A Deep Learning Model for Segmentation of Geographic Atrophy to Study Its Long-Term Natural History

Artificial intelligence for geographic atrophy: pearls and pitfalls

PURPOSE OF REVIEW

This review aims to address the recent advances of artificial intelligence (AI) in the context of clinical management of geographic atrophy (GA), a vision-impairing late-stage manifestation of age-related macular degeneration (AMD).

RECENT FINDINGS

Recent literature shows substantial advancements in the development of AI systems to segment GA lesions on multimodal retinal images, including color fundus photography (CFP), fundus autofluorescence (FAF) and optical coherence tomography (OCT), providing innovative solutions to screening and early diagnosis. Especially, the high resolution and 3D-nature of OCT has provided an optimal source of data for the training and validation of novel algorithms. The use of AI to measure progression in the context of newly approved GA therapies, has shown that AI methods may soon be indispensable for patient management. To date, while many AI models have been reported on, their implementation in the real-world has only just started. The aim is to make the benefits of AI-based personalized treatment accessible and far-reaching.

SUMMARY

The most recent advances (pearls) and challenges (pitfalls) associated with AI methods and their clinical implementation in the context of GA will be discussed.

Strong versus Weak Data Labeling for Artificial Intelligence Algorithms in the Measurement of Geographic Atrophy

Purpose

To gain an understanding of data labeling requirements to train deep learning models for measurement of geographic atrophy (GA) with fundus autofluorescence (FAF) images.

Design

Evaluation of artificial intelligence (AI) algorithms.

Subjects

The Age-Related Eye Disease Study 2 (AREDS2) images were used for training and cross-validation, and GA clinical trial images were used for testing.

Methods

Training data consisted of 2 sets of FAF images; 1 with area measurements only and no indication of GA location (Weakly labeled) and the second with GA segmentation masks (Strongly labeled).

Main Outcome Measures

Bland-Altman plots and scatter plots were used to compare GA area measurement between ground truth and AI measurements. The Dice coefficient was used to compare accuracy of segmentation of the Strong model.

Results

In the cross-validation AREDS2 data set (n = 601), the mean (standard deviation [SD]) area of GA measured by human grader, Weakly labeled AI model, and Strongly labeled AI model was 6.65 (6.3) mm2, 6.83 (6.29) mm2, and 6.58 (6.24) mm2, respectively. The mean difference between ground truth and AI was 0.18 mm2 (95% confidence interval, [CI], -7.57 to 7.92) for the Weakly labeled model and -0.07 mm2 (95% CI, -1.61 to 1.47) for the Strongly labeled model. With GlaxoSmithKline testing data (n = 156), the mean (SD) GA area was 9.79 (5.6) mm2, 8.82 (4.61) mm2, and 9.55 (5.66) mm2 for human grader, Strongly labeled AI model, and Weakly labeled AI model, respectively. The mean difference between ground truth and AI for the 2 models was -0.97 mm2 (95% CI, -4.36 to 2.41) and -0.24 mm2 (95% CI, -4.98 to 4.49), respectively. The Dice coefficient was 0.99 for intergrader agreement, 0.89 for the cross-validation data, and 0.92 for the testing data.

Conclusions

Deep learning models can achieve reasonable accuracy even with Weakly labeled data. Training methods that integrate large volumes of Weakly labeled images with small number of Strongly labeled images offer a promising solution to overcome the burden of cost and time for data labeling.

Financial Disclosures

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

Measuring Geographic Atrophy Area Using Column-Based Machine Learning Software on Spectral-Domain Optical Coherence Tomography versus Fundus Auto Fluorescence

BACKGROUND

The purpose of this study was to compare geographic atrophy (GA) area semi-automatic measurement using fundus autofluorescence (FAF) versus optical coherence tomography (OCT) annotation with the cRORA (complete retinal pigment epithelium and outer retinal atrophy) criteria.

METHODS

GA findings on FAF and OCT were semi-automatically annotated at a single time point in 36 pairs of FAF and OCT scans obtained from 36 eyes in 24 patients with dry age-related macular degeneration (AMD). The GA area, focality, perimeter, circularity, minimum and maximum Feret diameter, and minimum distance from the center were compared between FAF and OCT annotations.

RESULTS

The total GA area measured on OCT was 4.74 ± 3.80 mm2. In contrast, the total GA measured on FAF was 13.47 ± 8.64 mm2 (p < 0.0001), with a mean difference of 8.72 ± 6.35 mm2. Multivariate regression analysis revealed a significant correlation between the difference in area between OCT and FAF and the total baseline lesion perimeter and maximal lesion diameter measured on OCT (adjusted r2: 0.52; p < 0.0001) and the total baseline lesion area measured on FAF (adjusted r2: 0.83; p < 0.0001).

CONCLUSIONS

We report that the GA area measured on FAF differs significantly from the GA area measured on OCT. Further research is warranted in order to determine the clinical relevance of these findings.

Combination of optical coherence tomography-derived shape and texture features are associated with development of sub-foveal geographic atrophy in dry AMD

Geographic atrophy (GA) is an advanced form of dry age-related macular degeneration (AMD) that leads to progressive and irreversible vision loss. Identifying patients with greatest risk of GA progression is important for targeted utilization of emerging therapies. This study aimed to comprehensively evaluate the role of shape-based fractal dimension features ( F fd ) of sub-retinal pigment epithelium (sub-RPE) compartment and texture-based radiomics features ( F t ) of Ellipsoid Zone (EZ)-RPE and sub-RPE compartments for risk stratification for subfoveal GA (sfGA) progression. This was a retrospective study of 137 dry AMD subjects with a 5-year follow-up. Based on sfGA status at year 5, eyes were categorized as Progressors and Non-progressors. A total of 15 shape-based F fd of sub-RPE surface and 494 F t from each of sub-RPE and EZ-RPE compartments were extracted from baseline spectral domain-optical coherence tomography scans. The top nine features were identified from F fd and F t feature pool separately using minimum Redundancy maximum Relevance feature selection and used to train a Random Forest (RF) classifier independently using three-fold cross validation on the training set ( S t , N = 90) to distinguish between sfGA Progressors and Non-progressors. Combined F fd and F t was also evaluated in predicting risk of sfGA progression. The RF classifier yielded AUC of 0.85, 0.79 and 0.89 on independent test set ( S v , N = 47) using F fd , F t , and their combination, respectively. Using combined F fd and F t , the improvement in AUC was statistically significant on S v with p-values of 0.032 and 0.04 compared to using only F fd and only F t , respectively. Combined F fd and F t appears to identify high-risk patients. Our results show that FD and texture features could be potentially used for predicting risk of sfGA progression and future therapeutic response.

Oral Antioxidant and Lutein/Zeaxanthin Supplements Slow Geographic Atrophy Progression to the Fovea in Age-Related Macular Degeneration

PURPOSE

To determine whether oral micronutrient supplementation slows geographic atrophy (GA) progression in age-related macular degeneration (AMD).

DESIGN

Post hoc analysis of Age-Related Eye Disease Study (AREDS) and AREDS2, multicenter randomized placebo-controlled trials of oral micronutrient supplementation, each with 2 × 2 factorial design.

PARTICIPANTS

A total of 392 eyes (318 participants) with GA in AREDS and 1210 eyes (891 participants) with GA in AREDS2.

METHODS

The AREDS participants were randomly assigned to oral antioxidants (500 mg vitamin C, 400 IU vitamin E, 15 mg β-carotene), 80 mg zinc, combination, or placebo. The AREDS2 participants were randomly assigned to 10 mg lutein/2 mg zeaxanthin, 350 mg docosahexaenoic acid/650 mg eicosapentaenoic acid, combination, or placebo. Consenting AREDS2 participants were also randomly assigned to alternative AREDS formulations: original; no beta-carotene; 25 mg zinc instead of 80 mg; both.

