scholarly journals Deep learning segmentation of hyperautofluorescent fleck lesions in Stargardt disease

2020 ◽  
Vol 10 (1) ◽  
Author(s):  
Jason Charng ◽  
Di Xiao ◽  
Maryam Mehdizadeh ◽  
Mary S. Attia ◽  
Sukanya Arunachalam ◽  
...  
Keyword(s):  
Deep Learning ◽  
Vision Loss ◽  
Fundus Autofluorescence ◽  
Retinal Disease ◽  
Diagnostic Feature ◽  
Stargardt Disease ◽  
Inherited Retinal Disease ◽  
Trial Endpoint ◽  
Disease Testing

Abstract Stargardt disease is one of the most common forms of inherited retinal disease and leads to permanent vision loss. A diagnostic feature of the disease is retinal flecks, which appear hyperautofluorescent in fundus autofluorescence (FAF) imaging. The size and number of these flecks increase with disease progression. Manual segmentation of flecks allows monitoring of disease, but is time-consuming. Herein, we have developed and validated a deep learning approach for segmenting these Stargardt flecks (1750 training and 100 validation FAF patches from 37 eyes with Stargardt disease). Testing was done in 10 separate Stargardt FAF images and we observed a good overall agreement between manual and deep learning in both fleck count and fleck area. Longitudinal data were available in both eyes from 6 patients (average total follow-up time 4.2 years), with both manual and deep learning segmentation performed on all (n = 82) images. Both methods detected a similar upward trend in fleck number and area over time. In conclusion, we demonstrated the feasibility of utilizing deep learning to segment and quantify FAF lesions, laying the foundation for future studies using fleck parameters as a trial endpoint.

scholarly journals Deep Learning-Based Classification of Inherited Retinal Diseases Using Fundus Autofluorescence

2020 ◽  
Vol 9 (10) ◽  
pp. 3303
Author(s):  
Alexandra Miere ◽  
Thomas Le Meur ◽  
Karen Bitton ◽  
Carlotta Pallone ◽  
Oudy Semoun ◽  
...  
Keyword(s):  
Deep Learning ◽  
Learning Algorithm ◽  
Fundus Autofluorescence ◽  
Area Under The Curve ◽  
Retinal Disease ◽  
Relevant Information ◽  
Retinal Diseases ◽  
Autofluorescence Imaging ◽  
Comparable Group ◽  
Inherited Retinal Disease

Background. In recent years, deep learning has been increasingly applied to a vast array of ophthalmological diseases. Inherited retinal diseases (IRD) are rare genetic conditions with a distinctive phenotype on fundus autofluorescence imaging (FAF). Our purpose was to automatically classify different IRDs by means of FAF images using a deep learning algorithm. Methods. In this study, FAF images of patients with retinitis pigmentosa (RP), Best disease (BD), Stargardt disease (STGD), as well as a healthy comparable group were used to train a multilayer deep convolutional neural network (CNN) to differentiate FAF images between each type of IRD and normal FAF. The CNN was trained and validated with 389 FAF images. Established augmentation techniques were used. An Adam optimizer was used for training. For subsequent testing, the built classifiers were then tested with 94 untrained FAF images. Results. For the inherited retinal disease classifiers, global accuracy was 0.95. The precision-recall area under the curve (PRC-AUC) averaged 0.988 for BD, 0.999 for RP, 0.996 for STGD, and 0.989 for healthy controls. Conclusions. This study describes the use of a deep learning-based algorithm to automatically detect and classify inherited retinal disease in FAF. Hereby, the created classifiers showed excellent results. With further developments, this model may be a diagnostic tool and may give relevant information for future therapeutic approaches.

