To develop a computational framework for 3D segmentation and quantification of retinal microvasculature without manual annotation, enhancing efficiency and accuracy in clinical settings.
Key Findings:
RADAR outperformed standard segmentation tools, indicating its superior capability.
It resolved layer-specific morphological alterations not visible in 2D projections, which is critical for accurate diagnosis.
Quantitative analysis showed distinct patterns of compensatory remodeling and increased tortuosity in diabetic eyes, suggesting potential biomarkers for disease progression.
RADAR enables precise extraction of volumetric biomarkers such as vessel length and branching complexity, which can inform treatment strategies.
Interpretation:
RADAR provides a scalable and efficient tool for early detection and longitudinal assessment of ocular and systemic vascular diseases, potentially transforming monitoring practices.
Limitations:
Datasets generated are not publicly available due to patient privacy restrictions, limiting broader validation.
Custom code for RADAR is not publicly available due to intellectual property restrictions, which may hinder collaborative improvements.
Conclusion:
RADAR represents a significant advancement in the analysis of retinal microvasculature, facilitating better monitoring of vascular health and potentially improving patient outcomes.
A retrospective database study found a low absolute incidence but higher relative hazard of ischemic optic neuropathy following semaglutide initiation.