MAIN OUTCOME MEASURES

(1) Change in GA proximity to central macula over time and (2) change in square root GA area over time, each measured from color fundus photographs at annual visits and analyzed by mixed-model regression according to randomized assignments.

RESULTS

In AREDS eyes with noncentral GA (n = 208), proximity-based progression toward the central macula was significantly slower with randomization to antioxidants versus none, at 50.7 μm/year (95% confidence interval [CI], 38.0-63.4 μm/year) versus 72.9 μm/year (95% CI, 61.3-84.5 μm/year; P = 0.012), respectively. In AREDS2 eyes with noncentral GA, in participants assigned to AREDS antioxidants without β-carotene (n = 325 eyes), proximity-based progression was significantly slower with randomization to lutein/zeaxanthin versus none, at 80.1 μm/year (95% CI, 60.9-99.3 μm/year) versus 114.4 μm/year (95% CI, 96.2-132.7 μm/year; P = 0.011), respectively. In AREDS eyes with any GA (n = 392), area-based progression was not significantly different with randomization to antioxidants versus none (P = 0.63). In AREDS2 eyes with any GA, in participants assigned to AREDS antioxidants without β-carotene (n = 505 eyes), area-based progression was not significantly different with randomization to lutein/zeaxanthin versus none (P = 0.64).

CONCLUSIONS

Oral micronutrient supplementation slowed GA progression toward the central macula, likely by augmenting the natural phenomenon of foveal sparing.

FINANCIAL DISCLOSURE(S)

Proprietary or commercial disclosure may be found after the references.

Deep Learning Approaches for Detecting of Nascent Geographic Atrophy in Age-Related Macular Degeneration

Purpose

Nascent geographic atrophy (nGA) refers to specific features seen on OCT B-scans, which are strongly associated with the future development of geographic atrophy (GA). This study sought to develop a deep learning model to screen OCT B-scans for nGA that warrant further manual review (an artificial intelligence [AI]-assisted approach), and to determine the extent of reduction in OCT B-scan load requiring manual review while maintaining near-perfect nGA detection performance.

Design

Development and evaluation of a deep learning model.

Participants

One thousand eight hundred and eighty four OCT volume scans (49 B-scans per volume) without neovascular age-related macular degeneration from 280 eyes of 140 participants with bilateral large drusen at baseline, seen at 6-monthly intervals up to a 36-month period (from which 40 eyes developed nGA).

Methods

OCT volume and B-scans were labeled for the presence of nGA. Their presence at the volume scan level provided the ground truth for training a deep learning model to identify OCT B-scans that potentially showed nGA requiring manual review. Using a threshold that provided a sensitivity of 0.99, the B-scans identified were assigned the ground truth label with the AI-assisted approach. The performance of this approach for detecting nGA across all visits, or at the visit of nGA onset, was evaluated using fivefold cross-validation.

Main Outcome Measures

Sensitivity for detecting nGA, and proportion of OCT B-scans requiring manual review.

Results

The AI-assisted approach (utilizing outputs from the deep learning model to guide manual review) had a sensitivity of 0.97 (95% confidence interval [CI] = 0.93-1.00) and 0.95 (95% CI = 0.87-1.00) for detecting nGA across all visits and at the visit of nGA onset, respectively, when requiring manual review of only 2.7% and 1.9% of selected OCT B-scans, respectively.

Conclusions

A deep learning model could be used to enable near-perfect detection of nGA onset while reducing the number of OCT B-scans requiring manual review by over 50-fold. This AI-assisted approach shows promise for substantially reducing the current burden of manual review of OCT B-scans to detect this crucial feature that portends future development of GA.

Financial Disclosures

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

Development and validation of a deep learning model to predict axial length from ultra-wide field images

BACKGROUND

To validate the feasibility of building a deep learning model to predict axial length (AL) for moderate to high myopic patients from ultra-wide field (UWF) images.

METHODS

This study included 6174 UWF images from 3134 myopic patients during 2014 to 2020 in Eye and ENT Hospital of Fudan University. Of 6174 images, 4939 were used for training, 617 for validation, and 618 for testing. The coefficient of determination (R2), mean absolute error (MAE), and mean squared error (MSE) were used for model performance evaluation.

RESULTS

The model predicted AL with high accuracy. Evaluating performance of R2, MSE and MAE were 0.579, 1.419 and 0.9043, respectively. Prediction bias of 64.88% of the tests was under 1-mm error, 76.90% of tests was within the range of 5% error and 97.57% within 10% error. The prediction bias had a strong negative correlation with true AL values and showed significant difference between male and female (P < 0.001). Generated heatmaps demonstrated that the model focused on posterior atrophy changes in pathological fundus and peri-optic zone in normal fundus. In sex-specific models, R2, MSE, and MAE results of the female AL model were 0.411, 1.357, and 0.911 in female dataset and 0.343, 2.428, and 1.264 in male dataset. The corresponding metrics of male AL models were 0.216, 2.900, and 1.352 in male dataset and 0.083, 2.112, and 1.154 in female dataset.

CONCLUSIONS

It is feasible to utilize deep learning models to predict AL for moderate to high myopic patients with UWF images.

Phase 2 Trial Evaluating Minocycline for Geographic Atrophy in Age-Related Macular Degeneration: A Nonrandomized Controlled Trial

Importance

Existing therapies to slow geographic atrophy (GA) enlargement in age-related macular degeneration (AMD) have relatively modest anatomic efficacy, require intravitreal administration, and increase the risk of neovascular AMD. Additional therapeutic approaches are desirable.

Objective

To evaluate the safety and possible anatomic efficacy of oral minocycline, a microglial inhibitor, for the treatment of GA in AMD.

Design, Setting, and Participants

This was a phase 2, prospective, single-arm, 45-month, nonrandomized controlled trial conducted from December 2016 to April 2023. Patients with GA from AMD in 1 or both eyes were recruited from the National Institutes of Health (Bethesda, Maryland) and Bristol Eye Hospital (Bristol, UK). Study data were analyzed from September 2022 to May 2023.

Intervention

After a 9-month run-in phase, participants began oral minocycline, 100 mg, twice daily for 3 years.

Main Outcomes and Measures

The primary outcome measure was the difference in rate of change of square root GA area on fundus autofluorescence between the 24-month treatment phase and 9-month run-in phase.

Results

Of the 37 participants enrolled (mean [SD] age, 74.3 [7.6] years; 21 female [57%]), 36 initiated the treatment phase. Of these participants, 21 (58%) completed at least 33 months, whereas 15 discontinued treatment (8 by request, 6 for adverse events/illness, and 1 death). Mean (SE) square root GA enlargement rate in study eyes was 0.31 (0.03) mm per year during the run-in phase and 0.28 (0.02) mm per year during the treatment phase. The primary outcome measure of mean (SE) difference in enlargement rates between the 2 phases was -0.03 (0.03) mm per year (P = .39). Similarly, secondary outcome measures of GA enlargement rate showed no differences between the 2 phases. The secondary outcome measures of mean difference in rate of change between 2 phases were 0.2 letter score per month (95% CI, -0.4 to 0.9; P = .44) for visual acuity and 0.7 μm per month (-0.4 to 1.8; P = .20) for subfoveal retinal thickness. Of the 129 treatment-emergent adverse events among 32 participants, 49 (38%) were related to minocycline (with no severe or ocular events), including elevated thyrotropin level (15 participants) and skin hyperpigmentation/discoloration (8 participants).