Arcuate Vitelliform detachment with morning glory disc maculopathy: OCT findings of evolution and resolution after early surgery

2021 ◽  
Vol 14 (7) ◽  
pp. e244329
Author(s):  
Deependra Vikram Singh ◽  
Yog Sharma ◽  
Raja Rami Reddy ◽  
Ajay Sharma
Keyword(s):  
Visual Acuity ◽  
Vision Loss ◽  
Fundus Autofluorescence ◽  
Internal Limiting Membrane ◽  
Morning Glory ◽  
Internal Limiting Membrane Peeling ◽  
Excellent Outcome ◽  
Early Vitrectomy ◽  
Snellen Visual Acuity

Morning glory disc (MGD) is known to develop secondary maculopathy and vision loss. We followed a 7-year-old girl with MGD in right eye from 2010 to 2021. Her best-corrected Snellen visual acuity (BCVA) was 20/20 in both eyes till 2017. She presented with reduced vision in right eye with BCVA of 20/80 in 2018. Optical Coherence Tomography (OCT) revealed maculopathy related to MGD in right eye and arcuate Vitelliform neurosensory detachment (V-NSD) temporal to fovea. She underwent 25G vitrectomy with internal limiting membrane peeling. Resolution of retinoschisis and V-NSD was analysed by OCT and fundus autofluorescence (FAF) postoperatively. Arcuate V-NSD was hyperfluorescent on FAF and concentric to optic disc. It resolved slowly with BCVA improving to 20/20 at 18 and 30 months postoperatively. We report excellent outcome following early vitrectomy for MGD-related maculopathy and recommend serial follow-up and considering early vitrectomy whenever traction and BCVA worsens. We also describe arcuate V-NSD with MGD.

scholarly journals Estimation of impact of RPE65-mediated inherited retinal disease on quality of life and the potential benefits of gene therapy

2019 ◽  
Vol 103 (11) ◽  
pp. 1610-1614 ◽  
Author(s):  
Andrew Lloyd ◽  
Natalia Piglowska ◽  
Thomas Ciulla ◽  
Sarah Pitluck ◽  
Scott Johnson ◽  
...  
Keyword(s):  
Quality Of Life ◽  
Vision Loss ◽  
Health Utility Index ◽  
Retinal Disease ◽  
Health State ◽  
Health States ◽  
Substantial Impact ◽  
Hand Motion ◽  
Inherited Retinal Disease

Background/aimsIn rare diseases, health-related quality of life (HRQL) data can be difficult to capture. Given the ultrarare nature of RPE65-mediated inherited retinal disease (IRD), it was not feasible to recruit a patient sample and collect HRQL data prospectively. The objectives of this study were to develop health state descriptions of RPE65-mediated IRD, and to estimate associated patient utilities.MethodsVignette descriptions of IRD states were developed and then assessed to elicit utilities. The vignettes ranged from moderate vision loss through to hand motion to no light perception (NLP). Six retina specialists with additional expertise in IRDs provided a proxy valuation of the vignettes using generic measures of health—the 5-level version of EQ-5D-5L and Health Utility Index 3 (HUI3). The data were then scored using standard methods for each instrument.ResultsWeights from both HRQL measures revealed a large decline in scores with vision loss. The EQ-5D-5L weights ranged from 0.709 for moderate vision loss to 0.152 for hand motion to NLP. The HUI3 weights ranged from 0.519 to − 0.039, respectively. A decline was seen on both measures, and the degree of decline from moderate vision loss to NLP was identical on both (−0.56).ConclusionThis is the first study to report HRQL weights (or utilities) for health states describing different levels of vision loss in patients with IRD, specifically those with RPE65-mediated disease. The parallel decline in scores from the EQ-5D and HUI3 corroborates the substantial impact of progressive vision loss on HRQL.

scholarly journals Absence of Genotype/Phenotype Correlations Requires Molecular Diagnostic to Ascertain Stargardt and Stargardt-Like Swiss Patients

Genes ◽  
2021 ◽  
Vol 12 (6) ◽  
pp. 812
Author(s):  
Virginie M.M. Buhler ◽  
Lieselotte Berger ◽  
André Schaller ◽  
Martin S. Zinkernagel ◽  
Sebastian Wolf ◽  
...  
Keyword(s):  
Clinical Symptoms ◽  
Fundus Autofluorescence ◽  
Molecular Diagnostic ◽  
Stargardt Disease ◽  
Vitamin A Supplementation ◽  
Clinical Phenotypes ◽  
Number Of Patients ◽  
Increased Risk ◽  
Swiss Population

