Automated 3-D intraretinal layer segmentation of macular spectral-domain optical coherence tomography images

Unifying Structure and Function Towards a Comprehensive Macular Evaluation in Eye Disorders: A Multi-Modal Approach Using Microperimetry and Optical Coherence Tomography

OBJECTIVE

We present a new automatic and comprehensive framework to evaluate retinal sub-layer thickness and visual sensitivity at precise retinal locations through multi-modal registration of confocal scanning laser ophthalmoscope (cSLO) images obtained separately from an optical coherence tomography (OCT) device and a microperimeter.

METHODS

We map consecutive B-scans onto the cSLO images after accounting for eye motion. Next, we coarsely register the SLO-microperimetry and cSLO-OCT images using SIFT, followed by precise elastic image registration. Subsequently, we ensure the quality of the co-registered images through single-particle and object tracking of the warped microperimetry test locations. Finally, the retinal thickness is queried from the segmented retinal layers in the co-registered space. A manual mode involving projective transformation accounting for perspective distortions in the images arising from the two modalities is also included.

RESULTS

Validation using retinal images of 8 adults with albinism, 16 adults with amblyopia, 3 adults with macular diseases, and 15 visually healthy adults showed results with excellent reliability.

CONCLUSION

The proposed framework enables the evaluation of retinal thickness by utilizing precise structural and functional relationships of the eye through a self-assessing multi-tiered approach.

SIGNIFICANCE

Our framework lays the foundation towards a comprehensive structure-function assessment of the macular region in various eye disorders.

Novel 2D/3D vascular biomarkers reveal association between fundus changes and coronary heart disease

PURPOSE

To compare structural and vascular differences in the macular region of the retina using optical coherence tomography (OCT)/OCT angiography (OCTA) between coronary angiography (CAG)-confirmed coronary heart disease (CHD) patients and non-CHD individuals.

METHODS

The study included 340 eyes from 180 CHD patients and 136 eyes from 68 controls. Imaging was conducted using the AngioVue OCT device with a macula-centered 6 mm ∗ 6 mm field of view. Retinal thickness and 2D/3D vascular-related biomarkers were derived using existing retinal layer segmentation software, and our previously proposed 2D/3D vascular and 3D foveal avascular zone segmentation methods. Statistical analyses included t-tests, Mann-Whitney U tests, chi-square tests, and Pearson's correlation.

RESULTS

The CHD group exhibited significantly lower retinal nerve fiber layer (RNFL) thickness (r = -0.20, P < 0.001) in the inner inferior (I) region, based on macular region layer segmentation. For the 3D OCT images, as defined by the ETDRS grid, both the inner and outer retina layers in the outer superior (out-S) region were significantly thinner in the CHD group. The CHD group showed significantly lower overall 2D fractal dimension (FD) (1.72 ± 0.03 vs. 1.73 ± 0.02, P < 0.001) and vessel skeleton density (VSD) (26.61 ± 4.52 vs. 28.50 ± 3.40, P < 0.001) compared to the control group. The proposed 3D vascular density (VD) feature showed a significant difference between the groups (19.23 ± 5.67 vs. 20.69 ± 5.15, P = 0.048).

CONCLUSION

Thinning of retinal thickness and reduced vascular density are associated with CHD and may serve as valuable, cost-effective biomarkers for assessing coronary artery disease assessment.

Validation of Deep Learning-Based Automatic Retinal Layer Segmentation Algorithms for Age-Related Macular Degeneration with 2 Spectral-Domain OCT Devices

Purpose

Segmentations of retinal layers in spectral-domain OCT (SD-OCT) images serve as a crucial tool for identifying and analyzing the progression of various retinal diseases, encompassing a broad spectrum of abnormalities associated with age-related macular degeneration (AMD). The training of deep learning algorithms necessitates well-defined ground truth labels, validated by experts, to delineate boundaries accurately. However, this resource-intensive process has constrained the widespread application of such algorithms across diverse OCT devices. This work validates deep learning image segmentation models across multiple OCT devices by testing robustness in generating clinically relevant metrics.

Design

Prospective comparative study.

Participants

Adults >50 years of age with no AMD to advanced AMD, as defined in the Age-Related Eye Disease Study, in ≥1 eye, were enrolled. Four hundred two SD-OCT scans were used in this study.

Methods

We evaluate 2 separate state-of-the-art segmentation algorithms through a training process using images obtained from 1 OCT device (Heidelberg-Spectralis) and subsequent testing using images acquired from 2 OCT devices (Heidelberg-Spectralis and Zeiss-Cirrus). This assessment is performed on a dataset that encompasses a range of retinal pathologies, spanning from disease-free conditions to severe forms of AMD, with a focus on evaluating the device independence of the algorithms.

Main Outcome Measures

Performance metrics (including mean squared error, mean absolute error [MAE], and Dice coefficients) for the segmentations of the internal limiting membrane (ILM), retinal pigment epithelium (RPE), and RPE to Bruch's membrane region, along with en face thickness maps, volumetric estimations (in mm3). Violin plots and Bland-Altman plots comparing predictions against ground truth are also presented.

Results

The UNet and DeepLabv3, trained on Spectralis B-scans, demonstrate clinically useful outcomes when applied to Cirrus test B-scans. Review of the Cirrus test data by 2 independent annotators revealed that the aggregated MAE in pixels for ILM was 1.82 ± 0.24 (equivalent to 7.0 ± 0.9 μm) and for RPE was 2.46 ± 0.66 (9.5 ± 2.6 μm). Additionally, the Dice similarity coefficient for the RPE drusen complex region, comparing predictions to ground truth, reached 0.87 ± 0.01.

Conclusions

In the pursuit of task-specific goals such as retinal layer segmentation, a segmentation network has the capacity to acquire domain-independent features from a large training dataset. This enables the utilization of the network to execute tasks in domains where ground truth is hard to generate.

Financial Disclosures

Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

Characterizing Bruch's membrane: State-of-the-art imaging, computational segmentation, and biologic models in retinal disease and health

The Bruch's membrane (BM) is an acellular, extracellular matrix that lies between the choroid and retinal pigment epithelium (RPE). The BM plays a critical role in retinal health, performing various functions including biomolecule diffusion and RPE support. The BM is also involved in many retinal diseases, and insights into BM dysfunction allow for further understanding of the pathophysiology of various chorioretinal pathologies. Thus, characterization of the BM serves as an important area of research to further understand its involvement in retinal disease. In this article, we provide a review of various advancements in characterizing and visualizing the BM. We provide an overview of the BM in retinal health, as well as changes observed in aging and disease. We then describe current state-of-the-art imaging modalities and advances to further visualize the BM including various types of optical coherence tomography imaging, near-infrared reflectance (NIR), and autofluorescence imaging and tissue matrix-assisted laser desorption/ionization imaging mass spectrometry (MALDI-IMS). Following advances in imaging of the BM, we describe animal, cellular, and synthetic models that have been developed to further visualize the BM. Following this section, we provide an overview of deep learning in retinal imaging and describe advances in computational and artificial intelligence (AI) techniques to provide automated segmentation of the BM and BM opening. We conclude this section considering the clinical implications of these segmentation techniques. Ultimately, the diverse advances aimed to further characterize the BM may allow for deeper insights into the involvement of this critical structure in retinal health and disease.

SD-LayerNet: Robust and label-efficient retinal layer segmentation via anatomical priors

BACKGROUND AND OBJECTIVES

Automated, anatomically coherent retinal layer segmentation in optical coherence tomography (OCT) is one of the most important components of retinal disease management. However, current methods rely on large amounts of labeled data, which can be difficult and expensive to obtain. In addition, these systems tend often propose anatomically impossible results, which undermines their clinical reliability.

METHODS

This study introduces a semi-supervised approach to retinal layer segmentation that leverages large amounts of unlabeled data and anatomical prior knowledge related to the structure of the retina. During training, we use a novel topological engine that converts inferred retinal layer boundaries into pixel-wise structured segmentations. These compose a set of anatomically valid disentangled representations which, together with predicted style factors, are used to reconstruct the input image. At training time, the retinal layer boundaries and pixel-wise predictions are both guided by reference annotations, where available, but more importantly by innovatively exploiting anatomical priors that improve the performance, robustness and coherence of the method even if only a small amount of labeled data is available.

RESULTS

Exhaustive experiments with respect to label efficiency, contribution of unsupervised data and robustness to different acquisition settings were conducted. The proposed method showed state of-the-art performance on all the studied public and internal datasets, specially in low annotated data regimes. Additionally, the model was able to make use of unlabeled data from a different domain with only a small performance drop in comparison to a fully-supervised setting.

CONCLUSION

A novel, robust, label-efficient retinal layer segmentation method was proposed. The approach has shown state-of-the-art layer segmentation performance with a fraction of the training data available, while at the same time, its robustness against domain shift was also shown.

Progressive inner retinal neurodegeneration in non-proliferative macular telangiectasia type 2

PURPOSE

Patients with non-proliferative macular telangiectasia type 2 (MacTel) have ganglion cell layer (GCL) and nerve fibre layer (NFL) loss, but it is unclear whether the thinning is progressive. We quantified the change in retinal layer thickness over time in MacTel with and without diabetes.