Conclusions and Relevance

In this phase 2 nonrandomized controlled trial, oral minocycline was not associated with a decrease in GA enlargement over 24 months, compared with the run-in phase. This observation was consistent across primary and secondary outcome measures. Oral minocycline at this dose is likely not associated with slower rate of enlargement of GA in AMD.

Recurrent and Concurrent Prediction of Longitudinal Progression of Stargardt Atrophy and Geographic Atrophy

Stargardt disease and age-related macular degeneration are the leading causes of blindness in the juvenile and geriatric populations, respectively. The formation of atrophic regions of the macula is a hallmark of the end-stages of both diseases. The progression of these diseases is tracked using various imaging modalities, two of the most common being fundus autofluorescence (FAF) imaging and spectral-domain optical coherence tomography (SD-OCT). This study seeks to investigate the use of longitudinal FAF and SD-OCT imaging (month 0, month 6, month 12, and month 18) data for the predictive modelling of future atrophy in Stargardt and geographic atrophy. To achieve such an objective, we develop a set of novel deep convolutional neural networks enhanced with recurrent network units for longitudinal prediction and concurrent learning of ensemble network units (termed ReConNet) which take advantage of improved retinal layer features beyond the mean intensity features. Using FAF images, the neural network presented in this paper achieved mean (± standard deviation, SD) and median Dice coefficients of 0.895 (± 0.086) and 0.922 for Stargardt atrophy, and 0.864 (± 0.113) and 0.893 for geographic atrophy. Using SD-OCT images for Stargardt atrophy, the neural network achieved mean and median Dice coefficients of 0.882 (± 0.101) and 0.906, respectively. When predicting only the interval growth of the atrophic lesions with FAF images, mean (± SD) and median Dice coefficients of 0.557 (± 0.094) and 0.559 were achieved for Stargardt atrophy, and 0.612 (± 0.089) and 0.601 for geographic atrophy. The prediction performance in OCT images is comparably good to that using FAF which opens a new, more efficient, and practical door in the assessment of atrophy progression for clinical trials and retina clinics, beyond widely used FAF. These results are highly encouraging for a high-performance interval growth prediction when more frequent or longer-term longitudinal data are available in our clinics. This is a pressing task for our next step in ongoing research.

Local progression kinetics of macular atrophy in recessive Stargardt disease

BACKGROUND

To determine the effect of lesion topography on progression in Stargardt disease (STGD1).

METHODS

Fundus autofluoresence (excitation 488 nm) images of 193 eyes in patients with proven ABCA4 mutation were semi-automatically segmented for autofluoresence changes: (DDAF) and questionably decreased autofluoresence (QDAF), which are proxies for retinal pigment epithelial (RPE) atrophy. We calculated topographic incidence of DDAF and DDAF + QDAF, as well as velocity of progression of the border of lesions using Euclidean distance mapping.

RESULTS

Incidence of atrophy was highest near the fovea, then decreased in incidence with increased foveal eccentricity. However, the rate of atrophy progression followed the opposite pattern; rate of atrophy increased with distance from foveal center. The mean growth rate 500 microns from the foveal center for DDAF + QDAF was 39 microns per year (95% CI = 28-49), whereas the mean growth rate 3000 microns from the foveal center was 342 microns per year (95% CI = 194-522). No difference in growth rate was noted by axis around the fovea.

CONCLUSIONS

Incidence and progression of atrophy by fundus autofluorescence follow opposite patterns in STGD1. Further, atrophy progression increases significantly with distance from foveal center, which should be taken into consideration in clinical trials.

Deep-GA-Net for Accurate and Explainable Detection of Geographic Atrophy on OCT Scans

Objective

To propose Deep-GA-Net, a 3-dimensional (3D) deep learning network with 3D attention layer, for the detection of geographic atrophy (GA) on spectral domain OCT (SD-OCT) scans, explain its decision making, and compare it with existing methods.

Design

Deep learning model development.

Participants

Three hundred eleven participants from the Age-Related Eye Disease Study 2 Ancillary SD-OCT Study.

Methods

A dataset of 1284 SD-OCT scans from 311 participants was used to develop Deep-GA-Net. Cross-validation was used to evaluate Deep-GA-Net, where each testing set contained no participant from the corresponding training set. En face heatmaps and important regions at the B-scan level were used to visualize the outputs of Deep-GA-Net, and 3 ophthalmologists graded the presence or absence of GA in them to assess the explainability (i.e., understandability and interpretability) of its detections.

Main Outcome Measures

Accuracy, area under receiver operating characteristic curve (AUC), area under precision-recall curve (APR).

Results

Compared with other networks, Deep-GA-Net achieved the best metrics, with accuracy of 0.93, AUC of 0.94, and APR of 0.91, and received the best gradings of 0.98 and 0.68 on the en face heatmap and B-scan grading tasks, respectively.

Conclusions

Deep-GA-Net was able to detect GA accurately from SD-OCT scans. The visualizations of Deep-GA-Net were more explainable, as suggested by 3 ophthalmologists. The code and pretrained models are publicly available at https://github.com/ncbi/Deep-GA-Net.

Financial Disclosures

The author(s) have no proprietary or commercial interest in any materials discussed in this article.

Novel Fractal-Based Sub-RPE Compartment OCT Radiomics Biomarkers Are Associated With Subfoveal Geographic Atrophy in Dry AMD

OBJECTIVE

The purpose of this study was to quantitatively characterize the shape of the sub-retinal pigment epithelium (sub-RPE, i.e., space bounded by RPE and Bruch's membrane) compartment on SD-OCT using fractal dimension (FD) features and evaluate their impact on risk of subfoveal geographic atrophy (sfGA) progression.

METHODS

This was an IRB-approved retrospective study of 137 subjects with dry age-related macular degeneration (AMD) with subfoveal GA. Based on sfGA status at year five, eyes were categorized as "Progressors" and "Non-progressors". FD analysis allows quantification of the degree of shape complexity and architectural disorder associated with a structure. To characterize the structural irregularities along the sub-RPE surface between the two groups of patients, a total of 15 shape descriptors of FD were extracted from the sub-RPE compartment of baseline OCT scans. The top four features were identified using minimum Redundancy maximum Relevance (mRmR) feature selection method and evaluated with Random Forest (RF) classifier using three-fold cross validation from the training set (N = 90). Classifier performance was subsequently validated on the independent test set (N = 47).

RESULTS

Using the top four FD features, a RF classifier yielded an AUC of 0.85 on the independent test set. Mean fractal entropy (p-value = 4.8e-05) was identified as the most significant biomarker; higher values of entropy being associated with greater shape disorder and risk for sfGA progression.

CONCLUSIONS

FD assessment holds promise for identifying high-risk eyes for GA progression.

SIGNIFICANCE

With further validation, FD features could be potentially used for clinical trial enrichment and assessments for therapeutic response in dry AMD patients.

Detection of macular atrophy in age-related macular degeneration aided by artificial intelligence

INTRODUCTION

Age-related macular degeneration (AMD) is a leading cause of irreversible visual impairment worldwide. The endpoint of AMD, both in its dry or wet form, is macular atrophy (MA) which is characterized by the permanent loss of the RPE and overlying photoreceptors either in dry AMD or in wet AMD. A recognized unmet need in AMD is the early detection of MA development.

AREAS COVERED

Artificial Intelligence (AI) has demonstrated great impact in detection of retinal diseases, especially with its robust ability to analyze big data afforded by ophthalmic imaging modalities, such as color fundus photography (CFP), fundus autofluorescence (FAF), near-infrared reflectance (NIR), and optical coherence tomography (OCT). Among these, OCT has been shown to have great promise in identifying early MA using the new criteria in 2018.

