Premium
Automatic annotation of head movement among elderly people susceptible to Alzheimer's disease
Author(s) -
Helmi Garraoui
Publication year - 2020
Publication title -
alzheimer's and dementia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.713
H-Index - 118
eISSN - 1552-5279
pISSN - 1552-5260
DOI - 10.1002/alz.043862
Subject(s) - gesture , annotation , dialog box , computer science , nonverbal communication , recall , cognition , population , process (computing) , artificial intelligence , cognitive psychology , psychology , communication , medicine , world wide web , environmental health , neuroscience , operating system
Background Several researchers have revealed that sensory and motor changes can predate the cognitive symptoms of Alzheimer's disease (AD) by many years and may signify increased risk of developing AD. Therefore, studying the non‐verbal communication among the elderly susceptible to AD can contribute to a better understanding of their daily needs. Among these non‐verbal communications, we can mention head gestures, hand gestures, etc. This research discusses a new system model of nonverbal language annotation associated to hand gestures among elderly people susceptible to AD. Method The proposed approach aims at establishing a longitudinal study based on the use of recurrent neural network as a deep learning technique. We propose an interdisciplinary approach for automatic annotation of hand phases (rest position, preparation, hold, retraction), defined in many existing research works. To perform the classification, we rely on a ground truth, initiated by experts, known as CorpAGEst corpus. The latter is a collect of spontaneous conversations between interviewers and elderly people, which is used to train and test our model Result Experimentations show promising results, where the proposed approach succeeded in classifying the hand gesture phases with an 87.5% of precision and recall. Conclusion The proposed process focused on the automatic annotation of hand phases, reducing the cost of manual annotations and establishing a back and forth dialog between computer science communities and researchers in AD. The principal impact of this work is to contribute in establishing stronger gesture recognition techniques adapted to the aging population, giving the community of researchers tools to explore specificities in an automated fashion.