An Intervention Framework for Cognitively Impaired Patients

People suffering from memory impairments experience decreased concentration, with difficulty in recalling and thinking. Inhabitants in a smart home are surrounded by multiple sensors and the sensor data collected provide information about the inhabitant’s interaction with the environment. Sensor data can be used to monitor the cognitive disabilities of the person by studying their behaviour in relation to carrying out Activity of Daily Living (ADLs). In this paper we propose an intervention framework for assistance that incorporates duration information along with the partially observed low-level sensor information and time. The framework learns probabilistically about the user and the ongoing activity and then provides decision support to monitor and assist in completing the ADLs. Our results verify that adding duration information consistently improves the accuracy in predictions from partial observation of sensor activations.