Optic Disc and Cup Segmentation in Retinal Images for Glaucoma Diagnosis by Locally Statistical Active Contour Model...

Area of Science:

• Ophthalmology
• Medical Image Analysis
• Computer Vision

Background:

• Accurate segmentation of the optic disc and optic cup is vital for glaucoma diagnosis.
• Existing methods struggle with intensity inhomogeneity and fail to utilize spatial prior knowledge of these structures.

Purpose of the Study:

• To develop a novel approach for robust and accurate joint segmentation of the optic disc and optic cup.
• To overcome limitations of existing methods, particularly intensity variations and lack of structural prior integration.

Main Methods:

• A novel locally statistical active contour model with structure prior (LSACM-SP) was developed.
• Preprocessing techniques were used for initial contour extraction.
• The locally statistical active contour model (LSACM) was applied to handle intensity inhomogeneity.
• A structure prior was incorporated to leverage the anatomical relationship between the optic disc and optic cup.

Main Results:

• The LSACM-SP approach demonstrated robust performance in segmenting optic disc and optic cup structures.
• The method effectively handled intensity inhomogeneity in fundus images.
• Experimental results on DRISHTI-GS and RIM-ONE r2 databases showed superiority over existing algorithms.

Conclusions:

• The proposed LSACM-SP method offers an effective solution for optic disc and optic cup segmentation.
• This approach enhances diagnostic capabilities for glaucoma by providing more accurate structural segmentation.
• The integration of local statistics and structure priors is key to the model's success.