METHODS

In this retrospective, multicentre, comparative case series, subjects with MacTel with at least two optical coherence tomographic (OCT) scans separated by >9 months OCTs were segmented using the Iowa Reference Algorithms. Mean NFL and GCL thickness was computed across the total area of the early treatment diabetic retinopathy study grid and for the inner temporal region to determine the rate of thinning over time. Mixed effects models were fit to each layer and region to determine retinal thinning for each sublayer over time.

RESULTS

115 patients with MacTel were included; 57 patients (50%) had diabetes and 21 (18%) had a history of carbonic anhydrase inhibitor (CAI) treatment. MacTel patients with and without diabetes had similar rates of thinning. In patients without diabetes and untreated with CAIs, the temporal parafoveal NFL thinned at a rate of -0.25±0.09 µm/year (95% CI [-0.42 to -0.09]; p=0.003). The GCL in subfield 4 thinned faster in the eyes treated with CAI (-1.23±0.21 µm/year; 95% CI [-1.64 to -0.82]) than in untreated eyes (-0.19±0.16; 95% CI [-0.50, 0.11]; p<0.001), an effect also seen for the inner nuclear layer. Progressive outer retinal thinning was observed.

CONCLUSIONS

Patients with MacTel sustain progressive inner retinal neurodegeneration similar to those with diabetes without diabetic retinopathy. Further research is needed to understand the consequences of retinal thinning in MacTel.

Looking outside the box with a pathology aware AI approach for analyzing OCT retinal images in Stargardt disease

Stargardt disease type 1 (STGD1) is a genetic disorder that leads to progressive vision loss, with no approved treatments currently available. The development of effective therapies faces the challenge of identifying appropriate outcome measures that accurately reflect treatment benefits. Optical Coherence Tomography (OCT) provides high-resolution retinal images, serving as a valuable tool for deriving potential outcome measures, such as retinal thickness. However, automated segmentation of OCT images, particularly in regions disrupted by degeneration, remains complex. In this study, we propose a deep learning-based approach that incorporates a pathology-aware loss function to segment retinal sublayers in OCT images from patients with STGD1. This method targets relatively unaffected regions for sublayer segmentation, ensuring accurate boundary delineation in areas with minimal disruption. In severely affected regions, identified by a box detection model, the total retina is segmented as a single layer to avoid errors. Our model significantly outperforms standard models, achieving an average Dice coefficient of [Formula: see text] for total retina and [Formula: see text] for retinal sublayers. The most substantial improvement was in the segmentation of the photoreceptor inner segment, with Dice coefficient increasing by [Formula: see text]. This approach provides a balance between granularity and reliability, making it suitable for clinical application in tracking disease progression and evaluating therapeutic efficacy.

RetOCTNet: Deep Learning-Based Segmentation of OCT Images Following Retinal Ganglion Cell Injury

Purpose

We present RetOCTNet, a deep learning tool to segment the retinal nerve fiber layer (RNFL) and total retinal thickness automatically from optical coherence tomography (OCT) scans in rats following retinal ganglion cell (RGC) injury.

Methods

We created unilateral RGC injury by ocular hypertension (OHT) or optic nerve crush (ONC), and contralateral eyes were not injured. We manually segmented the RNFL and total retina of 3.0-mm radial OCT scans (1000 A-scans per B-scan, 20 frames per B-scan) as ground truth (n = 192 scans). We used these segmentations for training (80%), testing (10%), and validation (10%) to optimize the F1 score. To determine the generalizability of RetOCTNet, we tested it on volumetric scans of a separate cohort at baseline and 4, 8, and 12 weeks post-ONC.

Results

RetOCTNet's segmentations achieved high F1 scores relative to the ground-truth segmentations created by human annotators: 0.88 (RNFL) and 0.98 (retinal thickness) for control eyes, 0.84 and 0.98 for OHT eyes, and 0.78 and 0.96 for ONC eyes, respectively. On volumetric scans 12 weeks post-ONC, RetOCTNet calculated thinning of 29.49% and 10.82% in the RNFL and retina at the optic nerve head (ONH) and thinning of 38.34% and 9.85% in the RNFL and retina superior to the ONH.

Conclusions

RetOCTNet can segment the RNFL and total retinal thickness of both radial and volume OCT scans. RetOCTNet can be applied to longitudinally monitor RNFL in rodent models of RGC injury.

Translational Relevance

This tool automates RNFL and retinal thickness measurement for rat OCT scans following RGC injury, reducing analysis time and increasing the consistency between studies.

A Survey on Optical Coherence Tomography-Technology and Application

This paper reviews the main research on Optical Coherence Tomography (OCT), focusing on the progress and advancements made by researchers over the past three decades in its methods and medical imaging applications. By analyzing existing studies and developments, this review aims to provide a foundation for future research in the field.

OCT Changes Observed during the Progression of Early Age-Related Macular Degeneration

Purpose

Automated retinal cell layer segmentation empowers OCT as a precise tool for characterizing morphologic features of retinal health throughout age-related macular degeneration (AMD) progression, particularly in advance of more visible biomarkers such as drusen and macular pigmentary changes. Few studies have examined OCT changes in eyes progressing from early to intermediate disease, or combined examinations of cell layer thickness, reflectivity, and heterogeneity. Therefore, this study analyzed OCTs from eyes progressing from early to intermediate AMD to identify changes in retinal morphology and reflectivity that may serve as biomarkers of early progression.

Design

Retrospective cohort study.

Participants

Patients ≥50 years with a diagnosis of AMD and with high-quality ipsilateral OCTs in both early and intermediate stage disease.

Methods

Fifty OCTs from 25 patients were automatically segmented using a previously validated artificial intelligence-driven algorithm. Changes in the mean and standard deviation of cell layer thickness and reflectivity with progression through stages were calculated for 90 retinal volumes with the help of a novel Python-based analysis tool.

Main Outcome Measures

The primary outcomes were significant changes to cell layer thickness, reflectivity, and heterogeneity with progression of AMD.

Results

With progression from early to intermediate disease, photoreceptor outer segments diffusely thinned. Within the ellipsoid zone, the fovea and parafovea were thinned with a simultaneous increase in thickness variability and a decrease in parafoveal reflectivity. The retinal pigment epithelium-Bruch's membrane complex underwent diffuse thickening and increased thickness variability alongside a decrease in foveal and parafoveal reflectivity.

Conclusions

These findings correlate with the known histopathology of early AMD and identify measurable OCT trends through the earliest stages of disease.

Financial Disclosures

Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

Macular patterns of neuronal and visual field loss in recovered optic neuritis identified by machine learning

We used machine learning to investigate the residual visual field (VF) deficits and macula retinal ganglion cell (RGC) thickness loss patterns in recovered optic neuritis (ON). We applied archetypal analysis (AA) to 377 same-day pairings of 10-2 VF and optical coherence tomography (OCT) macula images from 93 ON eyes and 70 normal fellow eyes ≥ 90 days after acute ON. We correlated archetype (AT) weights (total weight = 100%) of VFs and total retinal thickness (TRT), inner retinal thickness (IRT), and macular ganglion cell-inner plexiform layer (GCIPL) thickness. AA showed most ON eyes had a 10-2 VF pattern like the normal fellow eye VF, despite having markedly thinner GCIPL patterns. AA identified 7 VF and 11 retinal thickness ATs for each OCT model. The normal VF AT constituted 80% of ON eyes and 90% of normal fellow eyes. The most common GCIPL AT consisted of diffuse thinning. We identified significant correlations for the normal AT weights using OCT AT weights of five GCIPL ATs (r = 0.45), four TRT ATs (0.53) and two IRT ATs (0.42). Following acute ON, most eyes had complete 10-2 VF recovery despite significant GCIPL thinning, suggesting compensatory mechanisms for vision.

VLFATRollout: Fully transformer-based classifier for retinal OCT volumes

BACKGROUND AND OBJECTIVE

Despite the promising capabilities of 3D transformer architectures in video analysis, their application to high-resolution 3D medical volumes encounters several challenges. One major limitation is the high number of 3D patches, which reduces the efficiency of the global self-attention mechanisms of transformers. Additionally, background information can distract vision transformers from focusing on crucial areas of the input image, thereby introducing noise into the final representation. Moreover, the variability in the number of slices per volume complicates the development of models capable of processing input volumes of any resolution while simple solutions like subsampling may risk losing essential diagnostic details.

METHODS

To address these challenges, we introduce an end-to-end transformer-based framework, variable length feature aggregator transformer rollout (VLFATRollout), to classify volumetric data. The proposed VLFATRollout enjoys several merits. First, the proposed VLFATRollout can effectively mine slice-level fore-background information with the help of transformer's attention matrices. Second, randomization of volume-wise resolution (i.e. the number of slices) during training enhances the learning capacity of the learnable positional embedding (PE) assigned to each volume slice. This technique allows the PEs to generalize across neighboring slices, facilitating the handling of high-resolution volumes at the test time.

