Comparison of Multimodal Imaging for the Characterization of Geographic Atrophy.

Purpose: The purpose of this study was to compare the performances of infrared (IR), fundus autofluorescence (FAF), and multicolor (MC) imaging in the characterization of geographic atrophy, with a focus on the possibility to detect incomplete retinal pigmented and outer retinal atrophy (iRORA) on en face imaging. Methods: The ground truth was established by two graders evaluating atrophy on spectral-domain optical coherence tomography (SD-OCT) images. A score for visibility of foveal sparing and margins of atrophy was attributed. Measurement of the atrophic area and the fovea-to-margin distance were performed. Accuracy of detection of foveal sparing was evaluated through comparison with B-scan images ground truth, with/without the inclusion of patients with foveal iRORA. Results: Seventy patients were included in this study. Foveal sparing and atrophy's margins subjective visibility were significantly higher rated on MC images compared to IR and FAF (P < 0.005 and P < 0.001). Agreement with OCT B-scan assessed foveal sparing revealed a significantly higher area under receiver operating characteristic curves (AUROC) for MC images at the analysis performed both with (0.876) and without (0.853) inclusion of patients with foveal iRORA (P < 0.001 and P = 0.006). Quantitative measurements revealed lower atrophy extension (P = 0.026) and fovea-to-margin distance (P = 0.019) with MC imaging. Conclusions: MC imaging performed better at foveal sparing assessment, especially in the setting of foveal iRORA. MC also resulted in higher visibility of atrophy's margins, lower atrophy extension measurements, and lower distance from the fovea to atrophy's margins compared to both FAF and IR. Translational Relevance: MC rated significantly higher in foveal sparing and atrophy detection, higher visibility of atrophy's margins, lower atrophy extension measurements, and lower distance from the fovea to atrophy's margins, compared to FAF and IR.

DEVELOPMENT AND VALIDATION OF AN EXPLAINABLE ARTIFICIAL INTELLIGENCE FRAMEWORK FOR MACULAR DISEASE DIAGNOSIS BASED ON OPTICAL COHERENCE TOMOGRAPHY IMAGES.

PURPOSE: To develop and validate an artificial intelligence framework for identifying multiple retinal lesions at image level and performing an explainable macular disease diagnosis at eye level in optical coherence tomography images. METHODS: A total of 26,815 optical coherence tomography images were collected from 865 eyes, and 9 retinal lesions and 3 macular diseases were labeled by ophthalmologists, including diabetic macular edema and dry/wet age-related macular degeneration. We applied deep learning to classify retinal lesions at image level and random forests to achieve an explainable disease diagnosis at eye level. The performance of the integrated two-stage framework was evaluated and compared with human experts. RESULTS: On testing data set of 2,480 optical coherence tomography images from 80 eyes, the deep learning model achieved an average area under curve of 0.978 (95% confidence interval, 0.971-0.983) for lesion classification. In addition, random forests performed accurate disease diagnosis with a 0% error rate, which achieved the same accuracy as one of the human experts and was better than the other three experts. It also revealed that the detection of specific lesions in the center of macular region had more contribution to macular disease diagnosis. CONCLUSION: The integrated method achieved high accuracy and interpretability in retinal lesion classification and macular disease diagnosis in optical coherence tomography images and could have the potential to facilitate the clinical diagnosis.

Intraocular iron injection induces oxidative stress followed by elements of geographic atrophy and sympathetic ophthalmia.

