Optimizing the trainable B-COSFIRE filter for retinal blood vessel segmentation

Abstract

Segmentation of the retinal blood vessels using filtering techniques is a widely used step in the development of an automated system for diagnostic retinal image analysis. This paper optimized the blood vessel segmentation, by extending the trainable B-COSFIRE filter via identification of more optimal parameters. The filter parameters are introduced using an optimization procedure to three public datasets (STARE, DRIVE, and CHASE-DB1). The suggested approach considers analyzing thresholding parameters selection followed by application of background artifacts removal techniques. The approach results are better than the other state of the art methods used for vessel segmentation. ANOVA analysis technique is also used to identify the most significant parameters that are impacting the performance results (p-value ¡ 0.05). The proposed enhancement has improved the vessel segmentation accuracy in DRIVE, STARE and CHASE-DB1 to 95.47, 95.30 and 95.30, respectively.

A Hybrid Unsupervised Approach for Retinal Vessel Segmentation

“…Similar methods were employed using MF [ 27 – 32 ], combined filters [ 33 ], COSFIRE filters [ 3 , 5 , 34 – 36 ], Gaussian filters [ 37 ], wavelet filters [ 38 ], and Frangi's filter [ 39 ]. The MM-based approaches are utilized for isolating retinal image segments such as optic disk, macula, fovea, and vasculature.…”

“…Critical diagnostic to eye diseases in human retinal images can be indicated by its shape analysis, its appearance, blood vessels, morphological features, and tortuosity [ 3 ]. Structure of RVS is also used for screening of brain and heart stock diseases [ 4 , 5 ].…”

Multiloss Function Based Deep Convolutional Neural Network for Segmentation of Retinal Vasculature into Arterioles and Venules

“…After completing the manual preparation of the labels in Figure 3, we have applied the following steps to enhance the prepared labels:Extracting the vessel tree of the original images in the black and white image [60]Combining the vessel tree results above with the manually prepared label to enhance the labels edgesMorphological processing to close the holes in the segmented vesselsNoise removal [61] to clean the backgroundPerforming the multiloss judgment for each generated label to enhance the thickness inconsistency of the pixels beyond the segmented edges as illustrated in and in Figures 4 and 5…”

D-GaussianNet: Adaptive Distorted Gaussian Matched Filter with Convolutional Neural Network for Retinal Vessel Segmentation

“…The retinal vascular tree is a major structure to be studied in the analysis of fundus images. Inspection of shape, width, tortuosity, and other blood vessel characteristics contributes to identify many retinal diseases [ 4 , 18 ]. Moreover, the detection and subtraction of the vascular tree facilitate the recognition of other lesions that appear in the retina, so a precise delineation of the vessels is required.…”