RESULTS

VLFATRollout was thoroughly tested on the retinal optical coherence tomography (OCT) volume classification task, demonstrating a notable average improvement of 5.47% in balanced accuracy over the leading convolutional models for a 5-class diagnostic task. These results emphasize the effectiveness of our framework in enhancing slice-level representation and its adaptability across different volume resolutions, paving the way for advanced transformer applications in medical image analysis. The code is available at https://github.com/marziehoghbaie/VLFATRollout/.

Influence of drusenoid pigment epithelial detachments on the progression of age-related macular degeneration and visual acuity

OBJECTIVE

To analyze the presence and morphologic characteristics of drusenoid pigment epithelial detachments (DPEDs) in spectral-domain optical coherence tomography (SD-OCT) in Caucasian patients with early and intermediate age-related macular degeneration (AMD) as well as the influence of these characteristics on best-corrected visual acuity (BCVA) and disease progression.

DESIGN

Prospective observational cohort study.

PARTICIPANTS

89 eyes of 56 patients with early and intermediate AMD.

METHODS

Examinations consisted of BCVA, SD-OCT, and indocyanine green angiography. Evaluated parameters included drusen type, mean drusen height and -volume, the presence of DPED, DPED maximum height, -maximum diameter, -volume, topographic location, the rate of DPED collapse, and the development of macular neovascularization (MNV) or geographic atrophy (GA).

RESULTS

DPED maximum height (162.34 µm ± 75.70 μm, p = 0.019) was significantly associated with the development of GA and MNV. For each additional 100 μm in maximum height, the odds of developing any late AMD (GA or MNV) increased by 2.23 (95% CI = 1.14-4.35). The presence of DPED (44 eyes, p = 0.01), its volume (0.20 mm ± 0.20 mm, p = 0.01), maximum diameter (1860.87 μm ± 880.74 μm, p = 0.03), maximum height (p < 0.001) and topographical location in the central millimetre (p = 0.004) of the Early Treatment Diabetic Retinopathy Study (ETDRS)-Grid were significantly correlated with BCVA at the last follow-up (0.15logMAR ± 0.20logMAR; Snellen equivalent approximately 20/28). DPEDs occurred significantly less in the outer quadrants than in the central millimetre and inner quadrants of ETDRS-Grid (all p values < 0.001).

CONCLUSIONS

The height of drusen and DPEDs is a biomarker that is significantly associated with the development of late AMD and visual loss. DPEDs affect predominantly the center and inner quadrants of the ETDRS-Grid.

Signature of Altered Retinal Microstructures and Electrophysiology in Schizophrenia Spectrum Disorders Is Associated With Disease Severity and Polygenic Risk

BACKGROUND

Optical coherence tomography and electroretinography studies have revealed structural and functional retinal alterations in individuals with schizophrenia spectrum disorders (SSDs). However, it remains unclear which specific retinal layers are affected; how the retina, brain, and clinical symptomatology are connected; and how alterations of the visual system are related to genetic disease risk.

METHODS

Optical coherence tomography, electroretinography, and brain magnetic resonance imaging were applied to comprehensively investigate the visual system in a cohort of 103 patients with SSDs and 130 healthy control individuals. The sparse partial least squares algorithm was used to identify multivariate associations between clinical disease phenotype and biological alterations of the visual system. The association of the revealed patterns with individual polygenic disease risk for schizophrenia was explored in a post hoc analysis. In addition, covariate-adjusted case-control comparisons were performed for each individual optical coherence tomography and electroretinography parameter.

RESULTS

The sparse partial least squares analysis yielded a phenotype-eye-brain signature of SSDs in which greater disease severity, longer duration of illness, and impaired cognition were associated with electrophysiological alterations and microstructural thinning of most retinal layers. Higher individual loading onto this disease-relevant signature of the visual system was significantly associated with elevated polygenic risk for schizophrenia. In case-control comparisons, patients with SSDs had lower macular thickness, thinner retinal nerve fiber and inner plexiform layers, less negative a-wave amplitude, and lower b-wave amplitude.

CONCLUSIONS

This study demonstrates multimodal microstructural and electrophysiological retinal alterations in individuals with SSDs that are associated with disease severity and individual polygenic burden.

AI in the clinical management of GA: A novel therapeutic universe requires novel tools

Regulatory approval of the first two therapeutic substances for the management of geographic atrophy (GA) secondary to age-related macular degeneration (AMD) is a major breakthrough following failure of numerous previous trials. However, in the absence of therapeutic standards, diagnostic tools are a key challenge as functional parameters in GA are hard to provide. The majority of anatomical biomarkers are subclinical, necessitating advanced and sensitive image analyses. In contrast to fundus autofluorescence (FAF), optical coherence tomography (OCT) provides high-resolution visualization of neurosensory layers, including photoreceptors, and other features that are beyond the scope of human expert assessment. Artificial intelligence (AI)-based methodology strongly enhances identification and quantification of clinically relevant GA-related sub-phenotypes. Introduction of OCT-based biomarker analysis provides novel insight into the pathomechanisms of disease progression and therapeutic, moving beyond the limitations of conventional descriptive assessment. Accordingly, the Food and Drug Administration (FDA) has provided a paradigm-shift in recognizing ellipsoid zone (EZ) attenuation as a primary outcome measure in GA clinical trials. In this review, the transition from previous to future GA classification and management is described. With the advent of AI tools, diagnostic and therapeutic concepts have changed substantially in monitoring and screening of GA disease. Novel technology combined with pathophysiological knowledge and understanding of the therapeutic response to GA treatments, is currently opening the path for an automated, efficient and individualized patient care with great potential to improve access to timely treatment and reduce health disparities.

Validation of Inter-Reader Agreement/Consistency for Quantification of Ellipsoid Zone Integrity and Sub-RPE Compartmental Features Across Retinal Diseases

BACKGROUND

An unmet need exists when clinically assessing retinal and layer-based features of retinal diseases. Therefore, quantification of retinal-layer-thicknesses/fluid volumes using deep-learning-augmented platforms to reproduce human-obtained clinical measurements is needed.

METHODS

In this analysis, 210 spectral-domain optical coherence tomography (SD-OCT) scans (30 without pathology, 60 dry age-related macular degeneration [AMD], 60 wet AMD, and 60 diabetic macular edema [total 23,625 B-scans]) were included. A fully automated segmentation platform segmented four retinal layers for compartmental assessment (internal limiting membrane, ellipsoid zone [EZ], retinal pigment epithelium [RPE], and Bruch's membrane). Two certified OCT readers independently completed manual segmentation and B-scan level validation of automated segmentation, with segmentation correction when needed (semi-automated). Certified reader metrics were compared to gold standard metrics using intraclass correlation coefficients (ICCs) to assess overall agreement. Across different diseases, several metrics generated from automated segmentations approached or matched human readers performance.

RESULTS

Absolute ICCs for retinal mean thickness measurements showed excellent agreement (range 0.980-0.999) across four cohorts. EZ-RPE thickness values and sub-RPE compartment ICCs demonstrated excellent agreement (ranges of 0.953-0.987 and 0.944-0.997, respectively) for full dataset, dry-AMD, and wet-AMD cohorts.

CONCLUSIONS

Analyses demonstrated high reliability and consistency of segmentation of outer retinal compartmental features using a completely human/manual approach or a semi-automated approach to segmentation. These results support the critical role that measuring features, such as photoreceptor preservation through EZ integrity, in future clinical trials may optimize clinical care.

Metadata-enhanced contrastive learning from retinal optical coherence tomography images

Deep learning has potential to automate screening, monitoring and grading of disease in medical images. Pretraining with contrastive learning enables models to extract robust and generalisable features from natural image datasets, facilitating label-efficient downstream image analysis. However, the direct application of conventional contrastive methods to medical datasets introduces two domain-specific issues. Firstly, several image transformations which have been shown to be crucial for effective contrastive learning do not translate from the natural image to the medical image domain. Secondly, the assumption made by conventional methods, that any two images are dissimilar, is systematically misleading in medical datasets depicting the same anatomy and disease. This is exacerbated in longitudinal image datasets that repeatedly image the same patient cohort to monitor their disease progression over time. In this paper we tackle these issues by extending conventional contrastive frameworks with a novel metadata-enhanced strategy. Our approach employs widely available patient metadata to approximate the true set of inter-image contrastive relationships. To this end we employ records for patient identity, eye position (i.e. left or right) and time series information. In experiments using two large longitudinal datasets containing 170,427 retinal optical coherence tomography (OCT) images of 7912 patients with age-related macular degeneration (AMD), we evaluate the utility of using metadata to incorporate the temporal dynamics of disease progression into pretraining. Our metadata-enhanced approach outperforms both standard contrastive methods and a retinal image foundation model in five out of six image-level downstream tasks related to AMD. We find benefits in both a low-data and high-data regime across tasks ranging from AMD stage and type classification to prediction of visual acuity. Due to its modularity, our method can be quickly and cost-effectively tested to establish the potential benefits of including available metadata in contrastive pretraining.

