Open AccessEstimation of posture and prediction of the elderly getting out of bed using a body pressure sensor
Author(s) -
Atsushi Hagihara,
Fumiko Harada,
Hiromitsu Shimakawa
Publication year - 2021
Publication title -
international journal of power electronics and drive systems/international journal of electrical and computer engineering
Language(s) - English
Resource type - Journals
eISSN - 2722-2578
pISSN - 2722-256X
DOI - 10.11591/ijece.v11i2.pp1208-1222
Subject(s) - hidden markov model , computer science , intervention (counseling) , state (computer science) , artificial intelligence , simulation , medicine , algorithm , nursing
We propose an IoT support system for estimating the posture of the care recipient on the bed from the body pressure of the care recipient measured by a sheet-type body pressure sensor, and detecting the posture related to leaving the bed in real time. In addition, we propose a method that predicts getting out of the bed before the care recipient takes a posture related to getting out of the bed by considering the state transition. Intervention experiment showed that using body pressure features as an explanatory variable and applying machine learning, 16 types of postures on the bed of care recipients with an F value of 0.7 or more could be identified. From the experiment without intervention, by applying the hidden Markov model, we calculated the transition probability to each hidden state when the care recipient getting out of the bed and the transition probability to each hidden state when the care recipient not getting out of the bed. As a result, there was a difference of about 0.1 in the transition probability of the state related to raising upper body.