EXPERT OPINION

There are few studies in which AI-OCT methods have been used to identify MA; however, results are very promising when compared to other imaging modalities. In this paper, we review the development and advances of ophthalmic imaging modalities and their combination with AI technology to detect MA in AMD. In addition, we emphasize the application of AI-OCT as an objective, cost-effective tool for the early detection and monitoring of the progression of MA in AMD.

Deep Learning to Predict Geographic Atrophy Area and Growth Rate from Multimodal Imaging

OBJECTIVE

To develop deep learning models for annualized geographic atrophy (GA) growth rate prediction using fundus autofluorescence (FAF) images and spectral-domain OCT volumes from baseline visits, which can be used for prognostic covariate adjustment to increase power of clinical trials.

DESIGN

This retrospective analysis estimated GA growth rate as the slope of a linear fit on all available measurements of lesion area over a 2-year period. Three multitask deep learning models-FAF-only, OCT-only, and multimodal (FAF and OCT)-were developed to predict concurrent GA area and annualized growth rate.

PARTICIPANTS

Patients were from prospective and observational lampalizumab clinical trials.

METHODS

The 3 models were trained on the development data set, tested on the holdout set, and further evaluated on the independent test sets. Baseline FAF images and OCT volumes from study eyes of patients with bilateral GA (NCT02247479; NCT02247531; and NCT02479386) were split into development (1279 patients/eyes) and holdout (443 patients/eyes) sets. Baseline FAF images from study eyes of NCT01229215 (106 patients/eyes) and NCT02399072 (169 patients/eyes) were used as independent test sets.

MAIN OUTCOME MEASURES

Model performance was evaluated using squared Pearson correlation coefficient (r²) between observed and predicted lesion areas/growth rates. Confidence intervals were calculated by bootstrap resampling (B = 10 000).

RESULTS

On the holdout data set, r² (95% confidence interval) of the FAF-only, OCT-only, and multimodal models for GA lesion area prediction was 0.96 (0.95-0.97), 0.91 (0.87-0.95), and 0.94 (0.92-0.96), respectively, and for GA growth rate prediction was 0.48 (0.41-0.55), 0.36 (0.29-0.43), and 0.47 (0.40-0.54), respectively. On the 2 independent test sets, r² of the FAF-only model for GA lesion area was 0.98 (0.97-0.99) and 0.95 (0.93-0.96), and for GA growth rate was 0.65 (0.52-0.75) and 0.47 (0.34-0.60).

CONCLUSIONS

We show the feasibility of using baseline FAF images and OCT volumes to predict individual GA area and growth rates using a multitask deep learning approach. The deep learning-based growth rate predictions could be used for covariate adjustment to increase power of clinical trials.

FINANCIAL DISCLOSURE(S)

Proprietary or commercial disclosure may be found after the references.

Training Deep Learning Models to Work on Multiple Devices by Cross-Domain Learning with No Additional Annotations

PURPOSE

To create an unsupervised cross-domain segmentation algorithm for segmenting intraretinal fluid and retinal layers on normal and pathologic macular OCT images from different manufacturers and camera devices.

DESIGN

We sought to use generative adversarial networks (GANs) to generalize a segmentation model trained on one OCT device to segment B-scans obtained from a different OCT device manufacturer in a fully unsupervised approach without labeled data from the latter manufacturer.

PARTICIPANTS

A total of 732 OCT B-scans from 4 different OCT devices (Heidelberg Spectralis, Topcon 1000, Maestro2, and Zeiss Plex Elite 9000).

METHODS

We developed an unsupervised GAN model, GANSeg, to segment 7 retinal layers and intraretinal fluid in Topcon 1000 OCT images (domain B) that had access only to labeled data on Heidelberg Spectralis images (domain A). GANSeg was unsupervised because it had access only to 110 Heidelberg labeled OCTs and 556 raw and unlabeled Topcon 1000 OCTs. To validate GANSeg segmentations, 3 masked graders manually segmented 60 OCTs from an external Topcon 1000 test dataset independently. To test the limits of GANSeg, graders also manually segmented 3 OCTs from Zeiss Plex Elite 9000 and Topcon Maestro2. A U-Net was trained on the same labeled Heidelberg images as baseline. The GANSeg repository with labeled annotations is at https://github.com/uw-biomedical-ml/ganseg.

MAIN OUTCOME MEASURES

Dice scores comparing segmentation results from GANSeg and the U-Net model with the manual segmented images.

RESULTS

Although GANSeg and U-Net achieved comparable Dice scores performance as human experts on the labeled Heidelberg test dataset, only GANSeg achieved comparable Dice scores with the best performance for the ganglion cell layer plus inner plexiform layer (90%; 95% confidence interval [CI], 68%-96%) and the worst performance for intraretinal fluid (58%; 95% CI, 18%-89%), which was statistically similar to human graders (79%; 95% CI, 43%-94%). GANSeg significantly outperformed the U-Net model. Moreover, GANSeg generalized to both Zeiss and Topcon Maestro2 swept-source OCT domains, which it had never encountered before.

CONCLUSIONS

GANSeg enables the transfer of supervised deep learning algorithms across OCT devices without labeled data, thereby greatly expanding the applicability of deep learning algorithms.

Machine Learning-Based Automated Detection and Quantification of Geographic Atrophy and Hypertransmission Defects Using Spectral Domain Optical Coherence Tomography

The current study describes the development and assessment of innovative, machine learning (ML)-based approaches for automated detection and pixel-accurate measurements of regions with geographic atrophy (GA) in late-stage age-related macular degeneration (AMD) using optical coherence tomography systems. 900 OCT volumes, 100266 B-scans, and en face OCT images from 341 non-exudative AMD patients with or without GA were included in this study from both Cirrus (Zeiss) and Spectralis (Heidelberg) OCT systems. B-scan and en face level ground truth GA masks were created on OCT B-scan where the segmented ellipsoid zone (EZ) line, retinal pigment epithelium (RPE) line, and bruchs membrane (BM) line overlapped. Two deep learning-based approaches, B-scan level and en face level, were trained. The OCT B-scan model had detection accuracy of 91% and GA area measurement accuracy of 94%. The en face OCT model had detection accuracy of 82% and GA area measurement accuracy of 96% with primary target of hypertransmission on en face OCT. Accuracy was good for both devices tested (92-97%). Automated lesion size stratification for CAM cRORA definition of 250um minimum lesion size was feasible. High-performance models for automatic detection and segmentation of GA area were achieved using OCT systems and deep learning. The automatic measurements showed high correlation with the ground truth. The en face model excelled at identification of hypertransmission defects. The models performance generalized well across device types tested. Future development will include integration of both models to enhance feature detection across GA lesions as well as isolating hypertransmission defects without GA for pre-GA biomarker extraction.

Automated Segmentation of Autofluorescence Lesions in Stargardt Disease

OBJECTIVE

To train a deep learning (DL) algorithm to perform fully automated semantic segmentation of multiple autofluorescence lesion types in Stargardt disease.

DESIGN

Cross-sectional study with retrospective imaging data.

SUBJECTS

The study included 193 images from 193 eyes of 97 patients with Stargardt disease.

METHODS

Fundus autofluorescence images obtained from patient visits between 2013 and 2020 were annotated with ground-truth labels. Model training and evaluation were performed using fivefold cross-validation.

MAIN OUTCOMES MEASURES

Dice similarity coefficients, intraclass correlation coefficients, and Bland-Altman analyses comparing algorithm-predicted and grader-labeled segmentations.

