A reinforcement learning and deep learning based intelligent system for the support of impaired patients in home treatment

Abstract A clinical treatment process typically carries out in two stages; i.e., hospital stay and treatment at home after hospitalization. The correct completion of the treatment process is essential, but it becomes challenging for elders and patients with any physical or cognitive disability since they need assistance in the execution of the treatment itself. This work presents an intelligent system able to provide automatic assistance to those patients that have to follow a planned treatment at home. The system can support the patient with both customized reminders whenever it is the time to take medication and alerts to avoid possible medication errors when the patient is going to assume an incorrect drug by mistake. The core of the proposed solution consists of a multi-agent system that relies on algorithms of both Reinforcement Learning and Deep Learning. Experimental results show that the system improves the quality of home assistance services reducing medication errors.