AI-driven mobile recommendation systems for diabetes management

Diabetes can be effectively controlled by maintaining a healthy diet, well-managed blood glucose level and regular physical activity. Evidence suggests that improving dietary habits can play a crucial role in preventing the onset or progression of diabetes. A large number of mobile apps have been recently introduced to assist individuals with self-management of diabetes. However, these studies often provide dietary advice based on the average responses of groups to specific foods, rather than considering individual glycemic responses.

This project aims to develop an AI-driven, personalised diabetes management system using a mobile app to rate foods based on the glycaemic response of an individual. AI models will be trained on both the food intake and blood glucose data, and learn from an individual’s eating patterns and associated blood glucose responses over time. The AI model will be then used for providing personalised recommendations by rating food items as poor or healthy choices. Such regular and suggestive feedback using the strategy of ‘nudging’ over time could significantly modify behaviour and prevent the development of diabetes in patients with pre-diabetes and improve control inpatients with early diabetes on oral therapy. 

The second stage of the project aims to develop an online web-based (Model-to-App) platform enabling new users from to upload their glucose and food data, automatically generating a personalised prediction model and a downloadable app. This early-stage development will demonstrate the scalability and broader applicability over a longer period and across a wider population, supporting future translation into routine primary care and community-based preventive programs.