Dataset Information

Machine learning OCT predictors of progression from intermediate age-related macular degeneration to geographic atrophy and vision loss.

ABSTRACT:

Objective

To describe optical coherence tomography (SD-OCT) features, age, gender, and systemic variables that may be used in machine/deep learning studies to identify high-risk patient subpopulations with high risk of progression to geographic atrophy (GA) and visual acuity (VA) loss in the short term.

Design

prospective, longitudinal study.

Subjects

We analyzed imaging data from patients with iAMD (N= 316) enrolled in Age-Related Eye Disease Study 2 (AREDS2) Ancillary SD-OCT with adequate SD-OCT imaging for repeated measures.

Methods

Qualitative and quantitative multimodal variables from the database were derived at each yearly visit over 5 years. Based on statistical analyses developed in the field of cardiology, an algorithm was developed and used to select person-years without GA on colour fundus photography or SD-OCT at baseline. The analysis employed machine learning approaches to generate classification trees. Eyes were stratified as low, average, above average and high risk in 1 or 2 years, based on OCT and demographic features by the risk of GA development or decreased VA by 5+ and 10+ letters.

Main outcome measures

new onset of SD-OCT-determined GA and VA loss.

Results

We identified multiple retinal and subretinal SD-OCT and demographic features from the baseline visit, each of which independently conveyed low to high risk of new-onset GA or VA loss on each of the follow-up visits at 1 or 2 years.

Conclusion

We propose a risk-stratified classification of iAMD based on the combination of OCT-derived retinal features, age, gender and systemic variables for progression to OCT-determined GA and/or VA loss. After external validation, the composite early endpoints may be used as exclusion or inclusion criteria for future clinical studies of iAMD focused on prevention of GA progression or VA loss.

SUBMITTER: Lad E

PROVIDER: S-EPMC9161427 | biostudies-literature | 2022 Jun

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Machine learning OCT predictors of progression from intermediate age-related macular degeneration to geographic atrophy and vision loss.

Lad Eleonora E Sleiman Karim K Banks David L DL Hariharan Sanjay S Clemons Traci T Herrmann Rolf R Dauletbekov Daniyar D Giani Andrea A Chong Victor V Chew Emily Y EY Toth Cynthia A CA

Ophthalmology science 20220601 2

<h4>Objective</h4>To describe optical coherence tomography (SD-OCT) features, age, gender, and systemic variables that may be used in machine/deep learning studies to identify high-risk patient subpopulations with high risk of progression to geographic atrophy (GA) and visual acuity (VA) loss in the short term.<h4>Design</h4>prospective, longitudinal study.<h4>Subjects</h4>We analyzed imaging data from patients with iAMD (N= 316) enrolled in Age-Related Eye Disease Study 2 (AREDS2) Ancillary SD- ...[more]

PMID: 35662803

Similar Datasets

Project description:PurposeIn macula-wide analyses, spectral-domain (SD) optical coherence tomography (OCT) features including drusen volume, hyperreflective foci, and OCT-reflective drusen substructures independently predict geographic atrophy (GA) onset secondary to age-related macular degeneration (AMD). We sought to identify SD OCT features in the location of new GA before its onset.DesignRetrospective study.ParticipantsAge-Related Eye Disease Study 2 Ancillary SD OCT Study participants.MethodsWe analyzed longitudinally captured SD OCT images and color photographs from 488 eyes of 488 participants with intermediate AMD at baseline. Sixty-two eyes with sufficient image quality demonstrated new-onset GA on color photographs during study years 2 through 7. The area of new-onset GA and one size-matched control region in the same eye were segmented separately, and corresponding spatial volumes on registered SD OCT images at the GA incident year and at 2, 3, and 4 years previously were defined. Differences in SD OCT features between paired precursor regions were evaluated through matched-pairs analyses.Main outcome measuresLocalized SD OCT features 2 years before GA onset.ResultsCompared with paired control regions, GA precursor regions at 2, 3, and 4 years before (n = 54, 33, and 25, respectively) showed greater drusen volume (P = 0.01, P = 0.003, and P = 0.003, respectively). At 2 and 3 years before GA onset, they were associated with the presence of hypertransmission (P < 0.001 and P = 0.03, respectively), hyperreflective foci (P < 0.001 and P = 0.045, respectively), OCT-reflective drusen substructures (P = 0.004 and P = 0.03, respectively), and loss or disruption of the photoreceptor zone, ellipsoid zone, and retinal pigment epithelium (RPE, P < 0.001 and P = 0.005-0.045, respectively). At 4 years before GA onset, precursor regions were associated with photoreceptor zone thinning (P = 0.007) and interdigitation zone loss (P = 0.045).ConclusionsEvolution to GA is heralded by early local photoreceptor changes and drusen accumulation, detectable 4 years before GA onset. These precede other anatomic heralds such as RPE changes and drusen substructure emergence detectable 1 to 2 years before GA. This study thus identified earlier end points for GA as potential therapeutic targets in clinical trials.

