WL05: Leveraging Machine Learning for CGM Data Analysis in Diabetes Research

Music City Center 

Continuous glucose monitoring (CGM) provides real-time, high-resolution data that captures glucose fluctuations, offering valuable insights into diabetes management and research. However, the complexity, high dimensionality, and temporal dependencies of CGM data present significant analytical challenges that require advanced modeling techniques. This roundtable discussion will explore the application of machine learning (ML) methods-including logistic regression, LogitBoost, random forest, artificial neural networks, and partial least squares discrimination analysis (PLS-DA)-to enhance CGM data analysis. By leveraging these ML techniques, we aim to advance glucose pattern recognition, risk prediction, and individualized diabetes management. The discussion will address methodological challenges, model interpretability, and the integration of ML models into clinical decision-making. Participants will exchange ideas on optimizing ML approaches for CGM data and explore future directions in AI-driven diabetes research.

Machine Learning

Diabetes

Continuous Glucose Monitoring

Neural Network

LogitBoost

Partial Least Square Discrimination Analysis 

Biopharmaceutical Section