Lamina Cribrosa Microstructure in Nonhuman Primates With Naturally Occurring Peripapillary Retinal Nerve Fiber Layer Thinning

Purpose

The lamina cribrosa (LC) is hypothesized to be the site of initial axonal damage in glaucoma with the circumpapillary retinal nerve fiber layer thickness (RNFL-T) widely used as a standard metric for quantifying the glaucomatous damage. The purpose of this study was to determine in vivo, 3-dimensional (3D) differences in the microstructure of the LC in eyes of nonhuman primates (NHPs) with naturally occurring glaucoma.

Methods

Spectral-domain optical coherence tomography (OCT) scans (Leica, Chicago, IL, USA) of the optic nerve head were acquired from a colony of 50 adult rhesus monkeys suspected of having high prevalence of glaucoma. The RNFL-T was analyzed globally and in quadrants using a semi-automated segmentation software. From a set of 100 eyes, 18 eyes with the thinnest global RNFL-T were selected as the study group and 18 eyes with RNFL-T values around the 50th percentile were used as controls. A previously described automated segmentation algorithm was used for LC microstructure analysis. Parameters included beam thickness, pore diameter and their ratio (beam-to-pore ratio [BPR]), pore area and shape parameters, beam and pore volume, and connective tissue volume fraction (CTVF; beam volume/total volume). The LC microstructure was analyzed globally and in the following volumetric sectors: quadrants, central and peripheral lamina, and three depth slabs (anterior, middle, and posterior).

Results

Although no significant difference was detected between groups for age, weight, or disc size, the study group had significantly thinner RNFL than the control group (P < 0.01). The study group had significantly smaller global and sectoral pore diameter and larger BPR compared with the control group. Across eyes, the global RNFL-T was associated positively with pore diameter globally. BPR and CTVF were significantly and negatively associated with the corresponding RNFL-T in the superior quadrant.

Conclusions

Global and sectoral microstructural differences were detected when comparing thin and normal RNFL-T eyes. Whether these LC differences are the cause of RNFL damage or the result of remodeling of the LC requires further investigation.

Translational Relevance

Our findings indicate structural alterations in the LC of NHP exhibiting natural thinning of the RNFL, a common characteristic of glaucomatous damage.

Morph-SSL: Self-Supervision With Longitudinal Morphing for Forecasting AMD Progression From OCT Volumes

The lack of reliable biomarkers makes predicting the conversion from intermediate to neovascular age-related macular degeneration (iAMD, nAMD) a challenging task. We develop a Deep Learning (DL) model to predict the future risk of conversion of an eye from iAMD to nAMD from its current OCT scan. Although eye clinics generate vast amounts of longitudinal OCT scans to monitor AMD progression, only a small subset can be manually labeled for supervised DL. To address this issue, we propose Morph-SSL, a novel Self-supervised Learning (SSL) method for longitudinal data. It uses pairs of unlabelled OCT scans from different visits and involves morphing the scan from the previous visit to the next. The Decoder predicts the transformation for morphing and ensures a smooth feature manifold that can generate intermediate scans between visits through linear interpolation. Next, the Morph-SSL trained features are input to a Classifier which is trained in a supervised manner to model the cumulative probability distribution of the time to conversion with a sigmoidal function. Morph-SSL was trained on unlabelled scans of 399 eyes (3570 visits). The Classifier was evaluated with a five-fold cross-validation on 2418 scans from 343 eyes with clinical labels of the conversion date. The Morph-SSL features achieved an AUC of 0.779 in predicting the conversion to nAMD within the next 6 months, outperforming the same network when trained end-to-end from scratch or pre-trained with popular SSL methods. Automated prediction of the future risk of nAMD onset can enable timely treatment and individualized AMD management.

Pretraining of 3D image segmentation models for retinal OCT using denoising-based self-supervised learning

Deep learning algorithms have allowed the automation of segmentation for many biomarkers in retinal OCTs, enabling comprehensive clinical research and precise patient monitoring. These segmentation algorithms predominantly rely on supervised training and specialised segmentation networks, such as U-Nets. However, they require segmentation annotations, which are challenging to collect and require specialized expertise. In this paper, we explore leveraging 3D self-supervised learning based on image restoration techniques, that allow to pretrain 3D networks with the aim of improving segmentation performance. We test two methods, based on image restoration and denoising. After pretraining on a large 3D OCT dataset, we evaluate our weights by fine-tuning them on two challenging fluid segmentation datasets utilising different amount of training data. The chosen methods are easy to set up while providing large improvements for fluid segmentation, enabling the reduction of the amount of required annotation or an increase in the performance. Overall, the best results were obtained for denoising-based SSL methods, with higher results on both fluid segmentation datasets as well as faster pretraining durations.

Long-term impact of low-energy femtosecond laser and manual cataract surgery on macular layer thickness: A prospective randomized study

PURPOSE

To evaluate change in retinal layers 18 months after femtosecond laser-assisted cataract surgery (LCS) and manual cataract surgery (MCS) in a representative age-related cataract population using artificial intelligence (AI)-based automated retinal layer segmentation.

METHODS

This was a prospective, randomized and intraindividual-controlled study including 60 patients at the Medical University of Vienna, Austria. Bilateral same-day LCS and MCS were performed in a randomized sequence. To provide insight into the development of cystoid macular oedema (CME), retinal layer thickness was measured pre-operatively and up to 18 months post-operatively in the central 1 mm, 3 mm and 6 mm.

RESULTS

Fifty-six patients completed all follow-up visits. LCS compared to MCS did not impact any of the investigated retinal layers at any follow-up visit (p > 0.05). For the central 1 mm, a significant increase in total retinal thickness (TRT) was seen after 1 week followed by an elevated plateau thereafter. For the 3 mm and 6 mm, TRT increased only after 3 weeks and 6 weeks and decreased again until 18 months. TRT remained significantly increased compared to pre-operative thickness (p < 0.001). Visual acuity remained unaffected by the macular thickening and no case of CME was observed. Inner nuclear layer (INL) and outer nuclear layer (ONL) were the main causative layers for the total TRT increase. Photoreceptors (PR) declined 1 week after surgery but regained pre-operative values 18 months after surgery.

CONCLUSION

Low-energy femtosecond laser pre-treatment did not influence thickness of the retinal layers in any topographic zone compared to manual high fluidic phacoemulsification. TRT did not return to pre-operative values 18 months after surgery. The causative layers for subclinical development of CME were successfully identified.

An Automated Comparative Analysis of the Exudative Biomarkers in Neovascular Age-Related Macular Degeneration, The RAP Study: Report 6

PURPOSE

To investigate differences in volume and distribution of the main exudative biomarkers across all types and subtypes of macular neovascularization (MNV) using artificial intelligence (AI).

DESIGN

Cross-sectional study.

METHODS

An AI-based analysis was conducted on 34,528 OCT B-scans consisting of 281 (250 unifocal, 31 multifocal) MNV3, 55 MNV2, and 121 (30 polypoidal, 91 non-polypoidal) MNV1 treatment-naive eyes. Means (SDs), medians and heat maps of cystic intraretinal fluid (IRF), subretinal fluid (SRF), pigment epithelial detachments (PED), and hyperreflective foci (HRF) volumes, as well as retinal thickness (RT) were compared among MNV types and subtypes.

RESULTS

MNV3 had the highest mean IRF with 291 (290) nL, RT with 357 (49) µm, and HRF with 80 (70) nL, P ≤ .05. MNV1 showed the greatest mean SRF with 492 (586) nL, whereas MNV3 exhibited the lowest with 218 (382) nL, P ≤ .05. Heat maps showed IRF confined to the center, whereas SRF was scattered in all types. SRF, HRF, and PED were more distributed in the temporal macular half in MNV3. Means of IRF, HRF, and PED were higher in the multifocal than in the unifocal MNV3 with 416 (309) nL,114 (95) nL, and 810 (850) nL, P ≤ .05. Compared to the non-polypoidal subtype, the polypoidal subtype had greater means of SRF with 695 (718) nL, HRF 69 (63) nL, RT 357 (45) µm, and PED 1115 (1170) nL, P ≤ .05.

CONCLUSIONS

This novel quantitative AI analysis shows that SRF is a biomarker of choroidal origin in MNV1, whereas IRF, HRF, and RT are retinal biomarkers in MNV3. Polypoidal MNV1 and multifocal MNV3 present with higher exudation compared to other subtypes.

Exploring autism via the retina: Comparative insights in children with autism spectrum disorder and typical development

Autism spectrum disorder (ASD) is a widely recognized neurodevelopmental disorder, yet the identification of reliable imaging biomarkers for its early diagnosis remains a challenge. Considering the specific manifestations of ASD in the eyes and the interconnectivity between the brain and the eyes, this study investigates ASD through the lens of retinal analysis. We specifically examined differences in the macular region of the retina using optical coherence tomography (OCT)/optical coherence tomography angiography (OCTA) images between children diagnosed with ASD and those with typical development (TD). Our findings present potential novel characteristics of ASD: the thickness of the ellipsoid zone (EZ) with cone photoreceptors was significantly increased in ASD; the large-caliber arteriovenous of the inner retina was significantly reduced in ASD; these changes in the EZ and arteriovenous were more significant in the left eye than in the right eye. These observations of photoreceptor alterations, vascular function changes, and lateralization phenomena in ASD warrant further investigation, and we hope that this work can advance interdisciplinary understanding of ASD.

