Human Digital Twin for Personalized Elderly Type 2 Diabetes Management

Area of Science:

• Biomedical Engineering
• Artificial Intelligence in Medicine
• Geriatric Medicine

Background:

• Managing type 2 diabetes in elderly patients presents unique challenges due to comorbidities and polypharmacy.
• Personalized medicine is crucial for effective diabetes management in this demographic.
• Existing treatment strategies often lack the adaptability required for complex geriatric conditions.

Purpose of the Study:

• To introduce a human digital twin (HDT) framework for personalized management of type 2 diabetes in the elderly (E-T2D).
• To develop and validate predictive and management models within the HDT framework to optimize E-T2D treatment.
• To enhance precision medicine approaches for E-T2D by integrating patient-specific data and advanced control algorithms.

Main Methods:

• Development of a human digital twin (HDT) framework integrating mathematical and deep-learning models.
• Utilization of patient-specific data for creating virtual patient models within the HDT.
• Implementation of an adaptive patient model and a learning-based model predictive control (LB-MPC) algorithm for personalized insulin infusion.
• Incorporation of geriatric conditions as parameters and constraints within the LB-MPC for tailored management.

Main Results:

• The HDT framework demonstrated significant improvements in glycemic control, increasing time-in-range from 75% to 86-97%.
• The system achieved a reduction in insulin infusion requirements by 14-29%.
• The framework was successfully deployed and validated using clinical trial data and simulations from 15 elderly patients.

Conclusions:

• The human digital twin (HDT) framework offers a promising approach for personalized diabetes management in elderly patients.
• HDT-based precision medicine can effectively address the complexities of geriatric conditions in type 2 diabetes.
• This innovative framework enhances treatment efficacy and patient outcomes in E-T2D management.