Segmentation of optic disc in retinal images using an improved gradient vector flow algorithm

Image segmentation plays an important role in the analysis of retinal images as the extraction of the optic disk provides important cues for accurate diagnosis of various retinopathic diseases. In recent years, gradient vector flow (GVF) based algorithms have been used successfully to successfully segment a variety of medical imagery. However, due to the compromise of internal and external energy forces within the resulting partial differential equations, these methods can lead to less accurate segmentation results in certain cases. In this paper, we propose the use of a new mean shift-based GVF segmentation algorithm that drives the internal/external energies towards the correct direction. The proposed method incorporates a mean shift operation within the standard GVF cost function to arrive at a more accurate segmentation. Experimental results on a large dataset of retinal images demonstrate that the presented method optimally detects the border of the optic disc.

Automatic Analysis of Digital Retinal Images for Glaucoma Detection

In this paper we propose a novel automated glaucoma detection framework for mass-screening that operates on inexpensive retinal cameras. The proposed methodology is based on the assumption that discriminative features for glaucoma diagnosis can be extracted from the optical nerve head structures,
such as the cup-to-disc ratio or the neuro-retinal rim variation. After automatically segmenting the cup and optical disc, these features are feed into a machine learning classifier. Experiments were performed using two different datasets and from the obtained results the proposed technique provides
better performance than approaches based on appearance. A main advantage of our approach is that it only requires a few training samples to provide high accuracy over several different glaucoma stages.

Feedback of personal retinal images appears to have a motivational impact in people with non-proliferative diabetic retinopathy and suboptimal HbA1c: findings of a pilot study

AIM:
To conduct a pilot study to explore the potential impact of visual feedback of personal retinal images on diabetes outcomes.

METHODS:
Twenty-five participants with non-proliferative diabetic retinopathy and suboptimal HbA1c (> 53 mmol/mol; > 7%) were randomized to receive visual feedback of their own retinal images or to a control group. At baseline and 3-month follow-up, HbA1c, standard measures of beliefs, diabetes-related distress and self-care activities were assessed.

RESULTS:
In unadjusted models, relative to controls, the intervention group showed significantly greater improvement in HbA1c at 3-month follow-up (–0.6% vs. +0.3%, P < 0.01), as well as enhanced motivation to improve blood glucose management (P < 0.05).

CONCLUSIONS:
This small pilot study provides preliminary evidence that visual feedback of personal retinal images may offer a practical educational strategy for clinicians in eye care services to improve diabetes outcomes in non-target compliant patients. A fully powered randomized controlled trial is required to confirm these findings and determine the optimal use of feedback to produce sustained effects.

Mathematical-morphology-based edge detection of retinal vessels in retinal images

Of all the important small and medium-sized blood vessels of the human body, retinal blood vessels are the only deep capillary that can be directly observed by a non-traumatic method. Retinal vascular morphology, such as vessel diameter, shape and distribution, is influenced by systemic diseases (Martinez-Perez, Hughes, Thom and Parker 2007). We can use digital fundus photography and analysis of retinal vascular morphology to find the relationship between the changes in vascular morphology and diabetes for the diagnosis of diseases. We aim at developing a retinal image processing system, that can analyze retinal images and provide helpful information for diagnosis. © 2013 Springer-Verlag.

Fast Retinal Vessel Detection and Measurement Using Wavelets and Edge Location Refinement

The relationship between changes in retinal vessel morphology and the onset and progression of diseases such as diabetes, hypertension and retinopathy of prematurity (ROP) has been the subject of several large scale clinical studies. However, the difficulty of quantifying changes in retinal vessels in a sufficiently fast, accurate and repeatable manner has restricted the application of the insights gleaned from these studies to clinical practice. This paper presents a novel algorithm for the efficient detection and measurement of retinal vessels, which is general enough that it can be applied to both low and high resolution fundus photographs and fluorescein angiograms upon the adjustment of only a few intuitive parameters. Firstly, we describe the simple vessel segmentation strategy, formulated in the language of wavelets, that is used for fast vessel detection. When validated using a publicly available database of retinal images, this segmentation achieves a true positive rate of 70.27%, false positive rate of 2.83%, and accuracy score of 0.9371. Vessel edges are then more precisely localised using image profiles computed perpendicularly across a spline fit of each detected vessel centreline, so that both local and global changes in vessel diameter can be readily quantified. Using a second image database, we show that the diameters output by our algorithm display good agreement with the manual measurements made by three independent observers. We conclude that the improved speed and generality offered by our algorithm are achieved without sacrificing accuracy. The algorithm is implemented in MATLAB along with a graphical user interface, and we have made the source code freely available. 