RESULTS

The overall Dice similarity coefficient across all lesion classes was 0.78 (95% confidence interval [CI], 0.69-0.86). Dice coefficients were 0.90 (95% CI, 0.85-0.94) for areas of definitely decreased autofluorescence (DDAF), 0.55 (95% CI, 0.35-0.76) for areas of questionably decreased autofluorescence (QDAF), and 0.88 (95% CI, 0.73-1.00) for areas of abnormal background autofluorescence (ABAF). Intraclass correlation coefficients comparing the ground-truth and automated methods were 0.997 (95% CI, 0.996-0.998) for DDAF, 0.863 (95% CI, 0.823-0.895) for QDAF, and 0.974 (95% CI, 0.966-0.980) for ABAF.

CONCLUSIONS

A DL algorithm performed accurate segmentation of autofluorescence lesions in Stargardt disease, demonstrating the feasibility of fully automated segmentation as an alternative to manual or semiautomated labeling methods.

Fundus autofluorescence and optical coherence tomography biomarkers associated with the progression of geographic atrophy secondary to age-related macular degeneration

OBJECTIVES

To investigate the impact of qualitatively graded and deep learning quantified imaging biomarkers on growth of geographic atrophy (GA) secondary to age-related macular degeneration.

METHODS

This prospective study included 1062 visits of 181 eyes of 100 patients with GA. Spectral-domain optical coherence tomography (SD-OCT) and fundus autofluorescence (FAF) images were acquired at each visit. Hyperreflective foci (HRF) were quantitatively assessed in SD-OCT volumes using a validated deep learning algorithm. FAF images were graded for FAF patterns, subretinal drusenoid deposits (SDD), GA lesion configuration and atrophy enlargement. Linear mixed models were calculated to investigate associations between all parameters and GA progression.

RESULTS

FAF patterns were significantly associated with GA progression (p < 0.001). SDD was associated with faster GA growth (p = 0.005). Eyes with higher HRF concentrations showed a trend towards faster GA progression (p = 0.072) and revealed a significant impact on GA enlargement in interaction with FAF patterns (p = 0.01). The fellow eye status had no significant effect on lesion enlargement (p > 0.05). The diffuse-trickling FAF pattern exhibited significantly higher HRF concentrations than any other pattern (p < 0.001).

CONCLUSION

Among a wide range of investigated biomarkers, SDD and FAF patterns, particularly in interaction with HRF, significantly impact GA progression. Fully automated quantification of retinal imaging biomarkers such as HRF is both reliable and merited as HRF are indicators of retinal pigment epithelium dysmorphia, a central pathogenetic mechanism in GA. Identifying disease markers using the combination of FAF and SD-OCT is of high prognostic value and facilitates individualized patient management in a clinical setting.

Prediction of visual function from automatically quantified optical coherence tomography biomarkers in patients with geographic atrophy using machine learning

Geographic atrophy (GA) is a vision-threatening manifestation of age-related macular degeneration (AMD), one of the leading causes of blindness globally. Objective, rapid, reliable, and scalable quantification of GA from optical coherence tomography (OCT) retinal scans is necessary for disease monitoring, prognostic research, and clinical endpoints for therapy development. Such automatically quantified biomarkers on OCT are likely to further elucidate structure-function correlation in GA and thus the pathophysiological mechanisms of disease development and progression. In this work, we aimed to predict visual function with machine-learning applied to automatically acquired quantitative imaging biomarkers in GA. A post-hoc analysis of data from a clinical trial and routine clinical care was conducted. A deep-learning automated segmentation model was applied on OCT scans from 476 eyes (325 patients) with GA. A separate machine learning prediction model (Random Forest) used the resultant quantitative OCT (qOCT) biomarkers to predict cross-sectional visual acuity under standard (VA) and low luminance (LLVA). The primary outcome was regression coefficient (r2) and mean absolute error (MAE) for cross-sectional VA and LLVA in Early Treatment Diabetic Retinopathy Study (ETDRS) letters. OCT parameters were predictive of VA (r2 0.40 MAE 11.7 ETDRS letters) and LLVA (r2 0.25 MAE 12.1). Normalised random forest feature importance, as a measure of the predictive value of the three constituent features of GA; retinal pigment epithelium (RPE)-loss, photoreceptor degeneration (PDR), hypertransmission and their locations, was reported both on voxel-level heatmaps and ETDRS-grid subfields. The foveal region (46.5%) and RPE-loss (31.1%) had greatest predictive importance for VA. For LLVA, however, non-foveal regions (74.5%) and PDR (38.9%) were most important. In conclusion, automated qOCT biomarkers demonstrate predictive significance for VA and LLVA in GA. LLVA is itself predictive of GA progression, implying that the predictive qOCT biomarkers provided by our model are also prognostic.

Automated segmentation and feature discovery of age-related macular degeneration and Stargardt disease via self-attended neural networks

Age-related macular degeneration (AMD) and Stargardt disease are the leading causes of blindness for the elderly and young adults respectively. Geographic atrophy (GA) of AMD and Stargardt atrophy are their end-stage outcomes. Efficient methods for segmentation and quantification of these atrophic lesions are critical for clinical research. In this study, we developed a deep convolutional neural network (CNN) with a trainable self-attended mechanism for accurate GA and Stargardt atrophy segmentation. Compared with traditional post-hoc attention mechanisms which can only visualize CNN features, our self-attended mechanism is embedded in a fully convolutional network and directly involved in training the CNN to actively attend key features for enhanced algorithm performance. We applied the self-attended CNN on the segmentation of AMD and Stargardt atrophic lesions on fundus autofluorescence (FAF) images. Compared with a preexisting regular fully convolutional network (the U-Net), our self-attended CNN achieved 10.6% higher Dice coefficient and 17% higher IoU (intersection over union) for AMD GA segmentation, and a 22% higher Dice coefficient and a 32% higher IoU for Stargardt atrophy segmentation. With longitudinal image data having over a longer time, the developed self-attended mechanism can also be applied on the visual discovery of early AMD and Stargardt features.

Automatic geographic atrophy segmentation using optical attenuation in OCT scans with deep learning

A deep learning algorithm was developed to automatically identify, segment, and quantify geographic atrophy (GA) based on optical attenuation coefficients (OACs) calculated from optical coherence tomography (OCT) datasets. Normal eyes and eyes with GA secondary to age-related macular degeneration were imaged with swept-source OCT using 6 × 6 mm scanning patterns. OACs calculated from OCT scans were used to generate customized composite en face OAC images. GA lesions were identified and measured using customized en face sub-retinal pigment epithelium (subRPE) OCT images. Two deep learning models with the same U-Net architecture were trained using OAC images and subRPE OCT images. Model performance was evaluated using DICE similarity coefficients (DSCs). The GA areas were calculated and compared with manual segmentations using Pearson's correlation and Bland-Altman plots. In total, 80 GA eyes and 60 normal eyes were included in this study, out of which, 16 GA eyes and 12 normal eyes were used to test the models. Both models identified GA with 100% sensitivity and specificity on the subject level. With the GA eyes, the model trained with OAC images achieved significantly higher DSCs, stronger correlation to manual results and smaller mean bias than the model trained with subRPE OCT images (0.940 ± 0.032 vs 0.889 ± 0.056, p = 0.03, paired t-test, r = 0.995 vs r = 0.959, mean bias = 0.011 mm vs mean bias = 0.117 mm). In summary, the proposed deep learning model using composite OAC images effectively and accurately identified, segmented, and quantified GA using OCT scans.

Progression of cRORA (Complete RPE and Outer Retinal Atrophy) in Dry Age-Related Macular Degeneration Measured Using SD-OCT

Purpose

The purpose of this study was to evaluate the long-term rate of progression and baseline predictors of geographic atrophy (GA) using complete retinal pigment epithelium and outer retinal atrophy (cRORA) annotation criteria.

