Detecting indicators of cognitive impairment via Graph Convolutional Networks
| Autor: | Hammadi Nait Charif, Abdelhamed Bouchachia, Damla Arifoglu |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2020 |
| Předmět: |
0209 industrial biotechnology
Computer science Daily life activities Health related 02 engineering and technology medicine.disease Activity recognition 020901 industrial engineering & automation Artificial Intelligence Control and Systems Engineering 0202 electrical engineering electronic engineering information engineering medicine Life expectancy Dementia Graph (abstract data type) 020201 artificial intelligence & image processing Electrical and Electronic Engineering Cognitive decline Cognitive impairment Brain syndromes Cognitive psychology |
| ISSN: | 0952-1976 |
| Popis: | While the life expectancy is on the rise all over the world, more people face health related problems such as cognitive decline. Dementia is a name used to describe progressive brain syndromes affecting memory, thinking, behaviour and emotion. People suffering from dementia may lose their abilities to perform daily life activities and they become on their caregivers. Hence, detecting the indicators of cognitive decline and warning the caregivers and medical doctors for further diagnosis would be helpful. In this study, we tackle the problem of activity recognition and abnormal behaviour detection in the context of dementia by observing daily life patterns of elderly people. Since there is no real-world data available, firstly a method is presented to simulate abnormal behaviour that can be observed in daily activity patterns of dementia sufferers. Secondly, Graph Convolutional Networks (GCNs) are exploited to recognise activities based on their granular-level sensor activations. Thirdly, abnormal behaviour related to dementia is detected using activity recognition confidence probabilities. Lastly, GCNs are compared against the state-of-the-art methods. The results obtained indicate that GCNs are able to recognise activities and flag abnormal behaviour related to dementia. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full Text from ScienceDirect