Predicting Long-Term Type 2 Diabetes with Artificial Intelligence (AI): A Scoping Review

Area of Science:

• Medical Informatics
• Computational Biology
• Public Health

Background:

• Type 2 diabetes mellitus (T2DM) is a prevalent chronic metabolic disorder globally.
• Predicting T2DM risk is crucial for early intervention and management.
• Artificial intelligence (AI) offers advanced capabilities for risk prediction.

Purpose of the Study:

• To conduct a scoping review of AI techniques for long-term T2DM risk prediction.
• To evaluate the performance of different AI models in T2DM prediction.
• To identify key validation metrics for AI-driven T2DM risk assessment.

Main Methods:

• Scoping review methodology adhering to PRISMA-ScR guidelines.
• Systematic search and inclusion of 40 relevant studies on AI for T2DM prediction.
• Analysis of AI techniques, including Machine Learning (ML) and Deep Learning (DL), and their performance metrics.

Main Results:

• Machine Learning (ML) was the most common AI technique (23 studies).
• Deep Learning (DL) models were used in 4 studies, and combined ML/DL in 13.
• Ensemble learning models were prevalent (8 studies), with SVM and RF as top individual classifiers.
• Accuracy (31 studies) and recall (29 studies) were the primary validation metrics.

Conclusions:

• AI, particularly ML and DL, demonstrates significant potential for long-term T2DM risk prediction.
• High predictive accuracy and recall are critical for effective T2DM case detection.
• Further research into advanced AI models and robust validation is warranted for clinical application.