LUNet: deep learning for the segmentation of arterioles and venules in high resolution fundus images

Objective.This study aims to automate the segmentation of retinal arterioles and venules (A/V) from digital fundus images (DFI), as changes in the spatial distribution of retinal microvasculature are indicative of cardiovascular diseases, positioning the eyes as windows to cardiovascular health.Approach.We utilized active learning to create a new DFI dataset with 240 crowd-sourced manual A/V segmentations performed by 15 medical students and reviewed by an ophthalmologist. We then developed LUNet, a novel deep learning architecture optimized for high-resolution A/V segmentation. The LUNet model features a double dilated convolutional block to widen the receptive field and reduce parameter count, alongside a high-resolution tail to refine segmentation details. A custom loss function was designed to prioritize the continuity of blood vessel segmentation.Main Results.LUNet significantly outperformed three benchmark A/V segmentation algorithms both on a local test set and on four external test sets that simulated variations in ethnicity, comorbidities and annotators.Significance.The release of the new datasets and the LUNet model (www.aimlab-technion.com/lirot-ai) provides a valuable resource for the advancement of retinal microvasculature analysis. The improvements in A/V segmentation accuracy highlight LUNet's potential as a robust tool for diagnosing and understanding cardiovascular diseases through retinal imaging.

Leuven-Haifa High-Resolution Fundus Image Dataset for Retinal Blood Vessel Segmentation and Glaucoma Diagnosis

The Leuven-Haifa dataset contains 240 disc-centered fundus images of 224 unique patients (75 patients with normal tension glaucoma, 63 patients with high tension glaucoma, 30 patients with other eye diseases and 56 healthy controls) from the University Hospitals of Leuven. The arterioles and venules of these images were both annotated by master students in medicine and corrected by a senior annotator. All senior segmentation corrections are provided as well as the junior segmentations of the test set. An open-source toolbox for the parametrization of segmentations was developed. Diagnosis, age, sex, vascular parameters as well as a quality score are provided as metadata. Potential reuse is envisioned as the development or external validation of blood vessels segmentation algorithms or study of the vasculature in glaucoma and the development of glaucoma diagnosis algorithms. The dataset is available on the KU Leuven Research Data Repository (RDR).

The Relevance of Arterial Blood Pressure in the Management of Glaucoma Progression: A Systematic Review

BACKGROUND

Glaucoma is one of the leading causes of global blindness and is expected to co-occur more frequently with vascular morbidities in the upcoming years, as both are aging-related diseases. Yet, the pathogenesis of glaucoma is not entirely elucidated and the interplay between intraocular pressure, arterial blood pressure (BP) and ocular perfusion pressure is poorly understood.

OBJECTIVES

This systematic review aims to provide clinicians with the latest literature regarding the management of arterial BP in glaucoma patients.

METHODS

A systematic search was performed in Medline, Embase, Web of Science and Cochrane Library. Articles written in English assessing the influence of arterial BP and systemic antihypertensive treatment of glaucoma and its management were eligible for inclusion. Additional studies were identified by revising references included in selected articles.

RESULTS

80 Articles were included in this systemic review. A bimodal relation between BP and glaucoma progression was found. Both high and low BP increase the risk of glaucoma. Glaucoma progression was, possibly via ocular perfusion pressure variation, strongly associated with nocturnal dipping and high variability in the BP over 24 h.

CONCLUSIONS

We concluded that systemic BP level associates with glaucomatous damage and provided recommendations for the management and study of arterial BP in glaucoma. Prospective clinical trials are needed to further support these recommendations.

Effect of 24-h blood pressure dysregulations and reduced ocular perfusion pressure in open-angle glaucoma progression

BACKGROUND

Low ocular perfusion pressure (OPP), which depends on the mean arterial pressure (MAP) and intraocular pressure (IOP), is associated with glaucoma. We studied 24-h MAP dysregulations and OPP in relation to the progression of glaucoma damage.

METHODS

We retrospectively analyzed 155 normal-tension glaucoma (NTG) and 110 primary open-angle glaucoma (POAG) patients aged 18 years old followed at the University Hospital Leuven with repeated visual field tests ( n  = 7000 measures, including both eyes) who underwent 24-h ambulatory blood pressure monitoring. Twenty-four-hour MAP dysregulations were variability independent of the mean (VIM), and the five lowest dips in MAP readings over 24 h. OPP was the difference between 2/3 of the MAP and IOP. Glaucoma progression was the deterioration of the visual field, expressed as decibel (dB) changes in mean deviation analyzed by applying multivariable linear mixed regression models.

