Joint optic disc and cup segmentation based on elliptical-like morphological feature and spatial geometry constraint.

Area of Science:

• Ophthalmology
• Medical Imaging
• Computer Vision

Background:

• Glaucoma is a leading cause of irreversible blindness globally.
• Accurate segmentation of the optic disc and cup is crucial for glaucoma screening.
• Current methods may lack precision in detecting these key ocular structures.

Purpose of the Study:

• To develop a novel, precise, and automatic method for optic disc and cup segmentation.
• To improve glaucoma screening by directly obtaining a screening indicator (vCDR).
• To leverage elliptical features and spatial relationships for enhanced segmentation accuracy.

Main Methods:

• Reformulating segmentation as an ellipse detection task, utilizing elliptical-like morphological features.
• Detecting minimum bounding boxes of ellipses and learning ellipse parameters for segmentation.
• Introducing Paired-Box RPN for simultaneous, coupled detection of optic disc and cup.
• Incorporating a boundary attention module to guide context aggregation using disc and cup edges.

Main Results:

• The proposed method achieves superior performance in optic disc and cup segmentation compared to state-of-the-art techniques.
• The approach demonstrates effective glaucoma screening performance by calculating the vertical cup-to-disc ratio (vCDR).
• Joint segmentation utilizing elliptical features and spatial constraints enhances segmentation accuracy.

Conclusions:

• The novel ellipse detection-based approach offers a significant advancement in optic disc and cup segmentation.
• This method provides a more accurate and automated tool for glaucoma screening.
• Combining morphological features with spatial priors improves the reliability of automated ocular structure segmentation for disease detection.