Transfer Learning-Based Approach for Thickness Estimation on Optical Coherence Tomography of Varicose Veins

In-depth mechanical characterization of veins is required for promising innovations of venous substitutes and for better understanding of venous diseases. Two important physical parameters of veins are shape and thickness, which are quite challenging in soft tissues. Here, we propose the method TREE (TransfeR learning-based approach for thicknEss Estimation) to predict both the segmentation map and thickness value of the veins. This model incorporates one encoder and two decoders which are trained in a special manner to facilitate transfer learning. First, an encoder-decoder pair is trained to predict segmentation maps, then this pre-trained encoder with frozen weights is paired with a second decoder that is specifically trained to predict thickness maps. This leverages the global information gained from the segmentation model to facilitate the precise learning of the thickness model. Additionally, to improve the performance we introduce a sensitive pattern detector (SPD) module which further guides the network by extracting semantic details. The swept-source optical coherence tomography (SS-OCT) is the imaging modality for saphenous varicose vein extracted from the diseased patients. To demonstrate the performance of the model, we calculated the segmentation accuracy-0.993, mean square error in thickness (pixels) estimation-2.409 and both these metrics stand out when compared with the state-of-art methods.

Retinal Sublayer Analysis in Autoimmune Retinopathy and Identification of New Optical Coherence Tomography Phenotypes

PURPOSE

Autoimmune retinopathy (AIR) is a poorly characterized disease with a wide phenotypic spectrum, complicating investigations of its underlying pathophysiology. We sought to analyze optical coherence tomography (OCT) retinal thickness changes in AIR patients.

METHODS

A retrospective chart review from 2007 to 2017 was performed evaluating AIR patients at a single academic, tertiary referral center. OCT retinal sublayer analysis was performed, and paradoxical thickening phenotypes were reviewed.

RESULTS

Twenty-nine AIR patients with positive anti-retinal antibodies and OCT imaging were identified. Overall, AIR patients had thinner retinal sublayers compared to controls; however, 12 patients (41.4%) had paradoxical thickening of the outer plexiform layer (OPL). This revealed two distinct OCT phenotypes. No association was found between retinal sublayer thickness and specific antiretinal antibodies.

CONCLUSIONS

While the pathogenicity of antiretinal antibodies remains unclear, the OCT phenotypes observed underscore the potential for identifying clues in the underlying disease processes and clinical diagnosis.

Assessment of OCT-Based Macular Curvature and Its Relationship with Macular Microvasculature in Children with Anisomyopia

INTRODUCTION

To evaluate the intraocular differences in optical coherence tomography (OCT)-based macular curvature index (MCI) among children with anisomyopia and to investigate the relationship between MCI and the macular microvasculature.

METHODS

Fifty-two schoolchildren with anisometropia > 2.00 D were enrolled and underwent comprehensive examinations including cycloplegic refraction, axial length (AL), and swept source OCT/OCT angiography. OCT-based MCIs were determined from horizontal and vertical B-scans by a customized curve fitting model in MATLAB R2022 at 1-mm-, 3-mm-, and 6-mm-diameter circles at fovea. Characteristics and topographic variation of MCI was analyzed, and the relationships with microvascularity and its associated factors were investigated.

RESULTS

MCI achieved high reliability and repeatability. There were overall larger MCIs in the more myopic eyes than the less myopic eyes in 1-mm-, 3-mm-, and 6-mm-diameter circles at fovea (all p < 0.001). For the topographic variation, horizontal MCI was significantly greater than vertical MCI (all p < 0.001), and was the largest in 6-mm circle, followed by 3-mm and 1-mm circles. Stronger correlation of horizontal MCI with myopic severity than vertical MCI was found. Partial Pearson's correlation found MCI was negatively associated with deep capillary plexus (DCP) vessel density (p = 0.016). Eyes with a higher MCI in a 6-mm circle were more likely to have longer AL (p < 0.001), lower DCP vessel density (p = 0.037), and thinner choroidal thickness (ChT) (p = 0.045).

CONCLUSION

Larger MCI was found in the more myopic eyes of children with anisomyopia and was significantly associated with smaller DCP density, suggesting that MCI was an important indicator of myopia-related retinal microvascularity change, and it could be a valuable metric for myopia assessment in children.

DIRECTIONAL OPTICAL COHERENCE TOMOGRAPHY IMAGING OF MACULAR PATHOLOGY

PURPOSE

To survey the impact of directional reflectivity on structures within optical coherence tomography images in retinal pathology.

METHODS

Sets of commercial optical coherence tomography images taken from multiple pupil positions were analyzed. These directional optical coherence tomography sets revealed directionally reflective structures within the retina. After ensuring sufficient image quality, resulting hybrid and composite images were characterized by assessing the Henle fiber layer, outer nuclear layer, ellipsoid zone, and interdigitation zone. Additionally, hybrid images were reviewed for novel directionally reflective pathological features.

RESULTS

Cross-sectional directional optical coherence tomography image sets were obtained in 75 eyes of 58 patients having a broad range of retinal pathologies. All cases showed improved visualization of the outer nuclear layer/Henle fiber layer interface, and outer nuclear layer thinning was, therefore, more apparent in several cases. The ellipsoid zone and interdigitation zone also demonstrated attenuation where a geometric impact of underlying pathology affected their orientation. Misdirected photoreceptors were also noted as a consistent direction-dependent change in ellipsoid zone reflectivity between regions of normal and absent ellipsoid zone.

CONCLUSION

Directional optical coherence tomography enhances the understanding of retinal anatomy and pathology. This optical contrast yields more accurate identification of retinal structures and possible imaging biomarkers for photoreceptor-related pathology.

In vivo endoscopic optical coherence elastography based on a miniature probe

Optical coherence elastography (OCE) is a functional extension of optical coherence tomography (OCT). It offers high-resolution elasticity assessment with nanoscale tissue displacement sensitivity and high quantification accuracy, promising to enhance diagnostic precision. However, in vivo endoscopic OCE imaging has not been demonstrated yet, which needs to overcome key challenges related to probe miniaturization, high excitation efficiency and speed. This study presents a novel endoscopic OCE system, achieving the first endoscopic OCE imaging in vivo. The system features the smallest integrated OCE probe with an outer diameter of only 0.9 mm (with a 1.2-mm protective tube during imaging). Utilizing a single 38-MHz high-frequency ultrasound transducer, the system induced rapid deformation in tissues with enhanced excitation efficiency. In phantom studies, the OCE quantification results match well with compression testing results, showing the system's high accuracy. The in vivo imaging of the rat vagina demonstrated the system's capability to detect changes in tissue elasticity continually and distinguish between normal tissue, hematomas, and tissue with increased collagen fibers precisely. This research narrows the gap for the clinical implementation of the endoscopic OCE system, offering the potential for the early diagnosis of intraluminal diseases.

Hybrid deep learning and optimal graph search method for optical coherence tomography layer segmentation in diseases affecting the optic nerve

Accurate segmentation of retinal layers in optical coherence tomography (OCT) images is critical for assessing diseases that affect the optic nerve, but existing automated algorithms often fail when pathology causes irregular layer topology, such as extreme thinning of the ganglion cell-inner plexiform layer (GCIPL). Deep LOGISMOS, a hybrid approach that combines the strengths of deep learning and 3D graph search to overcome their limitations, was developed to improve the accuracy, robustness and generalizability of retinal layer segmentation. The method was trained on 124 OCT volumes from both eyes of 31 non-arteritic anterior ischemic optic neuropathy (NAION) patients and tested on three cross-sectional datasets with available reference tracings: Test-NAION (40 volumes from both eyes of 20 NAION subjects), Test-G (29 volumes from 29 glaucoma subjects/eyes), and Test-JHU (35 volumes from 21 multiple sclerosis and 14 control subjects/eyes) and one longitudinal dataset without reference tracings: Test-G-L (155 volumes from 15 glaucoma patients/eyes). In the three test datasets with reference tracings (Test-NAION, Test-G, and Test-JHU), Deep LOGISMOS achieved very high Dice similarity coefficients (%) on GCIPL: 89.97±3.59, 90.63±2.56, and 94.06±1.76, respectively. In the same context, Deep LOGISMOS outperformed the Iowa reference algorithms by improving the Dice score by 17.5, 5.4, and 7.5, and also surpassed the deep learning framework nnU-Net with improvements of 4.4, 3.7, and 1.0. For the 15 severe glaucoma eyes with marked GCIPL thinning (Test-G-L), it demonstrated reliable regional GCIPL thickness measurement over five years. The proposed Deep LOGISMOS approach has potential to enhance precise quantification of retinal structures, aiding diagnosis and treatment management of optic nerve diseases.

Generalized 3D registration algorithm for enhancing retinal optical coherence tomography images

Significance

Optical coherence tomography (OCT) has emerged as the standard of care for diagnosing and monitoring the treatment of various ocular disorders due to its noninvasive nature and in vivo volumetric acquisition capability. Despite its widespread applications in ophthalmology, motion artifacts remain a challenge in OCT imaging, adversely impacting image quality. While several multivolume registration algorithms have been developed to address this issue, they are often designed to cater to one specific OCT system or acquisition protocol.

