TY - JOUR
T1 - Activity tracking and monitoring of patients with alzheimer’s disease
AU - Lam, Kam Yiu
AU - Tsang, Nelson Wai Hung
AU - Han, Song
AU - Zhang, Wenlong
AU - NG, Joseph K Y
AU - Nath, Ajit
N1 - Publisher Copyright:
© 2015, Springer Science+Business Media New York.
PY - 2017/1/1
Y1 - 2017/1/1
N2 - In this paper, by applying motion detection and machine learning technologies, we have designed and developed an activity tracking and monitoring system, called SmartMind, to help Alzheimer’s Disease (AD) patients to live independently within their living rooms while providing emergency assistances and supports when necessary. Allowing AD patients to handle their daily activities not only can release the burdens on their families and caregivers, it is also highly important to help them regain confidence towards a healthy life. The daily activities of a patient captured from SmartMind can also serve as an important indicator to describe his/her normal living habit (NLH). By checking NLH, the patient’s current health status can be estimated on a daily basis. In the testing experiments of SmartMind, we have demonstrated the accuracy of SmartMind in activity detection and investigated its performance when different machine learning algorithms were adopted for posture detection. The performance results indicate that both support vector machine (SVM) and naive bayes (NB) can achieve an accuracy of higher than 97 % while the random forrests (RF) only gives an accuracy of around 73 %.
AB - In this paper, by applying motion detection and machine learning technologies, we have designed and developed an activity tracking and monitoring system, called SmartMind, to help Alzheimer’s Disease (AD) patients to live independently within their living rooms while providing emergency assistances and supports when necessary. Allowing AD patients to handle their daily activities not only can release the burdens on their families and caregivers, it is also highly important to help them regain confidence towards a healthy life. The daily activities of a patient captured from SmartMind can also serve as an important indicator to describe his/her normal living habit (NLH). By checking NLH, the patient’s current health status can be estimated on a daily basis. In the testing experiments of SmartMind, we have demonstrated the accuracy of SmartMind in activity detection and investigated its performance when different machine learning algorithms were adopted for posture detection. The performance results indicate that both support vector machine (SVM) and naive bayes (NB) can achieve an accuracy of higher than 97 % while the random forrests (RF) only gives an accuracy of around 73 %.
KW - Dementia
KW - Health informatics Context-aware computing
KW - Motion detection
KW - Pervasive computing
UR - http://www.scopus.com/inward/record.url?scp=84946926007&partnerID=8YFLogxK
U2 - 10.1007/s11042-015-3047-x
DO - 10.1007/s11042-015-3047-x
M3 - Journal article
AN - SCOPUS:84946926007
SN - 1380-7501
VL - 76
SP - 489
EP - 521
JO - Multimedia Tools and Applications
JF - Multimedia Tools and Applications
IS - 1
ER -