Methods

This is a retrospective study. Columns of GA were manually annotated by two graders using a self-developed software on optical coherence tomography (OCT) B-scans and projected onto the infrared images. The primary outcomes were: (1) rate of area progression, (2) rate of square root area progression, and (3) rate of radial progression towards the fovea. The effects of 11 additional baseline predictors on the primary outcomes were analyzed: total area, focality (defined as the number of lesions whose area is >0.05 mm2), circularity, total lesion perimeter, minimum diameter, maximum diameter, minimum distance from the center, sex, age, presence/absence of hypertension, and lens status.

Results

GA was annotated in 33 pairs of baseline and follow-up OCT scans from 33 eyes of 18 patients with dry age-related macular degeneration (AMD) followed for at least 6 months. The mean rate of area progression was 1.49 ± 0.86 mm2/year (P < 0.0001 vs. baseline), and the mean rate of square root area progression was 0.33 ± 0.15 mm/year (P < 0.0001 vs. baseline). The mean rate of radial progression toward the fovea was 0.07 ± 0.11 mm/year. A multiple variable linear regression model (adjusted r2 = 0.522) revealed that baseline focality and female sex were significantly correlated with the rate of GA area progression.

Conclusions

GA area progression was quantified using OCT as an alternative to conventional measurements performed on fundus autofluorescence images. Baseline focality correlated with GA area progression rate and lesion's minimal distance from the center correlated with GA radial progression rate toward the center. These may be important markers for the assessment of GA activity.

Translational Relevance

Advanced method linking specific retinal micro-anatomy to GA area progression analysis.

Risk Classification for Progression to Subfoveal Geographic Atrophy in Dry Age-Related Macular Degeneration Using Machine Learning-Enabled Outer Retinal Feature Extraction

BACKGROUND AND OBJECTIVE

To evaluate the utility of spectral-domain optical coherence tomography biomarkers to predict the development of subfoveal geographic atrophy (sfGA).

PATIENTS AND METHODS

This was a retrospective cohort analysis including 137 individuals with dry age-related macular degeneration without sfGA with 5 years of follow-up. Multiple spectral-domain optical coherence tomography quantitative metrics were generated, including ellipsoid zone (EZ) integrity and subretinal pigment epithelium (sub-RPE) compartment features.

RESULTS

Reduced mean EZ-RPE central subfield thickness and increased sub-RPE compartment thickness were significantly different between sfGA convertors and nonconvertors at baseline in both 2-year and 5-year sfGA risk assessment. Longitudinal change assessment showed a significantly higher degradation of EZ integrity in sfGA convertors. The predictive performance of a machine learning classification model based on 5-year and 2-year risk conversion to sfGA demonstrated an area under the receiver operating characteristic curve of 0.92 ± 0.06 and 0.96 ± 0.04, respectively.

CONCLUSIONS

Quantitative outer retinal and sub-RPE feature assessment using a machine learning-enabled retinal segmentation platform provides multiple parameters that are associated with progression to sfGA. [Ophthalmic Surg Lasers Imaging. 2022;53:31-39.].

Trustworthy AI: Closing the gap between development and integration of AI systems in ophthalmic practice

An increasing number of artificial intelligence (AI) systems are being proposed in ophthalmology, motivated by the variety and amount of clinical and imaging data, as well as their potential benefits at the different stages of patient care. Despite achieving close or even superior performance to that of experts, there is a critical gap between development and integration of AI systems in ophthalmic practice. This work focuses on the importance of trustworthy AI to close that gap. We identify the main aspects or challenges that need to be considered along the AI design pipeline so as to generate systems that meet the requirements to be deemed trustworthy, including those concerning accuracy, resiliency, reliability, safety, and accountability. We elaborate on mechanisms and considerations to address those aspects or challenges, and define the roles and responsibilities of the different stakeholders involved in AI for ophthalmic care, i.e., AI developers, reading centers, healthcare providers, healthcare institutions, ophthalmological societies and working groups or committees, patients, regulatory bodies, and payers. Generating trustworthy AI is not a responsibility of a sole stakeholder. There is an impending necessity for a collaborative approach where the different stakeholders are represented along the AI design pipeline, from the definition of the intended use to post-market surveillance after regulatory approval. This work contributes to establish such multi-stakeholder interaction and the main action points to be taken so that the potential benefits of AI reach real-world ophthalmic settings.

Personalized Atrophy Risk Mapping in Age-Related Macular Degeneration

Purpose

To develop and validate an automatic retinal pigment epithelial and outer retinal atrophy (RORA) progression prediction model for nonexudative age-related macular degeneration (AMD) cases in optical coherence tomography (OCT) scans.

Methods

Longitudinal OCT data from 129 eyes/119 patients with RORA was collected and separated into training and testing groups. RORA was automatically segmented in all scans and additionally manually annotated in the test scans. OCT-based features such as layers thicknesses, mean reflectivity, and a drusen height map served as an input to the deep neural network. Based on the baseline OCT scan or the previous visit OCT, en face RORA predictions were calculated for future patient visits. The performance was quantified over time with the means of Dice scores and square root area errors.

Results

The average Dice score for segmentations at baseline was 0.85. When predicting progression from baseline OCTs, the Dice scores ranged from 0.73 to 0.80 for total RORA area and from 0.46 to 0.72 for RORA growth region. The square root area error ranged from 0.13 mm to 0.33 mm. By providing continuous time output, the model enabled creation of a patient-specific atrophy risk map.

Conclusions

We developed a machine learning method for RORA progression prediction, which provides continuous-time output. It was used to compute atrophy risk maps, which indicate time-to-RORA-conversion, a novel and clinically relevant way of representing disease progression.

Translational Relevance

Application of recent advances in artificial intelligence to predict patient-specific progression of atrophic AMD.

Progression of Geographic Atrophy: Epistemic Uncertainties Affecting Mathematical Models and Machine Learning

Purpose

The purpose of this study was to identify a taxonomy of epistemic uncertainties that affect results for geographic atrophy (GA) assessment and progression.

Methods

An important source of variability is called "epistemic uncertainty," which is due to incomplete system knowledge (i.e. limitations in measurement devices, artifacts, and human subjective evaluation, including annotation errors). In this study, different epistemic uncertainties affecting the analysis of GA were identified and organized into a taxonomy. The uncertainties were discussed and analyzed, and an example was provided in the case of model structure uncertainty by characterizing progression of GA by mathematical modelling and machine learning. It was hypothesized that GA growth follows a logistic (sigmoidal) function. Using case studies, the GA growth data were used to test the sigmoidal hypothesis.

Results

Epistemic uncertainties were identified, including measurement error (imperfect outcomes from measuring tools), subjective judgment (grading affected by grader's vision and experience), model input uncertainties (data corruption or entry errors), and model structure uncertainties (elucidating the right progression pattern). Using GA growth data from case studies, it was demonstrated that GA growth can be represented by a sigmoidal function, where growth eventually approaches an upper limit.

Conclusion

Epistemic uncertainties contribute to errors in study results and are reducible if identified and addressed. By prior identification of epistemic uncertainties, it is possible to (a) quantify uncertainty not accounted for by natural statistical variability, and (b) reduce the presence of these uncertainties in future studies.

Translational Relevance

Lowering epistemic uncertainty will reduce experimental error, improve consistency and reproducibility, and increase confidence in diagnostics.

Local Progression Kinetics of Geographic Atrophy Depends Upon the Border Location

Purpose

To assess the influence of lesion morphology and location on geographic atrophy (GA) growth rate.

Methods

We manually delineated GA on color fundus photographs of 237 eyes in the Age-Related Eye Disease Study. We calculated local border expansion rate (BER) as the linear distance that a point on the GA border traveled over 1 year based on a Euclidean distance map. Eye-specific BER was defined as the mean local BER of all points on the GA border in an eye. The percentage area affected by GA was defined as the GA area divided by the total retinal area in the region.

Results

GA enlarged 1.51 ± 1.96 mm2 in area and 0.13 ± 0.11 mm in distance over 1 year. The GA area growth rate (mm2/y) was associated with the baseline GA area (P < 0.001), perimeter (P < 0.001), lesion number (P < 0.001), and circularity index (P < 0.001); in contrast, eye-specific BER (mm/y) was not significantly associated with any of these factors. As the retinal eccentricity increased from 0 to 3.5 mm, the local BER increased from 0.10 to 0.24 mm/y (P < 0.001); in contrast, the percentage of area affected by GA decreased from 49.3% to 2.3%.

Conclusions

Using distance-based measurements allows GA progression evaluation without significant confounding effects from baseline GA morphology. Local GA progression rates increased as a function of retinal eccentricity within the macula which is opposite of the trend for GA distribution, suggesting that GA initiation and enlargement may be mediated by different biological processes.

Improving Interpretability in Machine Diagnosis

Purpose

Manually identifying geographic atrophy (GA) presence and location on OCT volume scans can be challenging and time consuming. This study developed a deep learning model simultaneously (1) to perform automated detection of GA presence or absence from OCT volume scans and (2) to provide interpretability by demonstrating which regions of which B-scans show GA.

Design

Med-XAI-Net, an interpretable deep learning model was developed to detect GA presence or absence from OCT volume scans using only volume scan labels, as well as to interpret the most relevant B-scans and B-scan regions.

Participants

One thousand two hundred eighty-four OCT volume scans (each containing 100 B-scans) from 311 participants, including 321 volumes with GA and 963 volumes without GA.

Methods

Med-XAI-Net simulates the human diagnostic process by using a region-attention module to locate the most relevant region in each B-scan, followed by an image-attention module to select the most relevant B-scans for classifying GA presence or absence in each OCT volume scan. Med-XAI-Net was trained and tested (80% and 20% participants, respectively) using gold standard volume scan labels from human expert graders.

Main Outcome Measures

Accuracy, area under the receiver operating characteristic (ROC) curve, F₁ score, sensitivity, and specificity.

Results

In the detection of GA presence or absence, Med-XAI-Net obtained superior performance (91.5%, 93.5%, 82.3%, 82.8%, and 94.6% on accuracy, area under the ROC curve, F₁ score, sensitivity, and specificity, respectively) to that of 2 other state-of-the-art deep learning methods. The performance of ophthalmologists grading only the 5 B-scans selected by Med-XAI-Net as most relevant (95.7%, 95.4%, 91.2%, and 100%, respectively) was almost identical to that of ophthalmologists grading all volume scans (96.0%, 95.7%, 91.8%, and 100%, respectively). Even grading only 1 region in 1 B-scan, the ophthalmologists demonstrated moderately high performance (89.0%, 87.4%, 77.6%, and 100%, respectively).

Conclusions

Despite using ground truth labels during training at the volume scan level only, Med-XAI-Net was effective in locating GA in B-scans and selecting relevant B-scans within each volume scan for GA diagnosis. These results illustrate the strengths of Med-XAI-Net in interpreting which regions and B-scans contribute to GA detection in the volume scan.

Deep Learning Applied to Automated Segmentation of Geographic Atrophy in Fundus Autofluorescence Images

Purpose

This study describes the development of a deep learning algorithm based on the U-Net architecture for automated segmentation of geographic atrophy (GA) lesions in fundus autofluorescence (FAF) images.

Methods

Image preprocessing and normalization by modified adaptive histogram equalization were used for image standardization to improve effectiveness of deep learning. A U-Net-based deep learning algorithm was developed and trained and tested by fivefold cross-validation using FAF images from clinical datasets. The following metrics were used for evaluating the performance for lesion segmentation in GA: dice similarity coefficient (DSC), DSC loss, sensitivity, specificity, mean absolute error (MAE), accuracy, recall, and precision.

Results

In total, 702 FAF images from 51 patients were analyzed. After fivefold cross-validation for lesion segmentation, the average training and validation scores were found for the most important metric, DSC (0.9874 and 0.9779), for accuracy (0.9912 and 0.9815), for sensitivity (0.9955 and 0.9928), and for specificity (0.8686 and 0.7261). Scores for testing were all similar to the validation scores. The algorithm segmented GA lesions six times more quickly than human performance.

Conclusions

The deep learning algorithm can be implemented using clinical data with a very high level of performance for lesion segmentation. Automation of diagnostics for GA assessment has the potential to provide savings with respect to patient visit duration, operational cost and measurement reliability in routine GA assessments.

Translational Relevance

A deep learning algorithm based on the U-Net architecture and image preprocessing appears to be suitable for automated segmentation of GA lesions on clinical data, producing fast and accurate results.

Use of deep learning for structural analysis of computer tomography images of soil samples

Soil samples from several European countries were scanned using medical computer tomography (CT) device and are now available as CT images. The analysis of these samples was carried out using deep learning methods. For this purpose, a VGG16 network was trained with the CT images (X). For the annotation (y) a new method for automated annotation, 'surrogate' learning, was introduced. The generated neural networks (NNs) were subjected to a detailed analysis. Among other things, transfer learning was used to check whether the NN can also be trained to other y-values. Visually, the NN was verified using a gradient-based class activation mapping (grad-CAM) algorithm. These analyses showed that the NN was able to generalize, i.e. to capture the spatial structure of the soil sample. Possible applications of the models are discussed.

Reliability of Retinal Pathology Quantification in Age-Related Macular Degeneration: Implications for Clinical Trials and Machine Learning Applications

Purpose

To investigate the interreader agreement for grading of retinal alterations in age-related macular degeneration (AMD) using a reading center setting.

Methods

In this cross-sectional case series, spectral-domain optical coherence tomography (OCT; Topcon 3D OCT, Tokyo, Japan) scans of 112 eyes of 112 patients with neovascular AMD (56 treatment naive, 56 after three anti-vascular endothelial growth factor injections) were analyzed by four independent readers. Imaging features specific for AMD were annotated using a novel custom-built annotation platform. Dice score, Bland-Altman plots, coefficients of repeatability, coefficients of variation, and intraclass correlation coefficients were assessed.

Results

Loss of ellipsoid zone, pigment epithelium detachment, subretinal fluid, and drusen were the most abundant features in our cohort. Subretinal fluid, intraretinal fluid, hypertransmission, descent of the outer plexiform layer, and pigment epithelium detachment showed highest interreader agreement, while detection and measures of loss of ellipsoid zone and retinal pigment epithelium were more variable. The agreement on the size and location of the respective annotation was more consistent throughout all features.

Conclusions

The interreader agreement depended on the respective OCT-based feature. A selection of reliable features might provide suitable surrogate markers for disease progression and possible treatment effects focusing on different disease stages.

Translational Relevance

This might give opportunities for a more time- and cost-effective patient assessment and improved decision making as well as have implications for clinical trials and training machine learning algorithms.

Digital Technology for AMD Management in the Post-COVID-19 New Normal

PURPOSE

The COVID-19 pandemic has put strain on healthcare systems and the availability and allocation of healthcare manpower, resources and infrastructure. With immediate priorities to protect the health and safety of both patients and healthcare service providers, ophthalmologists globally were advised to defer nonurgent cases, while at the same time managing sight-threatening conditions such as neovascular Age-related Macular Degeneration (AMD). The management of AMD patients both from a monitoring and treatment perspective presents a particular challenge for ophthalmologists. This review looks at how these pressures have encouraged the acceptance and speed of adoption of digitalization.

DESIGN AND METHODS

A literature review was conducted on the use of digital technology during COVID-19 pandemic, and on the transformation of medicine, ophthalmology and AMD screening through digitalization.

RESULTS

In the management of AMD, the implementation of artificial intelligence and "virtual clinics" have provided assistance in screening, diagnosis, monitoring of the progression and the treatment of AMD. In addition, hardware and software developments in home monitoring devices has assisted in self-monitoring approaches.

CONCLUSIONS

Digitalization strategies and developments are currently ongoing and underway to ensure early detection, stability and visual improvement in patients suffering from AMD in this COVID-19 era. This may set a precedence for the post COVID-19 new normal where digital platforms may be routine, standard and expected in healthcare delivery.

Relationship of Topographic Distribution of Geographic Atrophy to Visual Acuity in Nonexudative Age-Related Macular Degeneration

PURPOSE

To investigate the topographic distribution of geographic atrophy (GA) and to identify an anatomic endpoint that correlates with visual acuity (VA) in eyes with GA.

DESIGN

Retrospective analysis of a multicenter, prospective, randomized controlled trial.

PARTICIPANTS

The Age-Related Eye Disease Study participants with GA secondary to nonexudative age-related macular degeneration.

METHODS

We manually delineated GA on 1654 fundus photographs of 365 eyes. We measured GA areas in 9 subfields on the Early Treatment Diabetic Retinopathy Study (ETDRS) grid and correlated them with VA via a mixed-effects model. We determined the optimal diameter for the central zone by varying the diameter from 0 to 10 mm until the highest r² between GA area in the central zone and VA was achieved. We estimated the VA decline rate over 8 years using a linear mixed model.

MAIN OUTCOME MEASURES

Geographic atrophy area in macular subfields and VA.

RESULTS

The percentage of area affected by GA declined as a function of retinal eccentricity. GA area was higher in the temporal than the nasal region (1.30 ± 1.75 mm² vs. 1.10 ± 1.62 mm²; P = 0.005) and in the superior than the inferior region (1.26 ± 1.73 mm² vs. 1.03 ± 1.53 mm²; P < 0.001). Total GA area correlated poorly with VA (r² = 0.07). Among GA areas in 9 subfields, only GA area in the central zone was associated independently with VA (P < 0.001). We determined 1 mm as the optimal diameter for the central zone in which GA area correlated best with VA (r² = 0.45). On average, full GA coverage of the central 1-mm diameter zone corresponded to 34.8 letters' decline in VA. The VA decline rate was comparable between eyes with initial noncentral and central GA before GA covered the entire central 1-mm diameter zone (2.7 letters/year vs. 2.8 letters/year; P = 0.94).

CONCLUSIONS

The prevalence of GA varies significantly across different macular regions. Although total GA area was associated poorly with VA, GA area in the central 1-mm diameter zone was correlated significantly with VA and may serve as a surrogate endpoint in clinical trials.

Artificial Intelligence Algorithms for Analysis of Geographic Atrophy: A Review and Evaluation

Purpose

The purpose of this study was to summarize and evaluate artificial intelligence (AI) algorithms used in geographic atrophy (GA) diagnostic processes (e.g. isolating lesions or disease progression).

Methods

The search strategy and selection of publications were both conducted in accordance with the Preferred of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. PubMed and Web of Science were used to extract literary data. The algorithms were summarized by objective, performance, and scope of coverage of GA diagnosis (e.g. lesion automation and GA progression).

Results

Twenty-seven studies were identified for this review. A total of 18 publications focused on lesion segmentation only, 2 were designed to detect and classify GA, 2 were designed to predict future overall GA progression, 3 focused on prediction of future spatial GA progression, and 2 focused on prediction of visual function in GA. GA-related algorithms reported sensitivities from 0.47 to 0.98, specificities from 0.73 to 0.99, accuracies from 0.42 to 0.995, and Dice coefficients from 0.66 to 0.89.

Conclusions

Current GA-AI publications have a predominant focus on lesion segmentation and a minor focus on classification and progression analysis. AI could be applied to other facets of GA diagnoses, such as understanding the role of hyperfluorescent areas in GA. Using AI for GA has several advantages, including improved diagnostic accuracy and faster processing speeds.

Translational Relevance

AI can be used to quantify GA lesions and therefore allows one to impute visual function and quality-of-life. However, there is a need for the development of reliable and objective models and software to predict the rate of GA progression and to quantify improvements due to interventions.

Merging Information From Infrared and Autofluorescence Fundus Images for Monitoring of Chorioretinal Atrophic Lesions

Purpose

To develop a method for automated detection and progression analysis of chorioretinal atrophic lesions using the combined information of standard infrared (IR) and autofluorescence (AF) fundus images.

Methods

Eighteen eyes (from 16 subjects) with punctate inner choroidopathy were analyzed. Macular IR and blue AF images were acquired in all eyes with a Spectralis HRA+OCT device (Heidelberg Engineering, Heidelberg, Germany). Two clinical experts manually segmented chorioretinal lesions on the AF image. AF images were aligned to the corresponding IR. Two random forest models were trained to classify pixels of lesions, one based on the AF image only, the other based on the aligned IR-AF. The models were validated using a leave-one-out cross-validation and were tested against the manual segmentation to compare their performance. A time series from one eye was identified and used to evaluate the method based on the IR-AF in a case study.

Results

The method based on the AF images correctly classified 95% of the pixels (i.e., in vs. out of the lesion) with a Dice's coefficient of 0.80. The method based on the combined IR-AF correctly classified 96% of the pixels with a Dice's coefficient of 0.84.

Conclusions

The automated segmentation of chorioretinal lesions using IR and AF shows closer alignment to manual segmentation than the same method based on AF only. Merging information from multimodal images improves the automatic and objective segmentation of chorioretinal lesions even when based on a small dataset.

Translational Relevance

Merged information from multimodal images improves segmentation performance of chorioretinal lesions.

Accelerating ophthalmic artificial intelligence research: the role of an open access data repository

PURPOSE OF REVIEW

Artificial intelligence has already provided multiple clinically relevant applications in ophthalmology. Yet, the explosion of nonstandardized reporting of high-performing algorithms are rendered useless without robust and streamlined implementation guidelines. The development of protocols and checklists will accelerate the translation of research publications to impact on patient care.

RECENT FINDINGS

Beyond technological scepticism, we lack uniformity in analysing algorithmic performance generalizability, and benchmarking impacts across clinical settings. No regulatory guardrails have been set to minimize bias or optimize interpretability; no consensus clinical acceptability thresholds or systematized postdeployment monitoring has been set. Moreover, stakeholders with misaligned incentives deepen the landscape complexity especially when it comes to the requisite data integration and harmonization to advance the field. Therefore, despite increasing algorithmic accuracy and commoditization, the infamous 'implementation gap' persists. Open clinical data repositories have been shown to rapidly accelerate research, minimize redundancies and disseminate the expertise and knowledge required to overcome existing barriers. Drawing upon the longstanding success of existing governance frameworks and robust data use and sharing agreements, the ophthalmic community has tremendous opportunity in ushering artificial intelligence into medicine. By collaboratively building a powerful resource of open, anonymized multimodal ophthalmic data, the next generation of clinicians can advance data-driven eye care in unprecedented ways.

SUMMARY

This piece demonstrates that with readily accessible data, immense progress can be achieved clinically and methodologically to realize artificial intelligence's impact on clinical care. Exponentially progressive network effects can be seen by consolidating, curating and distributing data amongst both clinicians and data scientists.