Locating blood vessels in retinal images by piecewise threshold probing of a matched filter response

Abstract

We describe an automated method to locate and outline blood vessels in images of the ocular fundus. Such a tool should prove useful to eye care specialists for purposes of patient screening, treatment evaluation, and clinical study. Our method differs from previously known methods in that it uses local and global vessel features cooperatively to segment the vessel network. We evaluate our method using hand-labeled ground truth segmentations of 20 images. A plot of the operating characteristic shows that our method reduces false positives by as much as 15 times over basic thresholding of a matched filter response (MFR), at up to a 75% true positive rate. For a baseline, we also compared the ground truth against a second hand-labeling, yielding a 90% true positive and a 4% false positive detection rate, on average. These numbers suggest there is still room for a 15% true positive rate improvement, with the same false positive rate, over our method. We are making all our images and hand labelings publicly available for interested researchers to use in evaluating related methods.

Retinal Vessel Detection and Measurement for Computer-aided Medical Diagnosis

“…To make the collected images comparable with the retinal images of the public databases (e.g., STARE and DRIVE), the resolution of the collected images was reduced by half to 760×570. Besides these collected images, the retinal images of STARE (also known as HOOVER database), a public database made available by Hoover et al [3], were tested as well. STARE consists of 20 red-free retinal fundus images captured at 35°of field of view.…”

“…The error and accuracy rates of Hoover's method [3], Jiang's method [6], Staal's method [9], and most likely class for STARE images were calculated and reported in [9]. In order to compare the proposed algorithm with others, we repeated the tests reported in [9].…”

A Comparison and Evaluation of Computerized Methods for Blood Vessel Enhancement and Segmentation in Retinal Images

“…There are two hand-labeling available for the 20 images of test set made by two different human observers. The manually segmented images by 1st human observer are used as ground truth and the segmentations of set B are tested against set A, serving as a human observer reference for performance comparison truth [11], [12]. Table I summarizes the results of both techniques using DRIVE.…”

Segmentation of retinal blood vessels using a novel clustering algorithm (RACAL) with a partial supervision strategy

“…The performance is measured by calculating the false positive rates (FPR) and the true positive rates (TPR), these rates are defined in the same way as in [12]. To evaluate the performance of our proposed algorithm, a set of 20 images publicly available [13] are used, where 10 are normal and 10 contain pathology.…”