Iron has been implicated in the pathogenesis of age-related retinal diseases, including age-related macular degeneration (AMD). Previous work showed that intravitreal (IVT) injection of iron induces acute photoreceptor death, lipid peroxidation, and autofluorescence (AF). Herein, we extend this work, finding surprising chronic features of the model: geographic atrophy and sympathetic ophthalmia. We provide new mechanistic insights derived from focal AF in the photoreceptors, quantification of bisretinoids, and localization of carboxyethyl pyrrole, an oxidized adduct of docosahexaenoic acid associated with AMD. In mice given IVT ferric ammonium citrate (FAC), RPE died in patches that slowly expanded at their borders, like human geographic atrophy. There was green AF in the photoreceptor ellipsoid, a mitochondria-rich region, 4 h after injection, followed later by gold AF in rod outer segments, RPE and subretinal myeloid cells. The green AF signature is consistent with flavin adenine dinucleotide, while measured increases in the bisretinoid all-trans-retinal dimer are consistent with the gold AF. FAC induced formation carboxyethyl pyrrole accumulation first in photoreceptors, then in RPE and myeloid cells. Quantitative PCR on neural retina and RPE indicated antioxidant upregulation and inflammation. Unexpectedly, reminiscent of sympathetic ophthalmia, autofluorescent myeloid cells containing abundant iron infiltrated the saline-injected fellow eyes only if the contralateral eye had received IVT FAC. These findings provide mechanistic insights into the potential toxicity caused by AMD-associated retinal iron accumulation. The mouse model will be useful for testing antioxidants, iron chelators, ferroptosis inhibitors, anti-inflammatory medications, and choroidal neovascularization inhibitors.

Quantification of Key Retinal Features in Early and Late Age-Related Macular Degeneration Using Deep Learning.

PURPOSE: We sought to develop and validate a deep learning model for segmentation of 13 features associated with neovascular and atrophic age-related macular degeneration (AMD). DESIGN: Development and validation of a deep-learning model for feature segmentation. METHODS: Data for model development were obtained from 307 optical coherence tomography volumes. Eight experienced graders manually delineated all abnormalities in 2712 B-scans. A deep neural network was trained with these data to perform voxel-level segmentation of the 13 most common abnormalities (features). For evaluation, 112 B-scans from 112 patients with a diagnosis of neovascular AMD were annotated by 4 independent observers. The main outcome measures were Dice score, intraclass correlation coefficient, and free-response receiver operating characteristic curve. RESULTS: On 11 of 13 features, the model obtained a mean Dice score of 0.63 ± 0.15, compared with 0.61 ± 0.17 for the observers. The mean intraclass correlation coefficient for the model was 0.66 ± 0.22, compared with 0.62 ± 0.21 for the observers. Two features were not evaluated quantitatively because of a lack of data. Free-response receiver operating characteristic analysis demonstrated that the model scored similar or higher sensitivity per false positives compared with the observers. CONCLUSIONS: The quality of the automatic segmentation matches that of experienced graders for most features, exceeding human performance for some features. The quantified parameters provided by the model can be used in the current clinical routine and open possibilities for further research into treatment response outside clinical trials.

Age-related macular degeneration (AMD): More than meets the eye. The role of multimodal imaging in today's management of AMD-A review.

Multimodal imaging (MMI) allows a more granular grading of age-related macular degeneration (AMD) disease severity, with many novel risk factors having been recently identified. With this imaging information, we are better able to counsel our patients with more accurate and individualized progression scenarios. MMI also allows identification of anatomical features that increase our understanding of disease processes involved in progression to late AMD. Treatment protocols for neovascular AMD (nAMD) depend largely on the optical coherence tomography (OCT) appearance to determine disease activity, which allows us to individualize treatment. In geographic atrophy (GA), new intervention trials require the ability to define the extent of GA, so that GA growth rate can be determined. This is achieved through fundus autofluorescence (FAF) imaging, which allows greater accuracy of border identification, as well as revealing FAF patterns predictive of growth rates. As we strive to bring interventions earlier in the disease course, OCT imaging provides an ability to identify the first signs of atrophy, which may serve as novel surrogate biomarkers for GA, thereby facilitating trials. In the future, the use of artificial intelligence (AI) to automatically identify relevant features on MMI could further enhance our ability to determine disease severity, predict progression and assist in identifying disease activity parameters to support clinical decision making when treating nAMD. Newer developments may allow frequent, remote capturing of images, reducing clinic visits, detecting progression and monitoring neovascular activity in-between clinic visits. Being aware of these new imaging insights in AMD, greatly enhance our clinical management of AMD.

NOVEL METHOD FOR IMAGE AVERAGING OF OPTICAL COHERENCE TOMOGRAPHY ANGIOGRAPHY IMAGES.

