A Framework for Developing Smart Mobile Healthcare Companions for Chronic Diseases Using a Chatbot and Bayesian Reputation Model
摘要
Type 2 Diabetes Mellitus is a chronic noncommunicable disease that can develop from risk factors such as obesity, eating habits, and sedentary lifestyle, among others. In addition, metabolic syndrome has risk factors similar to diabetes. Reducing or controlling these factors can reduce the probability of the occurrence of these diseases. However, patients must adhere easily to existing mobile health (m-health) programs that implement remote monitoring and self-care healthcare programs. We propose a new framework for developing mHealth tools based on components that researchers can replace depending on the health problem being monitored. The components are as follows: an online repository for patient information and follow-up, a mobile tool for user follow-up and feedback, and a tool that personalizes remote patient follow-up between medical visits. We use the framework to build mHealth to help patients achieve the medical and lifestyle goals set by their healthcare professionals. The central part is a Chatbot that generates messages using a Bayesian reputation model to adjust the conversation based on patient achievement. A website was created that doctors can use to monitor patient achievement and adjust goals as needed. A usability test showed promising results on using m-health to ensure patients met and completed set goals. The proposed framework shows the potential to develop similar tools for other chronic diseases that would benefit from remote monitoring through personalized m-health.