Personal digital twin: A virtual human twin for therapy adjustment in people with diabetes (2024-2026). Reference: 2024 PROD 00051. Funded by AGAUR – 145.500€. P.I. Omer Mujahid.
SUMMARY:
The Personal Digital Twin (PDT) project aims to develop an AI-driven virtual human twin for optimizing insulin therapy in diabetes patients. Managing insulin therapy is a continuous challenge, requiring precise adjustments to maintain glycemic control. This project leverages conditional deep generative models to create personalized digital twins of diabetes patients that simulate blood glucose responses based on real-time patient data, including insulin dosages, and carbohydrate intake.
The PDT integrates with continuous glucose monitoring (CGM) sensors, smart insulin pens, and mobile applications, providing an interface where clinicians can analyze glycemic trends and explore different therapy adjustments. Through real-time simulations, the PDT helps predict glycemic trends, enabling clinicians to make informed decisions tailored to each patient’s specific needs.
The project aims to develop a practical, data-driven tool to support diabetes management by leveraging personalized simulations. Through the use of deep generative models, the PDT seeks to contribute to a more individualized approach to insulin therapy, potentially improving treatment precision and patient outcomes.
TEAM:
MICELAB:
Research leads:
- Josep Vehí Casellas
Research Team:
- Omer Mujahid
- Aleix Beneyto Tantiña
- Iván Contreras Fernández-Dávila
- Oriol Bustos Martínez
- New member
RESEARCH OUTCOMES:
- Therapy Adjustment Tool for Healthcare Professionals: Streamlined adjustments to insulin therapies, enabling more productive and effective care.
- Optimized Glycemic Control: Improved insulin regimens result in better glycemic profiles and reduced risks of diabetes-related complications.
- Reduced Hypoglycemia Risk: Fewer insulin-induced hypoglycemic events alleviate patients’ fear of hypoglycemia and reduce diabetes distress.