PURPOSE: To develop a method of averaging optical coherence tomography (OCT) angiography to improve visualization of choriocapillaris structure. METHODS: A stack of OCT angiographic data from vascular layers were placed into the red-green-blue channels of a conventional digital color image. The superficial plexus was placed in the blue channel, choriocapillaris in the green, and deep vascular plexus in the red channel. The red-green-blue images derived from nine separate OCT angiographic scans were registered using an automatic registration sequence and the images were averaged. The averaged red-green-blue image was then split into the three averaged component layers. The technique is flexible and any vascular layer, such as macular neovascularization, can be used as well. RESULTS: The utility of the imaging method was demonstrated by showing the imaging of two different diseases. A patient with a history of familial amyloidosis, hypertension, kidney failure, kidney transplantation, and prednisone use, followed by central serous chorioretinopathy treated by photodynamic therapy. She had alterations in retinal pigment epithelial pigmentation and widespread abnormalities of autofluorescence. She showed remarkably decreased vascular density and vessel configuration of her choriocapillaris. A patient with pseudoxanthoma elasticum with subretinal drusenoid deposits at an early age also showed marked decreased choriocapillaris density and vascular configuration. These findings were compared with healthy controls of similar age with no abnormalities. CONCLUSION: The detailed method is capable of averaging choriocapillaris OCT angiographic images using a simple automatic method. Image averaging offers opportunity to improve the noisy OCT angiographic images such that actual vascular structure is visible.

Geographic Atrophy and OCT Angiography: Descriptive Study and Correlation With Autofluorescence.

BACKGROUND AND OBJECTIVE: Geographic atrophy (GA) involves the progressive loss of retinal pigment epithelium (RPE), photoreceptors, and choriocapillaris (CC). CC flow within a GA area is severely impaired in patients with atrophic age-related macular degeneration. The aim of this study was to compare GA area measured on optical coherence tomography angiography (OCTA) (CC nonperfusion area) and on fundus autofluorescence (FAF). PATIENTS AND METHODS: In this prospective, observational, cross-sectional study, OCTA and FAF were performed in patients with GA. On OCTA (CC segmentation), the CC nonperfusion area was measured manually using calipers. On FAF, GA was manually delimited, and the total surface was obtained using Region Finder software. The primary endpoint was to compare the CC nonperfusion area measured on OCTA and on the gold standard method (FAF). RESULTS: Forty eyes of 34 patients with a mean age of 82.63 years ± 9.21 years (range: 66 years to 100 years) were included. The mean GA area measured on FAF and OCTA was, respectively, 2.184 ± 3.045 mm2 and 2.349 ± 3.237 mm2 (P = .035). The mean difference was 0.165 ± 0.290 mm2. A strong correlation was found between both measurements (r = 0.97; P < .0001; confidence interval: 0.98-0.99), although the CC nonperfusion area was larger than the GA area on FAF (P = .035). CONCLUSIONS: In this study, the authors showed that in GA, the CC nonperfusion area correlates linearly with the GA area assessed by FAF. Also, the CC nonperfusion area is larger than the GA area measured by FAF, suggesting that CC degeneration could occur before RPE degeneration in GA. [Ophthalmic Surg Lasers Imaging Retina. 2019;50:e222-e228.].

Quantifying Choriocapillaris Flow Voids in Patients With Geographic Atrophy Using Swept-Source OCT Angiography.

BACKGROUND AND OBJECTIVE: To compare choriocapillaris (CC) flow voids (FVs) throughout the macula in patients with age-related macular degeneration (AMD) and geographic atrophy (GA) to age-similar controls using swept-source optical coherence tomography angiography (SS-OCTA). PATIENTS AND METHODS: In this cross-sectional study, 12 subjects with GA secondary to nonexudative AMD and 12 age-similar controls participated. SS-OCTA was performed using a 6 mm × 6 mm scanning pattern. CC FVs were calculated using a one-standard deviation thresholding method developed from a normal database. RESULTS: CC FVs were significantly increased in patients with GA compared with age-similar controls (P < .001). FVs within 2° of GA were significantly increased compared with the area outside 2° (P < .001). FVs beyond 2° of GA were significantly increased compared with age-similar controls (P < .001). CONCLUSIONS: FV analysis of in vivo CC images revealed diffuse CC perfusion deficits throughout the macular region in subjects with GA secondary to nonexudative AMD. [Ophthalmic Surg Lasers Imaging Retina. 2019;50:e229-e235.].