Aim

We aim to generate an OCT volume free of motion artifacts using a system-agnostic registration algorithm that is independent of system specifications or protocol.

Approach

We developed a B-scan registration algorithm that removes motion and corrects for both translational eye movements and rotational angle differences between volumes. Tests were carried out on various datasets obtained from two different types of custom-built OCT systems and one commercially available system to determine the reliability of the proposed algorithm. Additionally, different system specifications were used, with variations in axial resolution, lateral resolution, signal-to-noise ratio, and real-time motion tracking. The accuracy of this method has further been evaluated through mean squared error (MSE) and multiscale structural similarity index measure (MS-SSIM).

Results

The results demonstrate improvements in the overall contrast of the images, facilitating detailed visualization of retinal vasculatures in both superficial and deep vasculature plexus. Finer features of the inner and outer retina, such as photoreceptors and other pathology-specific features, are discernible after multivolume registration and averaging. Quantitative analyses affirm that increasing the number of averaged registered volumes will decrease MSE and increase MS-SSIM as compared to the reference volume.

Conclusions

The multivolume registered data obtained from this algorithm offers significantly improved visualization of the retinal microvascular network as well as retinal morphological features. Furthermore, we have validated that the versatility of our methodology extends beyond specific OCT modalities, thereby enhancing the clinical utility of OCT for the diagnosis and monitoring of ocular pathologies.

Anomaly guided segmentation: Introducing semantic context for lesion segmentation in retinal OCT using weak context supervision from anomaly detection

Automated lesion detection in retinal optical coherence tomography (OCT) scans has shown promise for several clinical applications, including diagnosis, monitoring and guidance of treatment decisions. However, segmentation models still struggle to achieve the desired results for some complex lesions or datasets that commonly occur in real-world, e.g. due to variability of lesion phenotypes, image quality or disease appearance. While several techniques have been proposed to improve them, one line of research that has not yet been investigated is the incorporation of additional semantic context through the application of anomaly detection models. In this study we experimentally show that incorporating weak anomaly labels to standard segmentation models consistently improves lesion segmentation results. This can be done relatively easy by detecting anomalies with a separate model and then adding these output masks as an extra class for training the segmentation model. This provides additional semantic context without requiring extra manual labels. We empirically validated this strategy using two in-house and two publicly available retinal OCT datasets for multiple lesion targets, demonstrating the potential of this generic anomaly guided segmentation approach to be used as an extra tool for improving lesion detection models.

OCTA-500: A retinal dataset for optical coherence tomography angiography study

Optical coherence tomography angiography (OCTA) is a novel imaging modality that has been widely utilized in ophthalmology and neuroscience studies to observe retinal vessels and microvascular systems. However, publicly available OCTA datasets remain scarce. In this paper, we introduce the largest and most comprehensive OCTA dataset dubbed OCTA-500, which contains OCTA imaging under two fields of view (FOVs) from 500 subjects. The dataset provides rich images and annotations including two modalities (OCT/OCTA volumes), six types of projections, four types of text labels (age/gender/eye/disease) and seven types of segmentation labels (large vessel/capillary/artery/vein/2D FAZ/3D FAZ/retinal layers). Then, we propose a multi-object segmentation task called CAVF, which integrates capillary segmentation, artery segmentation, vein segmentation, and FAZ segmentation under a unified framework. In addition, we optimize the 3D-to-2D image projection network (IPN) to IPN-V2 to serve as one of the segmentation baselines. Experimental results demonstrate that IPN-V2 achieves an about 10% mIoU improvement over IPN on CAVF task. Finally, we further study the impact of several dataset characteristics: the training set size, the model input (OCT/OCTA, 3D volume/2D projection), the baseline networks, and the diseases. The dataset and code are publicly available at: https://ieee-dataport.org/open-access/octa-500.

Advancing Ocular Imaging: A Hybrid Attention Mechanism-Based U-Net Model for Precise Segmentation of Sub-Retinal Layers in OCT Images

This paper presents a novel U-Net model incorporating a hybrid attention mechanism for automating the segmentation of sub-retinal layers in Optical Coherence Tomography (OCT) images. OCT is an ophthalmology tool that provides detailed insights into retinal structures. Manual segmentation of these layers is time-consuming and subjective, calling for automated solutions. Our proposed model combines edge and spatial attention mechanisms with the U-Net architecture to improve segmentation accuracy. By leveraging attention mechanisms, the U-Net focuses selectively on image features. Extensive evaluations using datasets demonstrate that our model outperforms existing approaches, making it a valuable tool for medical professionals. The study also highlights the model's robustness through performance metrics such as an average Dice score of 94.99%, Adjusted Rand Index (ARI) of 97.00%, and Strength of Agreement (SOA) classifications like "Almost Perfect", "Excellent", and "Very Strong". This advanced predictive model shows promise in expediting processes and enhancing the precision of ocular imaging in real-world applications.

Deep Learning-Based Retinal Layer Segmentation in Optical Coherence Tomography Scans of Patients with Inherited Retinal Diseases

BACKGROUND

In optical coherence tomography (OCT) scans of patients with inherited retinal diseases (IRDs), the measurement of the thickness of the outer nuclear layer (ONL) has been well established as a surrogate marker for photoreceptor preservation. Current automatic segmentation tools fail in OCT segmentation in IRDs, and manual segmentation is time-consuming.

METHODS AND MATERIAL

Patients with IRD and an available OCT scan were screened for the present study. Additionally, OCT scans of patients without retinal disease were included to provide training data for artificial intelligence (AI). We trained a U-net-based model on healthy patients and applied a domain adaption technique to the IRD patients' scans.

RESULTS

We established an AI-based image segmentation algorithm that reliably segments the ONL in OCT scans of IRD patients. In a test dataset, the dice score of the algorithm was 98.7%. Furthermore, we generated thickness maps of the full retinal thickness and the ONL layer for each patient.

CONCLUSION

Accurate segmentation of anatomical layers on OCT scans plays a crucial role for predictive models linking retinal structure to visual function. Our algorithm for segmentation of OCT images could provide the basis for further studies on IRDs.

Radiation effects on retinal layers revealed by OCT, OCT-A, and perimetry as a function of dose and time from treatment

Optical coherence tomography (OCT) has become a key method for diagnosing and staging radiation retinopathy, based mainly on the presence of fluid in the central macula. A robust retinal layer segmentation method is required for identification of the specific layers involved in radiation-induced pathology in individual eyes over time, in order to determine damage driven by radiation injury to the microvessels and to the inner retinal neurons. Here, we utilized OCT, OCT-angiography, visual field testing, and patient-specific dosimetry models to analyze abnormal retinal layer thickening and thinning relative to microvessel density, visual function, radiation dose, and time from radiotherapy in a cross-sectional cohort of uveal melanoma patients treated with 125I-plaque brachytherapy. Within the first 24 months of radiotherapy, we show differential thickening and thinning of the two inner retinal layers, suggestive of microvessel leakage and neurodegeneration, mostly favoring thickening. Four out of 13 eyes showed decreased inner retinal capillary density associated with a corresponding normal inner retinal thickness, indicating early microvascular pathology. Two eyes showed the opposite: significant inner retinal layer thinning and normal capillary density, indicating early neuronal damage preceding a decrease in capillary density. At later time points, inner retinal thinning becomes the dominant pathology and correlates significantly with decreased vascularity, vision loss, and dose to the optic nerve. Stable multiple retinal layer segmentation provided by 3D graph-based methods aids in assessing the microvascular and neuronal response to radiation, information needed to target therapeutics for radiation retinopathy and vision loss.

Visualization of Optic Nerve Structural Patterns in Papilledema Using Deep Learning Variational Autoencoders

Purpose

To visualize and quantify structural patterns of optic nerve edema encountered in papilledema during treatment.

Methods

A novel bi-channel deep-learning variational autoencoder (biVAE) model was trained using 1498 optical coherence tomography (OCT) scans of 125 subjects over time from the Idiopathic Intracranial Hypertension Treatment Trial (IIHTT) and 791 OCT scans of 96 control subjects from the University of Iowa. An independent test dataset of 70 eyes from 70 papilledema subjects was used to evaluate the ability of the biVAE model to quantify and reconstruct the papilledema spatial patterns from input OCT scans using only two variables.

Results

The montage color maps of the retinal nerve fiber layer (RNFL) and total retinal thickness (TRT) produced by the biVAE model provided an organized visualization of the variety of morphological patterns of optic disc edema (including differing patterns at similar thickness levels). Treatment effects of acetazolamide versus placebo in the IIHTT were also demonstrated in the latent space. In image reconstruction, the mean signed peripapillary retinal nerve fiber layer thickness (pRNFLT) difference ± SD was -0.12 ± 17.34 µm, the absolute pRNFLT difference was 13.68 ± 10.65 µm, and the RNFL structural similarity index reached 0.91 ± 0.05.

