Objective:

To create a comprehensive, multimodal data set focused on type 2 diabetes mellitus (T2DM) to support artificial intelligence in ophthalmic research, enhancing understanding and treatment of the disease.

Key Findings:

• The data set includes 165,051 files, approximately 2 TB in size, covering over 15 modalities, providing a rich resource for AI model training.
• Participants include healthy individuals, those with prediabetes, and individuals managing diabetes with oral medications or insulin, ensuring diverse representation.
• Post-visit monitoring captures continuous glucose levels and environmental sensor readings, offering insights into real-world health dynamics.

Interpretation:

AI-READI provides a robust framework for understanding T2DM through a salutogenic lens, integrating various health determinants and facilitating advanced AI analyses that can lead to improved patient outcomes.

Limitations:

• The project is limited to participants from three specific sites, which may affect generalizability and the applicability of findings to broader populations.
• Data collection relies on self-reported surveys, which may introduce bias and affect the reliability of certain data points.

Conclusion:

AI-READI represents a significant advancement in diabetic eye research, offering a rich, standardized data set that can enhance AI applications in diabetes care, potentially transforming treatment strategies.