AI Model Enhances T2D Risk Prediction

Rachana Saha

An AI model incorporating polygenic risk scores and demographic variables that make accurate T2D risk prediction, as per a Taiwanese study.

Researchers used data from 68,911 participants, combining genetic data, medical imaging, and demographics for risk assessment.

The model achieved an initial AUC of 0.915, which increased to 0.949 with imaging data integration.

A simplified version using eight key variables showed a high AUC of 0.939 and was validated in a separate dataset.

Lead author Yi-Jia Huang emphasized the model's potential for early T2D detection but noted the need for external validation.