Multiple Preprocessing Hybrid Level Set Model for Optic Disc Segmentation in Fundus Images

Abstract

The accurate segmentation of the optic disc (OD) in fundus images is a crucial step for the analysis of many retinal diseases. However, because of problems such as vascular occlusion, parapapillary atrophy (PPA), and low contrast, accurate OD segmentation is still a challenging task. Therefore, this paper proposes a multiple preprocessing hybrid level set model (HLSM) based on area and shape for OD segmentation. The area-based term represents the difference of average pixel values between the inside and outside of a contour, while the shape-based term measures the distance between a prior shape model and the contour. The average intersection over union (IoU) of the proposed method was 0.9275, and the average four-side evaluation (FSE) was 4.6426 on a public dataset with narrow-angle fundus images. The IoU was 0.8179 and the average FSE was 3.5946 on a wide-angle fundus image dataset compiled from a hospital. The results indicate that the proposed multiple preprocessing HLSM is effective in OD segmentation.

Identifying the Edges of the Optic Cup and the Optic Disc in Glaucoma Patients by Segmentation

“…Sun et al used ResFPN-Net to learn the boundary features and the inner relation between OD and OC for automatic segmentation [ 29 ]. Xue et al used hybrid level set modeling for disc segmentation [ 30 ]. Zaaboub et al proposed a two-stage (OD localization and segmentation) approach to detect the contour of the OD [ 31 ].…”

Automated optic disk segmentation for optic disk edema classification using factorized gradient vector flow

“…Eccentricity and size are used for the final OD selection 2054 images from DRIONS, MESSIDOR, ONHSD, DIARETDB1, DRISHTI, RIM-ONE Sensitivity 96.49, Specificity 99.75, Accuracy 99.60, Wilson and Mahesh 30 Using superpixels with the k-mean algorithm 1310 images from DRIONS, MESSIDOR Jaccard 84.23, Dice 90.84, Accuracy 99.34 Rehman et al 31 Using a simple linear iterative clustering algorithm technique combined with the features-based classification. The features from the segmented superpixel clusters are Intensity-based statistical features, texton-map histogram features, and fractal features, are extracted 1409 images from DRIONS, MESSIDOR, ONHSD Sensitivity 96.9, Specificity 99.5, Dice 89.9, Accuracy 99.3 Dai et al 32 Using a combination of the three energies: phase-based boundary, PCA-based shape, and region energies 1409 images from MESSIDOR, ONHSD, DRIONS Overlap 90.54 Xue et al 33 The hybrid level set model (HLSM) included distance-regularized, line integral and area integral, area-based, and shape-based models 138 images from DRSHTI-GS, TMUEH Intersection over union 92.75, Four-side evaluation 464.36 Gao et al 34 Using saliency detection and thresholding techniques to get a rough OD boundary. The oval fitting model is used to segment higher accurate boundary 229 images from DIARETDB0, DRSHTI-GS Overlap 66.59, Accuracy 96.30, F1-score 95.1 Abdullah et al 35 Using the fuzzy clustering mean method to localize the location.…”