Conclusions

A wide array of structural patterns of papilledema, integrating the magnitude of disc edema with underlying disc and retinal morphology, can be quantified by just two latent variables.

Translational Relevance

A biVAE model encodes structural patterns, as well as the correlation between channels, and may be applied to visualize individuals or populations with papilledema throughout treatment.

A Technique to Enable Efficient Adaptive Radiation Therapy: Automated Contouring of Prostate and Adjacent Organs

Purpose

The purpose of this work was to investigate the use of a segmentation approach that could potentially improve the speed and reproducibility of contouring during magnetic resonance-guided adaptive radiation therapy.

Methods and Materials

The segmentation algorithm was based on a hybrid deep neural network and graph optimization approach that also allows rapid user intervention (Deep layered optimal graph image segmentation of multiple objects and surfaces [LOGISMOS] + just enough interaction [JEI]). A total of 115 magnetic resonance-data sets were used for training and quantitative assessment. Expert segmentations were used as the independent standard for the prostate, seminal vesicles, bladder, rectum, and femoral heads for all 115 data sets. In addition, 3 independent radiation oncologists contoured the prostate, seminal vesicles, and rectum for a subset of patients such that the interobserver variability could be quantified. Consensus contours were then generated from these independent contours using a simultaneous truth and performance level estimation approach, and the deviation of Deep LOGISMOS + JEI contours to the consensus contours was evaluated and compared with the interobserver variability.

Results

The absolute accuracy of Deep LOGISMOS + JEI generated contours was evaluated using median absolute surface-to-surface distance which ranged from a minimum of 0.20 mm for the bladder to a maximum of 0.93 mm for the prostate compared with the independent standard across all data sets. The median relative surface-to-surface distance was less than 0.17 mm for all organs, indicating that the Deep LOGISMOS + JEI algorithm did not exhibit a systematic under- or oversegmentation. Interobserver variability testing yielded a mean absolute surface-to-surface distance of 0.93, 1.04, and 0.81 mm for the prostate, seminal vesicles, and rectum, respectively. In comparison, the deviation of Deep LOGISMOS + JEI from consensus simultaneous truth and performance level estimation contours was 0.57, 0.64, and 0.55 mm for the same organs. On average, the Deep LOGISMOS algorithm took less than 26 seconds for contour segmentation.

Conclusions

Deep LOGISMOS + JEI segmentation efficiently generated clinically acceptable prostate and normal tissue contours, potentially limiting the need for time intensive manual contouring with each fraction.

Simultaneous alignment and surface regression using hybrid 2D-3D networks for 3D coherent layer segmentation of retinal OCT images with full and sparse annotations

Layer segmentation is important to quantitative analysis of retinal optical coherence tomography (OCT). Recently, deep learning based methods have been developed to automate this task and yield remarkable performance. However, due to the large spatial gap and potential mismatch between the B-scans of an OCT volume, all of them were based on 2D segmentation of individual B-scans, which may lose the continuity and diagnostic information of the retinal layers in 3D space. Besides, most of these methods required dense annotation of the OCT volumes, which is labor-intensive and expertise-demanding. This work presents a novel framework based on hybrid 2D-3D convolutional neural networks (CNNs) to obtain continuous 3D retinal layer surfaces from OCT volumes, which works well with both full and sparse annotations. The 2D features of individual B-scans are extracted by an encoder consisting of 2D convolutions. These 2D features are then used to produce the alignment displacement vectors and layer segmentation by two 3D decoders coupled via a spatial transformer module. Two losses are proposed to utilize the retinal layers' natural property of being smooth for B-scan alignment and layer segmentation, respectively, and are the key to the semi-supervised learning with sparse annotation. The entire framework is trained end-to-end. To the best of our knowledge, this is the first work that attempts 3D retinal layer segmentation in volumetric OCT images based on CNNs. Experiments on a synthetic dataset and three public clinical datasets show that our framework can effectively align the B-scans for potential motion correction, and achieves superior performance to state-of-the-art 2D deep learning methods in terms of both layer segmentation accuracy and cross-B-scan 3D continuity in both fully and semi-supervised settings, thus offering more clinical values than previous works.

Neuro-ophthalmic features of patients with spontaneous cerebrospinal fluid leaks

Background

Increased intracranial pressure is a potential cause of spontaneous cerebrospinal fluid (sCSF) leak. Associated neuro-ophthalmic features have not been well studied, particularly relationships with idiopathic intracranial hypertension (IIH). We hypothesized that neuro-ophthalmic features routinely used in evaluations for IIH can be useful in the investigation of a causal relationship between IIH and sCSF leak. We reviewed the neuro-ophthalmic examination and office-based ophthalmic imaging data of all consecutive patients with sCSF leaks and at least one repair to investigate the clinical and neuro-ophthalmic features of increased intracranial pressure.

Methods

We conducted a retrospective longitudinal study at a single institution by querying the electronic medical record system for CSF leak Current Procedural Terminology (CPT) codes (G96.00 and G96.01) from June 1, 2019, to July 31, 2022. For patients with a confirmed diagnosis of sCSF leak, demographic information, eye examination results, and ophthalmic imaging details for both eyes were collected.

Results

A total of 189 patients with CSF leaks were identified through CPT coding; 159 had iatrogenic or traumatic CSF leaks, and 30 individuals (3 male, 27 female) had confirmed sCSF leaks. The mean age of patients with sCSF leaks was 46 years (range: 29 - 81), with a mean body mass index of 35.2 kg/m2 (range: 18.2 - 54.1). Only 11 of 30 underwent eye examinations (8 before surgical repair and 10 after). The mean pre-repair and post-repair best-corrected visual acuity were 20/30 (range: 20/20 - 20/55) and 20/25 (range: 20/20 - 20/40), respectively (P = 0.188). The mean retinal nerve fiber layer thickness was 99 µm (range: 96 - 104) pre-repair and 97 µm (range: 84 - 103) post-repair (P = 0.195). The mean ganglion cell complex thickness was 84 µm (range: 72 - 94) pre-repair and 82 µm (range: 71 - 94) post-repair (P = 0.500). Humphrey visual field average mean deviation was -5.1 (range: -12.4 - -1.8) pre-repair and -1.0 (range: -10.1 - 2.1) post-repair (P = 0.063).

Conclusions

Serial neuro-ophthalmic examinations are recommended for patients with sCSF leaks to screen for signs of current or prior increased intracranial pressure. Larger studies are required to clarify the longitudinal changes in neuro-ophthalmic features, to investigate the incidence of IIH in cases of sCSF leak development or recurrence after surgical repair, and to explore potential causal relationships to guide post-repair management and prevent recurrent leaks. A multicenter consortium is also suggested to develop a standard clinical protocol for comprehensive management of sCSF leaks.

When Sex Matters: Differences in the Central Nervous System as Imaged by OCT through the Retina

BACKGROUND

Retinal texture has gained momentum as a source of biomarkers of neurodegeneration, as it is sensitive to subtle differences in the central nervous system from texture analysis of the neuroretina. Sex differences in the retina structure, as detected by layer thickness measurements from optical coherence tomography (OCT) data, have been discussed in the literature. However, the effect of sex on retinal interocular differences in healthy adults has been overlooked and remains largely unreported.

METHODS

We computed mean value fundus images for the neuroretina layers as imaged by OCT of healthy individuals. Texture metrics were obtained from these images to assess whether women and men have the same retina texture characteristics in both eyes. Texture features were tested for group mean differences between the right and left eye.

RESULTS

Corrected texture differences exist only in the female group.

CONCLUSIONS

This work illustrates that the differences between the right and left eyes manifest differently in females and males. This further supports the need for tight control and minute analysis in studies where interocular asymmetry may be used as a disease biomarker, and the potential of texture analysis applied to OCT imaging to spot differences in the retina.

Point-to-point associations of drusen and hyperreflective foci volumes with retinal sensitivity in non-exudative age-related macular degeneration

OBJECTIVES

To evaluate the quantitative impact of drusen and hyperreflective foci (HRF) volumes on mesopic retinal sensitivity in non-exudative age-related macular degeneration (AMD).

METHODS

In a standardized follow-up scheme of every three months, retinal sensitivity of patients with early or intermediate AMD was assessed by microperimetry using a custom pattern of 45 stimuli (Nidek MP-3, Gamagori, Japan). Eyes were consecutively scanned using Spectralis SD-OCT (20° × 20°, 1024 × 97 × 496). Fundus photographs obtained by the MP-3 allowed to map the stimuli locations onto the corresponding OCT scans. The volume and mean thickness of drusen and HRF within a circle of 240 µm centred at each stimulus point was determined using automated AI-based image segmentation algorithms.

RESULTS

8055 individual stimuli from 179 visits from 51 eyes of 35 consecutive patients were matched with the respective OCT images in a point-to-point manner. The patients mean age was 76.85 ± 6.6 years. Mean retinal sensitivity at baseline was 25.7 dB. 73.47% of all MP-spots covered drusen area and 2.02% of MP-spots covered HRF. A negative association between retinal sensitivity and the volume of underlying drusen (p < 0.001, Estimate -0.991 db/µm3) and HRF volume (p = 0.002, Estimate -5.230 db/µm3) was found. During observation time, no eye showed conversion to advanced AMD.

