The National Institute of Technology (NIT) Rourkela has announced the development of a cutting-edge AI-driven model aimed at enhancing blood sugar predictions for individuals with diabetes.
This innovative tool seeks to provide accurate predictions, enabling timely and personalized adjustments in insulin doses, diet, and physical activity.
Designed to outperform traditional methods, the AI model utilizes multi-head attention layers within a neural basis expansion network to focus on the most relevant data points, effectively filtering out noise. Optimized for use on smartphones and insulin pumps, it ensures easy accessibility for daily diabetes management.
Lead researcher, Professor Mirza Khalid Baig, emphasized the model’s potential to significantly improve diabetes care by offering reliable predictions tailored to individual glucose patterns without requiring extensive training data or computing power.
The model is set to undergo extensive clinical trials in collaboration with senior diabetologists in Odisha, supported by the Department of Science and Technology, the Department of Biotechnology, and NIT Rourkela. This AI-driven approach promises to be a game-changer in the field of diabetes management, integrating seamlessly into digital health solutions.
