Machine Learning-Based Factor Analysis of Carbohydrate Metabolism Compensation for TDM2 Patients.

Area of Science:

• Endocrinology
• Medical Informatics
• Computational Biology

Background:

• Type 2 diabetes (T2DM) is a prevalent global chronic disease with escalating patient numbers.
• Poor carbohydrate metabolism control in T2DM significantly increases cardiovascular risks and leads to early disability.
• Effective therapy selection is crucial for managing T2DM and mitigating associated health complications.

Purpose of the Study:

• To develop and validate an automated method for selecting optimal, personalized therapies for Type 2 diabetes patients.
• To predict the likelihood of compensating carbohydrate metabolism within a six-month timeframe for individual patients.
• To enhance clinical decision-making for healthcare providers managing T2DM.

Main Methods:

• Development of predictive models utilizing historical medical data and current patient indicators.
• Creation of personalized optimal therapy recommendations with a high probability of success.
• Integration of the method into Support and Decision-Making Systems for T2DM management.

Main Results:

• The proposed method demonstrates high-quality predictions for therapy outcomes.
• Personalized therapies show the highest probability of compensating carbohydrate metabolism.
• The system provides reliable medical recommendations for T2DM treatment.

Conclusions:

• The automated therapy selection method offers a valuable tool for optimizing T2DM management.
• Personalized treatment strategies can significantly improve patient outcomes and reduce cardiovascular risk.
• This approach supports healthcare professionals in making informed decisions for Type 2 diabetes care.