CONCLUSION

A direct correlation between drusen and lower sensitivity of the overlying photoreceptors can be observed. For HRF, a small but significant correlation was shown, which is compromised by their small size. Biomarker quantification using AI-methods allows to determine the impact of sub-clinical features in the progression of AMD.

Retinotopic degeneration of the retina and optic tracts in autosomal dominant Alzheimer's disease

INTRODUCTION

We investigated the correlation between retinal thickness and optic tract integrity in subjects with autosomal dominant Alzheimer's disease (ADAD) causing mutations.

METHODS

Retinal thicknesses and diffusion tensor images (DTI) were obtained using optical coherence tomography and magnetic resonance imaging, respectively. The association between retinal thickness and DTI measures was adjusted for age, sex, retinotopy, and correlation between eyes.

RESULTS

Optic tract mean diffusivity and axial diffusivity were negatively correlated with retinotopically defined ganglion cell inner plexiform thickness (GCIPL). Fractional anisotropy was negatively correlated with retinotopically defined retinal nerve fiber layer thickness. There was no correlation between outer nuclear layer (ONL) thickness and any DTI measure.

DISCUSSION

In ADAD, GCIPL thickness is significantly associated with retinotopic optic tract DTI measures even in minimally symptomatic subjects. Similar associations were not present with ONL thickness or when ignoring retinotopy. We provide in vivo evidence for optic tract changes resulting from ganglion cell pathology in ADAD.

Graph Attention U-Net for Retinal Layer Surface Detection and Choroid Neovascularization Segmentation in OCT Images

Choroidal neovascularization (CNV) is a typical symptom of age-related macular degeneration (AMD) and is one of the leading causes for blindness. Accurate segmentation of CNV and detection of retinal layers are critical for eye disease diagnosis and monitoring. In this paper, we propose a novel graph attention U-Net (GA-UNet) for retinal layer surface detection and CNV segmentation in optical coherence tomography (OCT) images. Due to retinal layer deformation caused by CNV, it is challenging for existing models to segment CNV and detect retinal layer surfaces with the correct topological order. We propose two novel modules to address the challenge. The first module is a graph attention encoder (GAE) in a U-Net model that automatically integrates topological and pathological knowledge of retinal layers into the U-Net structure to achieve effective feature embedding. The second module is a graph decorrelation module (GDM) that takes reconstructed features by the decoder of the U-Net as inputs, it then decorrelates and removes information unrelated to retinal layer for improved retinal layer surface detection. In addition, we propose a new loss function to maintain the correct topological order of retinal layers and the continuity of their boundaries. The proposed model learns graph attention maps automatically during training and performs retinal layer surface detection and CNV segmentation simultaneously with the attention maps during inference. We evaluated the proposed model on our private AMD dataset and another public dataset. Experiment results show that the proposed model outperformed the competing methods for retinal layer surface detection and CNV segmentation and achieved new state of the arts on the datasets.

Improving OCT Image Segmentation of Retinal Layers by Utilizing a Machine Learning Based Multistage System of Stacked Multiscale Encoders and Decoders

Optical coherence tomography (OCT)-based retinal imagery is often utilized to determine influential factors in patient progression and treatment, for which the retinal layers of the human eye are investigated to assess a patient's health status and eyesight. In this contribution, we propose a machine learning (ML)-based multistage system of stacked multiscale encoders and decoders for the image segmentation of OCT imagery of the retinal layers to enable the following evaluation regarding the physiological and pathological states. Our proposed system's results highlight its benefits compared to currently investigated approaches by combining commonly deployed methods from deep learning (DL) while utilizing deep neural networks (DNN). We conclude that by stacking multiple multiscale encoders and decoders, improved scores for the image segmentation task can be achieved. Our retinal-layer-based segmentation results in a final segmentation performance of up to 82.25±0.74% for the Sørensen-Dice coefficient, outperforming the current best single-stage model by 1.55% with a score of 80.70±0.20%, given the evaluated peripapillary OCT data set. Additionally, we provide results on the data sets Duke SD-OCT, Heidelberg, and UMN to illustrate our model's performance on especially noisy data sets.

Retinal OCT Layer Segmentation via Joint Motion Correction and Graph-Assisted 3D Neural Network

Optical Coherence Tomography (OCT) is a widely used 3D imaging technology in ophthalmology. Segmentation of retinal layers in OCT is important for diagnosis and evaluation of various retinal and systemic diseases. While 2D segmentation algorithms have been developed, they do not fully utilize contextual information and suffer from inconsistency in 3D. We propose neural networks to combine motion correction and segmentation in 3D. The proposed segmentation network utilizes 3D convolution and a novel graph pyramid structure with graph-inspired building blocks. We also collected one of the largest OCT segmentation dataset with manually corrected segmentation covering both normal examples and various diseases. The experimental results on three datasets with multiple instruments and various diseases show the proposed method can achieve improved segmentation accuracy compared with commercial softwares and conventional or deep learning methods in literature. Specifically, the proposed method reduced the average error from 38.47% to 11.43% compared to clinically available commercial software for severe deformations caused by diseases. The diagnosis and evaluation of diseases with large deformation such as DME, wet AMD and CRVO would greatly benefit from the improved accuracy, which impacts tens of millions of patients.

Automated inter-device 3D OCT image registration using deep learning and retinal layer segmentation

Optical coherence tomography (OCT) is the most widely used imaging modality in ophthalmology. There are multiple variations of OCT imaging capable of producing complementary information. Thus, registering these complementary volumes is desirable in order to combine their information. In this work, we propose a novel automated pipeline to register OCT images produced by different devices. This pipeline is based on two steps: a multi-modal 2D en-face registration based on deep learning, and a Z-axis (axial axis) registration based on the retinal layer segmentation. We evaluate our method using data from a Heidelberg Spectralis and an experimental PS-OCT device. The empirical results demonstrated high-quality registrations, with mean errors of approximately 46 µm for the 2D registration and 9.59 µm for the Z-axis registration. These registrations may help in multiple clinical applications such as the validation of layer segmentations among others.

Deep survival modeling of longitudinal retinal OCT volumes for predicting the onset of atrophy in patients with intermediate AMD

In patients with age-related macular degeneration (AMD), the risk of progression to late stages is highly heterogeneous, and the prognostic imaging biomarkers remain unclear. We propose a deep survival model to predict the progression towards the late atrophic stage of AMD. The model combines the advantages of survival modelling, accounting for time-to-event and censoring, and the advantages of deep learning, generating prediction from raw 3D OCT scans, without the need for extracting a predefined set of quantitative biomarkers. We demonstrate, in an extensive set of evaluations, based on two large longitudinal datasets with 231 eyes from 121 patients for internal evaluation, and 280 eyes from 140 patients for the external evaluation, that this model improves the risk estimation performance over standard deep learning classification models.

Characteristics of Henle's fiber layer in healthy and glaucoma eyes assessed by polarization-sensitive optical coherence tomography

Using conventional optical coherence tomography (OCT), it is difficult to image Henle fibers (HF) due to their low backscattering potential. However, fibrous structures exhibit form birefringence, which can be exploited to visualize the presence of HF by polarization-sensitive (PS) OCT. We found a slight asymmetry in the retardation pattern of HF in the fovea region that can be associated with the asymmetric decrease of cone density with eccentricity from the fovea. We introduce a new measure based on a PS-OCT assessment of optic axis orientation to estimate the presence of HF at various eccentricities from the fovea in a large cohort of 150 healthy subjects. By comparing a healthy age-matched sub-group (N = 87) to a cohort of 64 early-stage glaucoma patients, we found no significant difference in HF extension but a slightly decreased retardation at about 2° to 7.5° eccentricity from the fovea in the glaucoma patients. This potentially indicates that glaucoma affects this neuronal tissue at an early state.

Accurate drusen segmentation in optical coherence tomography via order-constrained regression of retinal layer heights

Drusen are an important biomarker for age-related macular degeneration (AMD). Their accurate segmentation based on optical coherence tomography (OCT) is therefore relevant to the detection, staging, and treatment of disease. Since manual OCT segmentation is resource-consuming and has low reproducibility, automatic techniques are required. In this work, we introduce a novel deep learning based architecture that directly predicts the position of layers in OCT and guarantees their correct order, achieving state-of-the-art results for retinal layer segmentation. In particular, the average absolute distance between our model's prediction and the ground truth layer segmentation in an AMD dataset is 0.63, 0.85, and 0.44 pixel for Bruch's membrane (BM), retinal pigment epithelium (RPE) and ellipsoid zone (EZ), respectively. Based on layer positions, we further quantify drusen load with excellent accuracy, achieving 0.994 and 0.988 Pearson correlation between drusen volumes estimated by our method and two human readers, and increasing the Dice score to 0.71 ± 0.16 (from 0.60 ± 0.23) and 0.62 ± 0.23 (from 0.53 ± 0.25), respectively, compared to a previous state-of-the-art method. Given its reproducible, accurate, and scalable results, our method can be used for the large-scale analysis of OCT data.