Personalized Blood Glucose Prediction Using Multimodal Real-Life Data and Artificial Intelligence

Diabetes, especially type 1, requires continuous monitoring of blood glucose levels to prevent complications like hyperglycemia or hypoglycemia. While new technologies like Continuous Glucose Monitoring (CGM) and insulin pumps offer better management, predicting future glucose levels remains difficult due to the complexity of individual physiology and lifestyle factors. This project aims to create a smart system using artificial intelligence (AI) to accurately predict glucose levels. By combining CGM data with real-world lifestyle information—such as diet, exercise, sleep, stress, and emotions—collected via wearable sensors and apps, the system will learn to personalize predictions for each patient. The goal is to improve daily diabetes management and quality of life by helping patients and doctors take proactive steps to prevent dangerous glucose spikes or drops.