Vascular changes precede tomographic changes in diabetic eyes without retinopathy and improve artificial intelligence...

Area of Science:

• Ophthalmology
• Medical Imaging
• Artificial Intelligence

Background:

• Diabetic retinopathy is a leading cause of vision loss.
• Early detection and intervention are crucial for managing diabetic eye disease.
• Traditional imaging methods may not capture subtle, early retinal changes.

Purpose of the Study:

• To evaluate early vascular and tomographic retinal changes in diabetic patients using artificial intelligence (AI).
• To compare the efficacy of optical coherence tomography (OCT) and OCT angiography (OCTA) in detecting early diabetic changes.
• To assess the diagnostic performance of AI models integrating OCT and OCTA data.

Main Methods:

• Inclusion of normal eyes, diabetic eyes without retinopathy (DWR), and mild non-proliferative diabetic retinopathy (NPDR) eyes.
• Acquisition of optical coherence tomography angiography (OCTA) B-scans for all participants.
• Derivation of tomographic features (thickness, volume) from OCTA scans for AI model input.

Main Results:

• Significant differences in OCT and OCTA features were observed between the study groups (P < .05).
• OCTA features demonstrated superior ability to detect early retinal changes in DWR eyes compared to OCT (P < .05).
• An AI model utilizing both OCT and OCTA features achieved an area under the curve of 0.91 ± 0.02 (P < .05).

Conclusions:

• The integration of AI with OCT and OCTA significantly enhances the early diagnosis of diabetic retinal changes.
• OCTA provides valuable insights into early vascular alterations in diabetic eyes.
• AI-driven analysis of multimodal imaging data holds promise for improved diabetic eye care.