Retinal Vascular Caliber, Iris Color, and Age-Related Macular Degeneration in the Irish Nun Eye Study

Pupose. To evaluate the relationship between retinal vascular caliber (RVC), iris color and age-related macular degeneration (AMD) in elderly Irish nuns. Methods. Data from 1233 participants in the cross-sectional observational Irish Nun Eye Study were assessed from digital photographs with a standardized protocol using computer-assisted software. Macular images were graded according to the modified Wisconsin age-related maculopathy grading system. Regression models were used to assess associations, adjusting for age, mean arterial blood pressure, body mass index, refraction and fellow RVC. Results. In total, 1122 (91%) participants had gradable retinal images of sufficient quality for vessel assessment (mean age: 76.3 years [range: 56-100 years]). In an unadjusted analysis, we found some support for a previous finding that individuals with blue iris color had narrower retinal venules compared to those with brown iris color (P<0.05) but this was no longer significant after adjustment. AMD status was categorized as no AMD, any AMD and late AMD only. Individuals with any AMD (early or late AMD) had significantly narrower arterioles and venules compared to those with no AMD in an unadjusted analysis but this was no longer significant after adjustment. A non-significant reduced risk of any AMD or late AMD only was observed in association with brown compared to blue iris color, in both unadjusted and adjusted analyses. Conclusions. RVC was not significantly associated with iris color or early/late AMD after adjustment for confounders. A lower but non-significant AMD risk was observed in those with brown compared to blue iris color.

Retinal microvascular network attenuation in Alzheimer's disease

INTRODUCTION:Cerebral small-vessel disease has been implicated in the development of Alzheimer’sdisease (AD). The retinal microvasculature enables non-invasive visualization andevaluation of the systemic microcirculation. We evaluated retinal microvascular parametersin a case-control study of AD patients and cognitively-normal controls. 

METHODS:Retinal images were computationally analyzed and quantitative retinal parameters (caliber,fractal dimension, tortuosity, and bifurcation) measured. Regression models were used tocompute odds ratios (OR) and confidence intervals (CI) for AD with adjustment forconfounders. 

RESULTS:Retinal images were available in 213 AD participants and 294 cognitively-normal controls.Persons with lower venular fractal dimension (OR per standard deviation [SD] increase, 0.77[CI: 0.62–0.97]) and lower arteriolar tortuosity (OR per SD increase, 0.78 [CI: 0.63–0.97])were more likely to have AD following appropriate adjustment. 
DISCUSSION:Patients with AD have a sparser retinal microvascular network and retinal microvascularvariation may represent similar pathophysiological events within the cerebralmicrovasculature of patients with AD.

Evaluation of the Retinal Vasculature in Hypertension and Chronic Kidney Disease in an elderly population of Irish nuns

Background: Chronic kidney disease (CKD) and hypertension are global public health problems associated with considerable morbidity, premature mortality and attendant healthcare costs. Previous studies have highlighted that non-invasive examination of the retinal microcirculation can detect microvascular pathology that is associated with systemic disorders of the circulatory system such as hypertension. We examined the associations between retinal vessel caliber (RVC) and fractal dimension (DF), with both hypertension and CKD in elderly Irish nuns.

Methods: Data from 1233 participants in the cross-sectional observational Irish Nun Eye Study (INES) were assessed from digital photographs with a standardized protocol using computer-assisted software. Multivariate regression analyses were used to assess associations with hypertension and CKD, with adjustment for age, body mass index (BMI), refraction, fellow eye RVC, smoking, alcohol consumption, ischemic heart disease (IHD), cerebrovascular accident (CVA), diabetes and medication use.

