生活活動度に基づく日常活動状態の判別法

概要

論文の詳細を見る
In the field of rehabilitation or care, it is extremely important that medical staff clearly understand the appearance of daily activities of the subjects to make and improve rehabilitation programs or care plans. Information on daily activities of the subjects is usually obtained in each facility by questionnaires answered by subjects or their families, and through observation by medical staff. However these methods have some difficulties; questionnaires might include unreliable information; observations differ due to each observers subjectivity; observing is stressful for medical staff and observations invade subject's privacy. With these points in mind, several works on some sensors attached to the subjects and the sensing data analyzed in order to understand daily activities of the subjects have been reported. In this paper, a novel method is proposed that can judge daily activities, such as lying, sitting, standing, walking and wheelchair-driving, by using neural networks from the sensing data of the accelerometer attached to the subject. The network concerning posture of the subject when lying, sitting and standing and the network also concerning actions when walking and wheelchair-driving are constructed in a cascade. From the experimental results for 7 subjects in a facility, their matching ratios of evaluated the Level of Daily Activities between a use of video and the proposed method are 86.0% to 95.7%, which shows the effectiveness of the proposed method.
社団法人日本生体医工学会の論文
2008-12-10

生活活動度に基づく日常活動状態の判別法

スポンサーリンク

概要

著者

関連論文

スポンサーリンク