RESULTS

The mean age was 68 years (53% were women). High 24-h VIMmap was associated with glaucoma progression in POAG ( P  < 0.001) independently of the 24-h MAP level. The estimated changes in mean deviation in relation to dip MAP measures ranged from -2.84 dB [95% confidence interval (CI) -4.12 to -1.57] to -2.16 dB (95% CI -3.46 to -0.85) in POAG. Reduced OPP along with high variability and dips in MAP resulted in worse mean deviation deterioration.

CONCLUSION

The progression of glaucoma damage associates with repetitive and extreme dips in MAP caused by high variability in MAP throughout 24 h. This progression exacerbates if 24-h MAP dysregulations occur along with reduced OPP.

FundusQ-Net: A regression quality assessment deep learning algorithm for fundus images quality grading

OBJECTIVE

Ophthalmological pathologies such as glaucoma, diabetic retinopathy and age-related macular degeneration are major causes of blindness and vision impairment. There is a need for novel decision support tools that can simplify and speed up the diagnosis of these pathologies. A key step in this process is to automatically estimate the quality of the fundus images to make sure these are interpretable by a human operator or a machine learning model. We present a novel fundus image quality scale and deep learning (DL) model that can estimate fundus image quality relative to this new scale.

METHODS

A total of 1245 images were graded for quality by two ophthalmologists within the range 1-10, with a resolution of 0.5. A DL regression model was trained for fundus image quality assessment. The architecture used was Inception-V3. The model was developed using a total of 89,947 images from 6 databases, of which 1245 were labeled by the specialists and the remaining 88,702 images were used for pre-training and semi-supervised learning. The final DL model was evaluated on an internal test set (n=209) as well as an external test set (n=194).

RESULTS

The final DL model, denoted FundusQ-Net, achieved a mean absolute error of 0.61 (0.54-0.68) on the internal test set. When evaluated as a binary classification model on the public DRIMDB database as an external test set the model obtained an accuracy of 99%.

SIGNIFICANCE

the proposed algorithm provides a new robust tool for automated quality grading of fundus images.

Normal tension glaucoma: A dynamic optical coherence tomography angiography study

Purpose

Vascular dysregulation seems to play a role in the pathogenesis of glaucoma, in particular normal tension glaucoma (NTG). The development of optical coherence tomography angiography (OCTA) enabled the measurement of the retinal microvasculature non-invasively and with high repeatability. Nonetheless, only a few studies transformed OCTA into a dynamic examination employing a sympathomimetic stimulus. The goal of this study was to use this dynamic OCTA exam (1) to differentiate healthy individuals from glaucoma patients and (2) to distinguish glaucoma subcategories, NTG and high-tension primary open angle glaucoma (POAG).

Methods

Retinal vessel density (VD) in NTG patients (n = 16), POAG patients (n = 12), and healthy controls (n = 14) was compared before and during a hand grip test with a hydraulic dynamometer.

Results

At baseline, mean peripapillary VD was lower in POAG and NTG (42.6 and 48.5%) compared to healthy controls (58.1%; p < 0.001) and higher in NTG compared to POAG (p = 0.024) when corrected for mean arterial pressure (MAP). Peripapillary and macular (superficial and deep) VD differences were found for gender, age, and baseline MAP. No change in VD occurred (pre-/post-stimulus) in any of the groups.

Conclusion

Retinal VD loss in glaucoma patients was confirmed and the necessity to correct for gender, age and especially MAP was established. Although replication in a larger population is necessary, OCTA might not be the most suitable method to dynamically evaluate the retinal microvasculature.

Automatic Segmentation of the Optic Nerve Head Region in Optical Coherence Tomography: A Methodological Review

The optic nerve head (ONH) represents the intraocular section of the optic nerve, which is prone to damage by intraocular pressure (IOP). The advent of optical coherence tomography (OCT) has enabled the evaluation of novel ONH parameters, namely the depth and curvature of the lamina cribrosa (LC). Together with the Bruch's membrane minimum-rim-width (BMO-MRW), these seem to be promising ONH parameters for diagnosis and monitoring of retinal diseases such as glaucoma. Nonetheless, these OCT derived biomarkers are mostly extracted through manual segmentation, which is time-consuming and prone to bias, thus limiting their usability in clinical practice. The automatic segmentation of ONH in OCT scans could further improve the current clinical management of glaucoma and other diseases. This review summarizes the current state-of-the-art in automatic segmentation of the ONH in OCT. PubMed and Scopus were used to perform a systematic review. Additional works from other databases (IEEE, Google Scholar and ARVO IOVS) were also included, resulting in a total of 29 reviewed studies. For each algorithm, the methods, the size and type of dataset used for validation, and the respective results were carefully analysed. The results show a lack of consensus regarding the definition of segmented regions, extracted parameters and validation approaches, highlighting the importance and need of standardized methodologies for ONH segmentation. Only with a concrete set of guidelines, these automatic segmentation algorithms will build trust in data-driven segmentation models and be able to enter clinical practice.

