Towards ambient intelligent care and assistance systems for patients with dementia

Abstract

This paper presents a monitor assistant system for detection of dangerous events and health diagnostics for patients with dementia. The proposed framework brings together three different types of ambient intelligence: (i) sensors for detecting life-threatening events that can occur at home, like falls and anomalies in vital signs predicting risk of myocardial infarction; (ii) sensors placed under bed-posts for detecting unhealthy poor movement while patients are lying on the bed which can cause bedsores (decubitus); and (iii) analysis of the current cognitive state of patients with dementia by means of serious computer games. This paper presents the framework, methods and sensors used by our system to assist patients with dementia by triggering different type of alarm signals informing which patient needs what and how some help could be provided.