An Active Contour Model for Segmenting and Measuring Retinal Vessels

By Bashir Al-Diri, Andrew Hunter, David H. Steel. IEEE Transactions on Medical Imaging, 2009

472 citations recorded by the source corpus.

This paper presents an algorithm for segmenting and measuring retinal vessels, by growing a ldquoRibbon of Twinsrdquo active contour model, which uses two pairs of contours to capture each vessel edge, while maintaining width consistency. The algorithm is initialized using a generalized morphological order filter to identify approximate vessels centerlines. Once the vessel segments are identified the network topology is determined using an implicit neural cost function to resolve junction configurations. The algorithm is robust, and can accurately locate vessel edges under difficult conditions, including noisy blurred edges, closely parallel vessels, light reflex phenomena, and very fine vessels. It yields precise vessel width measurements, with subpixel average width errors. We compare the algorithm with several benchmarks from the literature, demonstrating higher segmentation sensitivity and more accurate width measurement.

Subfields: Computer Vision and Pattern Recognition, Ophthalmology, Radiology, Nuclear Medicine and Imaging

View Paper: https://www.semanticscholar.org/paper/43813d78eddd6149eba353bc45131a1cc8cf8322

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

In Ophthalmology

In Radiology, Nuclear Medicine and Imaging

Citations in the corpus

A Discriminatively Trained Fully Connected Conditional Random Field Model for Blood Vessel Segmentation in Fundus Images (2017, background): “In general, most of these strategies are based on applying thresholding, vessel tracking techniques [16] or region-oriented approaches—such as region growing [17]–[19] or active contours [20], [21]—after vessel enhancement.”
Blood vessel segmentation methodologies in retinal images - A survey (2012, methodology): “[88] 2009 Ribbon of Twin active contour model SN, SP Cvr Sum and Cheung [89] 2008 Chan-Vese contour model Visual Nse Zhang et al.”
Trainable COSFIRE filters for vessel delineation with application to retinal images (2015, methodology): “Growing a ‘‘Ribbon of Twins’’ active contour model has been used both for segmentation and for width measurement of the vessels (Al-Diri et al., 2009).”
Joint Segment-Level and Pixel-Wise Losses for Deep Learning Based Retinal Vessel Segmentation (2018, methodology): “[9] designed an active contour model using two pairs of contours to locate each vessel edge.”
Glaucoma risk index: Automated glaucoma detection from color fundus images (2010, methodology): “Most of these techniques exploit the local appearance of the vessel including (i) edge-based (Al-Diri et al. 2009; Martinez-Perez et al. 2007; Sofka and Stewart 2006), template-matching (Ricci and Perfetti 2007; Sofka and Stewart 2006; Wang et al.”
Robust Retinal Vessel Segmentation via Locally Adaptive Derivative Frames in Orientation Scores (2016, methodology): “Model-based approaches extract vessels by applying explicit vessel profile models [11] or deformable models [12].”
Retinal vessel segmentation by improved matched filtering: evaluation on a new high-resolution fundus image database (2013, methodology): “Other authors used an active contour model [25] or model-based region growing segmentation approach [26].”
A Three-Stage Deep Learning Model for Accurate Retinal Vessel Segmentation (2019, methodology): “[18] proposed an active contour model using two pairs of contours to locate each vessel edge, and Zhao et al.”
Fast Retinal Vessel Detection and Measurement Using Wavelets and Edge Location Refinement (2012, methodology): “The data included in the top part of the table are reproduced from [26] ( 2009 IEEE), to which we have added the results of the graph-based algorithm [34] and those obtained by applying the same tests using our algorithm.”
An Ensemble Classification-Based Approach Applied to Retinal Blood Vessel Segmentation (2012, methodology): “The model-based approaches include the vessel profile models [14]–[17], active contour models [18], and geometric models based on level sets [19].”
Blood Vessel Segmentation of Fundus Images by Major Vessel Extraction and Subimage Classification (2015, methodology): “Another model-based vessel segmentation approach proposed in [21] uses active contour models, but suffers from computational complexity as well.”
Hierarchical retinal blood vessel segmentation based on feature and ensemble learning (2015, methodology): “The model based approaches include the vessel profile models [13–16], active contour models [17], and level sets based geometric models [18].”
An approach to localize the retinal blood vessels using bit planes and centerline detection (2012, unknown): “in [75].”