Project description:ImportanceSensitive outcome measures for disease progression are needed for treatment trials in geographic atrophy (GA) secondary to age-related macular degeneration (AMD).ObjectiveTo quantify photoreceptor degeneration outside regions of GA in eyes with nonexudative AMD, to evaluate its association with future GA progression, and to characterize its spatio-temporal progression.Design, setting, and participantsMonocenter cohort study (Directional Spread in Geographic Atrophy [NCT02051998]) and analysis of data from a normative data study at a tertiary referral center. One hundred fifty-eight eyes of 89 patients with a mean (SD) age of 77.7 (7.1) years, median area of GA of 8.87 mm2 (IQR, 4.09-15.60), and median follow-up of 1.1 years (IQR, 0.52-1.7 years), as well as 93 normal eyes from 93 participants.ExposuresLongitudinal spectral-domain optical coherence tomography (SD-OCT) volume scans (121 B-scans across 30° × 25°) were segmented with a deep-learning pipeline and standardized in a pointwise manner with age-adjusted normal data (z scores). Outer nuclear layer (ONL), photoreceptor inner segment (IS), and outer segment (OS) thickness were quantified along evenly spaced contour lines surrounding GA lesions. Linear mixed models were applied to assess the association between photoreceptor-related imaging features and GA progression rates and characterize the pattern of photoreceptor degeneration over time.Main outcomes and measuresAssociation of ONL thinning with follow-up time (after adjusting for age, retinal topography [z score], and distance to the GA boundary).ResultsThe study included 158 eyes of 89 patients (51 women and 38 men) with a mean (SD) age of 77.7 (7.1) years. The fully automated B-scan segmentation was accurate (dice coefficient, 0.82; 95% CI, 0.80-0.85; compared with manual markings) and revealed a marked interpatient variability in photoreceptor degeneration. The ellipsoid zone (EZ) loss-to-GA boundary distance and OS thickness were prognostic for future progression rates. Outer nuclear layer and IS thinning over time was significant even when adjusting for age and proximity to the GA boundary (estimates of -0.16 μm/y; 95% CI, -0.30 to -0.02; and -0.17 μm/y; 95% CI, -0.26 to -0.09).Conclusions and relevanceDistinct and progressive alterations of photoreceptor laminae (exceeding GA spatially) were detectable and quantifiable. The degree of photoreceptor degeneration outside of regions of retinal pigment epithelium atrophy varied markedly between eyes and was associated with future GA progression. Macula-wide photoreceptor laminae thinning represents a potential candidate end point to monitor treatment effects beyond mere GA lesion size progression.

Project description:PURPOSE:In an eye with geographic atrophy (GA) secondary to age-related macular degeneration, we correlated ex vivo histologic features with findings recorded in vivo using optical coherence tomography (OCT), near-infrared reflectance imaging, and fundus autofluorescence. METHODS:In the left eye of an 86-year-old white woman, in vivo near-infrared reflectance and eye-tracked OCT B-scans at each of 6 clinic visits and a baseline fundus autofluorescence image were correlated with high-resolution histologic images of the preserved donor eye. RESULTS:Clinical imaging showed a small parafoveal multilobular area of GA, subfoveal soft drusen, refractile drusen, hyperreflective lines near the Bruch membrane, subretinal drusenoid deposit (reticular pseudodrusen), and absence of hyperautofluorescent foci at the GA margin. By histology, soft drusen end-stages included avascular fibrosis with highly reflective cholesterol crystals. These accounted for hyperreflective lines near the Bruch membrane in OCT and plaques in near-infrared reflectance imaging. Subretinal drusenoid deposit was thick, continuous, extracellular, extensive outside the fovea, and associated with distinctive retinal pigment epithelium dysmorphia and photoreceptor degeneration. A hyporeflective wedge corresponded to ordered Henle fibers without cellular infiltration. The external limiting membrane descent, which delimits GA, was best visualized in high-quality OCT B-scans. Retinal pigment epithelium and photoreceptor changes at the external limiting membrane descent were consistent with our recent histologic survey of donor eyes. CONCLUSION:This case informs on the extent, topography, and lifecycle of extracellular deposits. High-quality OCT scans are required to reveal all tissue features relevant to age-related macular degeneration progression to GA, especially the external limiting membrane descent. Histologically validated signatures of structural OCT B-scans can serve as references for other imaging modalities.

