatment. Early detection of dementia is therefore essential. Therefore, we focused on gait disturbance, one of the symptoms of dementia. As an experimental method, we conducted an outdoor walking experiment using an insole with a pressure sensor placed and a lumbar acceleration/angular velocity sensor. Ten cognitive function estimation parameters were extracted from the obtained foot pressure data and lumbar acceleration/angular velocity data. The Mini-Mental State Examination(MMSE) was also used to measure the subjects' cognitive function. Multiple regression analysis and binomial logistic regression analysis were performed to examine the relationship between cognitive function and the estimated parameters of cognitive function. As a result, highly accurate cognitive function estimation equations were obtained for the MCI group. In addition, discriminant equations were obtained with high accuracy for healthy subjects and MCI, and for dementia and healthy subjects with MCI. This study is expected to confirm the decline of cognitive function from walking behavior in daily activities and to be useful for early detection and treatment of dementia.