2021/10/19 by Francesca Palermo, Palermo, Francesca, Honglin Li +15
Computer Science · Medicine · #Artificial Intelligence (cs.AI) #Context-Aware Activity Recognition Systems #Dementia and Cognitive Impairment Research #FOS: Computer and information sciences #Machine Learning (cs.LG) #Machine Learning in Healthcare
paper · pdf · doi:10.48550/arxiv.2110.09868
openalex publication_date 2021/10/19 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Agitation is one of the neuropsychiatric symptoms with high prevalence in\ndementia which can negatively impact the Activities of Daily Living (ADL) and\nthe independence of individuals. Detecting agitation episodes can assist in\nproviding People Living with Dementia (PLWD) with early and timely\ninterventions. Analysing agitation episodes will also help identify modifiable\nfactors such as ambient temperature and sleep as possible components causing\nagitation in an individual. This preliminary study presents a supervised\nlearning model to analyse the risk of agitation in PLWD using in-home\nmonitoring data. The in-home monitoring data includes motion sensors,\nphysiological measurements, and the use of kitchen appliances from 46 homes of\nPLWD between April 2019-June 2021. We apply a recurrent deep learning model to\nidentify agitation episodes validated and recorded by a clinical monitoring\nteam. We present the experiments to assess the efficacy of the proposed model.\nThe proposed model achieves an average of 79.78% recall, 27.66% precision and\n37.64% F1 scores when employing the optimal parameters, suggesting a good\nability to recognise agitation events. We also discuss using machine learning\nmodels for analysing the behavioural patterns using continuous monitoring data\nand explore clinical applicability and the choices between sensitivity and\nspecificity in-home monitoring applications.\n