Project description:PurposeWe sought to determine whether genotype is associated with rate of growth of geographic atrophy (GA) in eyes with age-related macular degeneration (AMD).DesignProspective analysis of participants in a randomized controlled clinical trial.ParticipantsWe included 114 eyes of 114 participants in the Age-Related Eye Disease Study (AREDS).MethodsFundus photographs from AREDS participants with GA from whom a DNA specimen had been obtained and serial photographs had been taken over a minimum of 2 years were evaluated for progression as determined by change in cumulative area of GA. All fundus photographs were scanned, digitized, and centrally graded longitudinally for area of GA. The relationship of GA progression with previously identified genetic variants associated with AMD was assessed.Main outcome measuresGenotype frequencies and change in cumulative area of GA.ResultsThe mean growth rate of GA for the 114 eyes was 1.79 mm(2)/year (range, 0.17-4.76). No association between growth rate and genotype was present for variants in the CFH, C2, C3, APOE, and TLR3 genes. For the single nucleotide polymorphism rs10490924 in LOC387715/ARMS2, there was a significant association of GA growth rate, both adjusted and unadjusted for initial lesion size, with the homozygous risk genotype as compared with the homozygous nonrisk genotype (unadjusted P = 0.002; Bonferroni-corrected P = 0.014) and for allelic association (Bonferroni-corrected P value = 0.011). Analyses of other measures of GA progression (progression to central GA from extrafoveal GA and development of bilateral GA in those initially with unilateral GA) showed no statistically significant association between progression and the LOC387715/ARMS2/HTRA1 genotype.ConclusionsGrowth rates of GA calculated from digitized serial fundus photographs showed no association with variants in the CFH, C2, C3, APOE, or TLR3 genes. There was a nominally significant association with the LOC387715/ARMS2/HTRA1 genotype, although this finding was not supported by analyses of secondary measures of GA progression. Replication in other populations is needed to establish the existence of an association.

Project description:PurposeManually 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.DesignMed-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.ParticipantsOne 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.MethodsMed-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 measuresAccuracy, area under the receiver operating characteristic (ROC) curve, F1 score, sensitivity, and specificity.ResultsIn 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).ConclusionsDespite 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.

Project description:PurposeTo examine disease progression in age-related macular degeneration (AMD) at 2 distinct stages, progression to geographic atrophy (GA) versus GA expansion, by comparison of the risk and protective factors at each stage.DesignPerspective.SubjectsIndividuals at risk of GA or with GA.Main outcome measuresProgression to GA and GA expansion rate.MethodsCritical synthesis of the literature on risk and protective factors, both environmental and genetic, for progression to GA versus GA expansion in AMD.ResultsComparison of the risk and protective factors demonstrates partially overlapping but partially distinct risk and protective factors for progression to GA versus GA expansion. Some factors are shared (i.e., operating in the same direction at both stages), others are not shared, and others seem to operate in different directions at each stage. Risk variants at ARMS2/HTRA1 increase both risk of progression to GA and GA expansion rate, presumably through the same mechanism. By contrast, risk and protective variants at CFH/CFHR alter risk of GA but not GA expansion rate. A risk variant at C3 increases risk of GA but is associated with slower GA expansion. In environmental factors, cigarette smoking is associated with increased risk of GA and faster GA expansion, whereas increased age is associated with the former but not the latter. The Mediterranean diet is associated with decreased progression at both stages, although the food components with the largest contributions seem to differ between the 2 stages. Some phenotypic features, such as reticular pseudodrusen and hyperreflective foci, are associated with increased progression at both stages.ConclusionsAnalysis of the risk and protective factors for progression to GA and GA expansion demonstrates partially overlapping but partially distinct elements at each stage: some are shared, some are relevant to 1 stage only, and some even seem active in opposite directions at each stage. Aside from ARMS2/HTRA1, the overlap between the genetic risk factors for the 2 stages is minimal. This suggests that the biologic mechanisms differ at least partially between the 2 disease stages. This has implications for therapeutic approaches and suggests that treatment aimed at the underlying disease processes may need to be tailored by stage.Financial disclosuresProprietary or commercial disclosure may be found after the references.

Dataset Information

Machine learning OCT predictors of progression from intermediate age-related macular degeneration to geographic atrophy and vision loss.

Objective

Design

Subjects

Methods

Main outcome measures

Results

Conclusion

Publications

Machine learning OCT predictors of progression from intermediate age-related macular degeneration to geographic atrophy and vision loss.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets