Retinal Blood Vessel Segmentation Algorithms: A Comparative Survey

Abstract

Automated segmentation and delineation of morphological properties of retinal vascular network had now become the most important research area in the treatment of ophthalmologic disorders. With the advancement of computational efficiency, image processing methodologies are widely used in ophthalmology. This paper gives the review of various segmentation techniques implemented by various authors in conjunction with performance metrics like sensitivity, specificity, accuracy and area under the curve. Results of various algorithms has been compared and analyzed. For the extraction of retinal vasculature, 2D retinal image from various databases has been considered.

A proposal to measure the similarity between retinal vessel segmentations images

“…These are the so-called true DRIVE and STARE which are currently the most common datasets for the evaluation of retinal vessel segmentation methods, can be found in Refs. (Fraz et al, 2012;Garg & Gupta, 2016;Vostatek et al, 2017).…”

“…A plethora of techniques, algorithms, and methodologies for the segmentation of retinal blood vessels can be found in the literature (Abràmoff et al, 2010;Fathi & Naghsh-Nilchi, 2013;Felkel et al, 2001;Garg & Gupta, 2016;Jelinek & Cree, 2009;Kolar et al, 2013;Kuri, 2015;Siddalingaswamy & Prabhu, 2010). A particularly useful reference for the present work is Ref.…”

“…A particularly useful reference for the present work is Ref. (Vostatek et al, 2017) (Garg & Gupta, 2016). Other more specific categorizations propose vessel tracking, multi-scale approaches, and vessel profiling as independent classifications from the more general morphology-based techniques (Fraz et al, 2012;Kirbas, C. Quek, 2004), and intensity-based segmentation algorithm (Gonzalez & Woods, 2001).…”

“…Next, we describe our method for the comparison of segmented retinal blood vessel images that is based on linking the standard performance metrics with the structural properties of the images through the FD. Our approach can be readily extended to a more complete assessment by including metrics of the connectivity, area, and length of the segmented vessels (Aquino, Gegúndez, Bravo, & Marín, 2010;Garg & Gupta, 2016;Kolar et al, 2013). For our numerical experiments, we made use of the retinal images from the DRIVE database (Staal et al, 2004).…”

Automatic detection and classification of diabetic retinopathy from retinal fundus images

“… The output image of the CGF can be used for more advanced image processing techniques including classification of image pixels into vessels or non-vessels; this can be helpful in filtering blood vessels [45].…”