Suprachoroidal Adipose Tissue-Derived Mesenchymal Stem Cell Implantation in Patients with Dry-Type Age-Related Macular Degeneration and Stargardt's Macular Dystrophy: 6-Month Follow-Up Results of a Phase 2 Study.

This prospective clinical case series aimed to investigate the safety and efficacy of suprachoroidal adipose tissue-derived mesenchymal stem cell (ADMSC) implantation in patients with dry-type age-related macular degeneration (AMD) and Stargardt's macular dystrophy (SMD). This study included four patients with advanced-stage dry-type AMD and four patients with SMD who underwent suprachoroidal implantation of ADMSCs. The best-corrected visual acuity (BCVA) in the study was 20/200. The worse eye of the patient was operated on. Patients were evaluated on the first day, first week, and first, third, and sixth months postoperatively. BCVA, anterior segment and fundus examination, color photography, fundus autofluorescence, optical coherence tomography, and visual field examination were carried out at each visit. Fundus fluorescein angiography and multifocal electroretinography (mf-ERG) recordings were performed at the end of the first, third, and sixth months and anytime if necessary during the follow-up. All eight patients completed the sixth month follow-up. None of them had any systemic or ocular complications. All of the eight patients experienced visual acuity improvement, visual field improvement, and improvement in mf-ERG recordings. Stem cell treatment with suprachoroidal implantation of ADMSCs seems to be safe and effective in the treatment of dry-type AMD and SMD.

The Role of New Imaging Methods in Managing Age-Related Macular Degeneration.

The use of imaging for age-related macular degeneration (AMD) depends on how it benefits clinical management and on reimbursement. The latter should relate to the former. This review assesses how different forms of AMD can be imaged and what information this provides. For nonneovascular AMD high-resolution optical coherence tomography (OCT), autofluorescence, and near infrared imaging can identify the type of drusen, such as reticular pseudodrusen, which influences prognosis, and the amount of atrophy, for which phase 3 trials are underway. Clarifying the correct diagnosis for late-onset Stargardt and macular telangiectasia, if treatment becomes available, will be especially important. Choroidal thickness can be measured and changes with anti‒vascular endothelial growth factor treatment, but how this influences management is less clear. The finding of a thick choroid may alter the diagnosis to pachychoroid neovasculopathy, which may have a different treatment response. Peripheral retinal changes are commonly found on ultrawide-field imaging but their importance is not yet determined. The mainstay of imaging is OCT, which can detect neovascular AMD by detecting thickening and be used for follow-up, as the presence or absence of thickening is the main determinant of treatment. Higher resolution systems and now OCT angiography are able to distinguish neovascular type, especially type 2 choroidal neovascularization but also polypoidal choroidal vasculopathy and retinal angiomatous proliferation. Fundus fluorescein and indocyanine green angiographies still have a role, although that partly depends on whether photodynamic therapy is being considered. Automated image analysis and machine learning will be increasingly important in supporting clinician decisions.

A view of the current and future role of optical coherence tomography in the management of age-related macular degeneration.

Optical coherence tomography (OCT) has become an established diagnostic technology in the clinical management of age-related macular degeneration (AMD). OCT is being used for primary diagnosis, evaluation of therapeutic efficacy, and long-term monitoring. Computer-based advances in image analysis provide complementary imaging tools such as OCT angiography, further novel automated analysis methods as well as feature detection and prediction of prognosis in disease and therapy by machine learning. In early AMD, pathognomonic features such as drusen, pseudodrusen, and abnormalities of the retinal pigment epithelium (RPE) can be imaged in a qualitative and quantitative way to identify early signs of disease activity and define the risk of progression. In advanced AMD, disease activity can be monitored clearly by qualitative and quantified analyses of fluid pooling, such as intraretinal cystoid fluid, subretinal fluid, and pigment epithelial detachment (PED). Moreover, machine learning methods detect a large spectrum of new biomarkers. Evaluation of treatment efficacy and definition of optimal therapeutic regimens are an important aim in managing neovascular AMD. In atrophic AMD hallmarked by geographic atrophy (GA), advanced spectral domain (SD)-OCT imaging largely replaces conventional fundus autofluorescence (FAF) as it adds insight into the condition of the neurosensory layers and associated alterations at the level of the RPE and choroid. Exploration of imaging features by computerized methods has just begun but has already opened relevant and reliable horizons for the optimal use of OCT imaging for individualized and population-based management of AMD-the leading retinal epidemic of modern times.