Retinal imaging biomarkers of neurodegenerative diseases

The timely detection of neurodegenerative diseases is central to improving clinical care as well as enabling the development and deployment of disease-modifying therapies. Retinal imaging is emerging as a method to detect features of a number of neurodegenerative diseases, given the anatomical and functional similarities between the retina and the brain. This review provides an overview of the current status of retinal imaging biomarkers of neurodegenerative diseases including Alzheimer's disease, Parkinson's disease, Lewy body dementia, frontotemporal dementia, Huntington's disease and multiple sclerosis. Whilst research findings are promising, efforts to harmonise study designs and imaging methods will be important in translating these findings into clinical care. Doing so may mean that eye care providers will play important roles in the detection of a variety of neurodegenerative diseases in future.

Combination of snapshot hyperspectral retinal imaging and optical coherence tomography to identify Alzheimer's disease patients

INTRODUCTION

The eye offers potential for the diagnosis of Alzheimer's disease (AD) with retinal imaging techniques being explored to quantify amyloid accumulation and aspects of neurodegeneration. To assess these changes, this proof-of-concept study combined hyperspectral imaging and optical coherence tomography to build a classification model to differentiate between AD patients and controls.

METHODS

In a memory clinic setting, patients with a diagnosis of clinically probable AD (n = 10) or biomarker-proven AD (n = 7) and controls (n = 22) underwent non-invasive retinal imaging with an easy-to-use hyperspectral snapshot camera that collects information from 16 spectral bands (460-620 nm, 10-nm bandwidth) in one capture. The individuals were also imaged using optical coherence tomography for assessing retinal nerve fiber layer thickness (RNFL). Dedicated image preprocessing analysis was followed by machine learning to discriminate between both groups.

RESULTS

Hyperspectral data and retinal nerve fiber layer thickness data were used in a linear discriminant classification model to discriminate between AD patients and controls. Nested leave-one-out cross-validation resulted in a fair accuracy, providing an area under the receiver operating characteristic curve of 0.74 (95% confidence interval [0.60-0.89]). Inner loop results showed that the inclusion of the RNFL features resulted in an improvement of the area under the receiver operating characteristic curve: for the most informative region assessed, the average area under the receiver operating characteristic curve was 0.70 (95% confidence interval [0.55, 0.86]) and 0.79 (95% confidence interval [0.65, 0.93]), respectively. The robust statistics used in this study reduces the risk of overfitting and partly compensates for the limited sample size.

CONCLUSIONS

This study in a memory-clinic-based cohort supports the potential of hyperspectral imaging and suggests an added value of combining retinal imaging modalities. Standardization and longitudinal data on fully amyloid-phenotyped cohorts are required to elucidate the relationship between retinal structure and cognitive function and to evaluate the robustness of the classification model.

Hyperspectral Imaging and the Retina: Worth the Wave?

Purpose

Hyperspectral imaging is gaining attention in the biomedical field because it generates additional spectral information to study physiological and clinical processes. Several technologies have been described; however an independent, systematic literature overview is lacking, especially in the field of ophthalmology. This investigation is the first to systematically overview scientific literature specifically regarding retinal hyperspectral imaging.

Methods

A systematic literature review was conducted, in accordance with PRISMA Statement 2009 criteria, in four bibliographic databases: Medline, Embase, Cochrane Database of Systematic Reviews, and Web of Science.

Results

Fifty-six articles were found that meet the review criteria. A range of techniques was reported: Fourier analysis, liquid crystal tunable filters, tunable laser sources, dual-slit monochromators, dispersive prisms and gratings, computed tomography, fiber optics, and Fabry-Perrot cavity filter covered complementary metal oxide semiconductor. We present a narrative synthesis and summary tables of findings of the included articles, because methodologic heterogeneity and diverse research topics prevented a meta-analysis being conducted.

Conclusions

Application in ophthalmology is still in its infancy. Most previous experiments have been performed in the field of retinal oximetry, providing valuable information in the diagnosis and monitoring of various ocular diseases. To date, none of these applications have graduated to clinical practice owing to the lack of sufficiently large validation studies.

Translational Relevance

Given the promising results that smaller studies show for hyperspectral imaging (e.g., in Alzheimer's disease), advanced research in larger validation studies is warranted to determine its true clinical potential.