Results: In total, 1122 (91%) participants (mean age: 76.3 [range: 56-100] years) had gradable retinal images of sufficient quality for blood vessel assessment. Hypertension was significantly associated with a narrower central retinal arteriolar equivalent (CRAE) in a fully adjusted analysis (P = 0.002; effect size= -2.16 μm; 95% confidence intervals [CI]: -3.51, -0.81 μm). No significant associations between other retinal vascular parameters and hypertension or between any retinal vascular parameters and CKD were found.

Conclusions: Individuals with hypertension have significantly narrower retinal arterioles which may afford an earlier opportunity for tailored prevention and treatment options to optimize the structure and function of the microvasculature, providing additional clinical utility. No significant associations between retinal vascular parameters and CKD were detected.

Automated retinal image quality assessment on the UK Biobank dataset for epidemiological studies

Morphological changes in the retinal vascular network are associated with future risk of many systemic and vascular diseases. However, uncertainty over the presence and nature of some of these associations exists. Analysis of data from large population based studies will help to resolve these uncertainties. The QUARTZ (QUantitative Analysis of Retinal vessel Topology and siZe) retinal image analysis system allows automated processing of large numbers of retinal images. However, an image quality assessment module is needed to achieve full automation. In this paper, we propose such an algorithm, which uses the segmented vessel map to determine the suitability of retinal images for use in the creation of vessel morphometric data suitable for epidemiological studies. This includes an effective 3-dimensional feature set and support vector machine classification. A random subset of 800 retinal images from UK Biobank (a large prospective study of 500,000 middle aged adults; where 68,151 underwent retinal imaging) was used to examine the performance of the image quality algorithm. The algorithm achieved a sensitivity of 95.33% and a specificity of 91.13% for the detection of inadequate images. The strong performance of this image quality algorithm will make rapid automated analysis of vascular morphometry feasible on the entire UK Biobank dataset (and other large retinal datasets), with minimal operator involvement, and at low cost.

Evaluation of segmentation algorithms for extraction of RNFL in OCT images

The thickness of the retinal nerve fiber layer (RFNL) has become a diagnose measure for glaucoma assessment. To measure this thickness, accurate segmentation of the RFNL in optical coherence tomography (OCT) images is essential. Identification of a suitable segmentation algorithm will facilitate the enhancement of the RNFL thickness measurement accuracy. This paper investigates the performance of six algorithms in the segmentation of RNFL in OCT images. The algorithms are: normalised cuts, region growing, k-means clustering, active contour, level sets segmentation: Piecewise Gaussian Method (PGM) and Kernelized Method (KM). The performance of the six algorithms are determined through a set of experiments on OCT retinal images. An experimental procedure is used to measure the performance of the tested algorithms. The measured segmentation precision-recall results of the six algorithms are compared and discussed.

Comparison of subjective and objective methods to determine the retinal arterio-venous ratio using fundus photography

Purpose: To assess the inter and intra observer variability of subjective grading of the retinal arterio-venous ratio (AVR) using a visual grading and to compare the subjectively derived grades to an objective method using a semi-automated computer program. Methods: Following intraocular pressure and blood pressure measurements all subjects underwent dilated fundus photography. 86 monochromatic retinal images with the optic nerve head centred (52 healthy volunteers) were obtained using a Zeiss FF450+ fundus camera. Arterio-venous ratios (AVR), central retinal artery equivalent (CRAE) and central retinal vein equivalent (CRVE) were calculated on three separate occasions by one single observer semi-automatically using the software VesselMap (ImedosSystems, Jena, Germany). Following the automated grading, three examiners graded the AVR visually on three separate occasions in order to assess their agreement. Results: Reproducibility of the semi-automatic parameters was excellent (ICCs: 0.97 (CRAE); 0.985 (CRVE) and 0.952 (AVR)). However, visual grading of AVR showed inter grader differences as well as discrepancies between subjectively derived and objectively calculated AVR (all p < 0.000001). Conclusion: Grader education and experience leads to inter-grader differences but more importantly, subjective grading is not capable to pick up subtle differences across healthy individuals and does not represent true AVR when compared with an objective assessment method. Technology advancements mean we no longer rely on opthalmoscopic evaluation but can capture and store fundus images with retinal cameras, enabling us to measure vessel calibre more accurately compared to visual estimation; hence it should be integrated in optometric practise for improved accuracy and reliability of clinical assessments of retinal vessel calibres. © 2014 Spanish General Council of Optometry.

Reliability of retinal vessel calibre measurements using a retinal oximeter

Background: Summarised retinal vessel diameters are linked to systemic vascular pathology. Monochromatic images provide best contrast to measure vessel calibres. However, when obtaining images with a dual wavelength oximeter the red-free image can be extracted as the green channel information only which in turn will reduce the number of photographs taken at a given time. This will reduce patient exposure to the camera flash and could provide sufficient quality images to reliably measure vessel calibres. Methods: We obtained retinal images of one eye of 45 healthy participants. Central retinal arteriolar and central retinal venular equivalents (CRAE and CRVE, respectively) were measured using semi-automated software from two monochromatic images: one taken with a red-free filter and one extracted from the green channel of a dual wavelength oximetry image. Results: Participants were aged between 21 and 62 years, all were normotensive (SBP: 115 (12) mmHg; DBP: 72 (10) mmHg) and had normal intra-ocular pressures (12 (3) mmHg). Bland-Altman analysis revealed good agreement of CRAE and CRVE as obtained from both images (mean bias CRAE = 0.88; CRVE = 2.82). Conclusions: Summarised retinal vessel calibre measurements obtained from oximetry images are in good agreement to those obtained using red-free photographs.

Speckle reduction in optical coherence tomography imaging by affine-motion image registration

Signal-degrading speckle is one factor that can reduce the quality of optical coherence tomography images. We demonstrate the use of a hierarchical model-based motion estimation processing scheme based on an affine-motion model to reduce speckle in optical coherence tomography imaging, by image registration and the averaging of multiple B-scans. The proposed technique is evaluated against other methods available in the literature. The results from a set of retinal images show the benefit of the proposed technique, which provides an improvement in signal-to-noise ratio of the square root of the number of averaged images, leading to clearer visual information in the averaged image. The benefits of the proposed technique are also explored in the case of ocular anterior segment imaging.

Influence of stop position on spectacle magnification

Purpose To develop and use equations of spectacle magnification when the limiting stop is either the entrance pupil of the eye or an artificial pupil in front of a lens. Methods Spectacle magnification was determined for ophthalmic lenses in air and for water environments. The reference was the retinal image for an uncorrected eye in air with a natural pupil. Results When an artificial pupil is placed in front of lenses, spectacle magnification is hardly affected by lens power, unlike the usual situation where the natural pupil is used. The water environment provides interesting influences in which spectacle magnification is highly sensitive to the distance between the cornea and eye entrance pupil. In water, retinal images are approximately 18% bigger than in air. Wearing air-filled goggles in water increases retinal image size by about 13% compared with that when they are not worn. Conclusions The equations extend earlier understanding of spectacle magnification and should be useful for those wishing to determine magnification of ophthalmic lens systems when artificial pupils and environments such as water are used.

Approximate Nearest Neighbour Field based Optic Disk Detection

Approximate Nearest Neighbour Field maps are commonly used by computer vision and graphics community to deal with problems like image completion, retargetting, denoising, etc. In this paper, we extend the scope of usage of ANNF maps to medical image analysis, more specifically to optic disk detection in retinal images. In the analysis of retinal images, optic disk detection plays an important role since it simplifies the segmentation of optic disk and other retinal structures. The proposed approach uses FeatureMatch, an ANNF algorithm, to find the correspondence between a chosen optic disk reference image and any given query image. This correspondence provides a distribution of patches in the query image that are closest to patches in the reference image. The likelihood map obtained from the distribution of patches in query image is used for optic disk detection. The proposed approach is evaluated on five publicly available DIARETDB0, DIARETDB1, DRIVE, STARE and MESSIDOR databases, with total of 1540 images. We show, experimentally, that our proposed approach achieves an average detection accuracy of 99% and an average computation time of 0.2 s per image. (C) 2013 Elsevier Ltd. All rights reserved.