ASSOCIATION BETWEEN THE VITREOMACULAR INTERFACE AND OPTICAL COHERENCE TOMOGRAPHY CHARACTERISTICS IN WET AGE-RELATED MACULAR DEGENERATION.

PURPOSE: To study the effect of the vitreomacular interface on various wet age-related macular degeneration (AMD) characteristics including the size and type of choroidal neovascularization (CNV), choroidal thickness, and activity of the CNV. METHODS: This was a retrospective observational cross-sectional study. The study included 43 patients (51 eyes) with treatment-naive age-related macular degeneration. Twenty-six patients with wet AMD in one eye and dry AMD in the other eye were included in a paired-eye analysis. Patients underwent optical coherence tomography examination using Heidelberg Spectralis (spectral domain optical coherence tomography) at presentation to determine the type of CNV and the vitreomacular status. In addition, various parameters were measured including the choroidal thickness and horizontal width and vertical height measurements of the CNV. RESULTS: There was no correlation between the height, width, activity or type of the CNV, and the presence or absence of vitreomacular adhesion. The mean choroidal thickness (using enhanced depth imaging) in cases with vitreomacular adhesion was 272.57 µm compared with 197.32 µm in cases with no vitreomacular adhesion, a statistically significant difference (P = 0.003). In the paired-eye study (21 patients), there was no significant difference between the eyes with wet AMD and dry AMD with regard to vitreomacular status or the choroidal thickness. In a subgroup analysis, patients with Type 1 CNV had a significantly higher percentage of vitreomacular adhesion compared with the other eye with dry AMD (P = 0.034). CONCLUSION: In conclusion, the vitreomacular interface does seem to be associated with an increased choroidal thickness in cases of wet AMD. Furthermore, the association between the vitreomacular interface and wet AMD is more significant for Type 1 CNV.

Prevalence and quantification of geographic atrophy associated with newly diagnosed and treatment-naïve exudative age-related macular degeneration.

OBJECTIVE: To identify and quantify geographic atrophy (GA) associated with neovascular age-related macular degeneration (AMD) at initial presentation using a fundus autofluorescence (FAF) semi-automated software and to correlate the results with demographic and clinical data. DESIGN: Retrospective, observational study. METHODS: The study population consisted of treatment-naïve patients with newly diagnosed neovascular AMD. Best-corrected visual acuity, fundus photographs, infrared reflectance, FAF and spectral-domain optical coherence tomography were performed, associated with fluorescein and indocyanine green angiographies. Identification of GA was independently performed by three readers. Quantification of atrophy areas was done using RegionFinder Software (RFA), a semi-automated software embedded in Spectralis device (Heidelberg Engineering, Germany). RESULTS: We included 206 eyes of 173 consecutive patients (72% female, mean age: 79.7±9.1 years). Type I choroidal neovascularisation (CNV) was observed in 44.2% of eyes, type II CNV was observed in 20.9% and mixed CNV lesion was observed in 11.7%. Polypoidal choroidal vasculopathy was diagnosed in 7.7% and type III CNV was diagnosed in 15.5%. Analysis of FAF frames showed that GA was associated with nAMD in 46/206 eyes (22.3%). Taking into account data both from Region Finder and multimodal imaging, our results suggest that GA was present in 24.3% of eyes newly diagnosed with exudative AMD. Mean size of GA was 1.23±1.76 mm2 (range 0.03-7.39). CONCLUSION: GA is associated with nAMD in 1/4 of cases at initial presentation. Combined imaging, including RFA is an effective tool to identify and quantify GA at diagnosis.