We genetically characterized 22 Swiss patients who had been diagnosed with Stargardt disease after clinical examination. We identified in 11 patients (50%) pathogenic bi-allelic ABCA4 variants, c.1760+2T>C and c.4496T>C being novel. The dominantly inherited pathogenic ELOVL4 c.810C>G p.(Tyr270*) and PRPH2-c.422A>G p.(Tyr141Cys) variants were identified in eight (36%) and three patients (14%), respectively. All patients harboring the ELOVL4 c.810C>G p.(Tyr270*) variant originated from the same small Swiss area, identifying a founder mutation. In the ABCA4 and ELOVL4 cohorts, the clinical phenotypes of “flecks”, “atrophy”, and “bull”s eye like” were observed by fundus examination. In the small number of patients harboring the pathogenic PRPH2 variant, we could observe both “flecks” and “atrophy” clinical phenotypes. The onset of disease, progression of visual acuity and clinical symptoms, inheritance patterns, fundus autofluorescence, and optical coherence tomography did not allow discrimination between the genetically heterogeneous Stargardt patients. The genetic heterogeneity observed in the relatively small Swiss population should prompt systematic genetic testing of clinically diagnosed Stargardt patients. The resulting molecular diagnostic is required to prevent potentially harmful vitamin A supplementation, to provide genetic counseling with respect to inheritance, and to schedule appropriate follow-up visits in the presence of increased risk of choroidal neovascularization.

scholarly journals Deep Learning to Distinguish ABCA4-Related Stargardt Disease from PRPH2-Related Pseudo-Stargardt Pattern Dystrophy

2021 ◽  
Vol 10 (24) ◽  
pp. 5742
Author(s):  
Alexandra Miere ◽  
Olivia Zambrowski ◽  
Arthur Kessler ◽  
Carl-Joe Mehanna ◽  
Carlotta Pallone ◽  
...  
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Good Accuracy ◽  
Fundus Autofluorescence ◽  
Retinal Diseases ◽  
Proof Of Concept ◽  
Stargardt Disease ◽  
Test Dataset ◽  
Pattern Dystrophy ◽  
Learning Classifier

(1) Background: Recessive Stargardt disease (STGD1) and multifocal pattern dystrophy simulating Stargardt disease (“pseudo-Stargardt pattern dystrophy”, PSPD) share phenotypic similitudes, leading to a difficult clinical diagnosis. Our aim was to assess whether a deep learning classifier pretrained on fundus autofluorescence (FAF) images can assist in distinguishing ABCA4-related STGD1 from the PRPH2/RDS-related PSPD and to compare the performance with that of retinal specialists. (2) Methods: We trained a convolutional neural network (CNN) using 729 FAF images from normal patients or patients with inherited retinal diseases (IRDs). Transfer learning was then used to update the weights of a ResNet50V2 used to classify the 370 FAF images into STGD1 and PSPD. Retina specialists evaluated the same dataset. The performance of the CNN and that of retina specialists were compared in terms of accuracy, sensitivity, and precision. (3) Results: The CNN accuracy on the test dataset of 111 images was 0.882. The AUROC was 0.890, the precision was 0.883 and the sensitivity was 0.883. The accuracy for retina experts averaged 0.816, whereas for retina fellows it averaged 0.724. (4) Conclusions: This proof-of-concept study demonstrates that, even with small databases, a pretrained CNN is able to distinguish between STGD1 and PSPD with good accuracy.

scholarly journals Recent advances in the management of dry age-related macular degeneration: A review

F1000Research ◽  
2017 ◽  
Vol 6 ◽  
pp. 245 ◽  
Author(s):  
Francesco Bandello ◽  
Riccardo Sacconi ◽  
Lea Querques ◽  
Eleonora Corbelli ◽  
Maria Vittoria Cicinelli ◽  
...  
Keyword(s):  
Macular Degeneration ◽  
Vision Loss ◽  
Morphological Changes ◽  
Pigment Epithelium ◽  
Fundus Autofluorescence ◽  
Retinal Pigment ◽  
Age Related Macular Degeneration ◽  
Choroidal Blood Flow ◽  
Age Related

Age-related macular degeneration (AMD), the most important cause of vision loss in elderly people, is a degenerative disorder of the central retina with a multifactorial etiopathology. AMD is classified in dry AMD (d-AMD) or neovascular AMD depending on the presence of choroidal neovascularization. Currently, no therapy is approved for geographic atrophy, the late form of d-AMD, because no treatment can restore the damage of retinal pigment epithelium (RPE) or photoreceptors. For this reason, all treatment approaches in d-AMD are only likely to prevent and slow down the progression of existing atrophy. This review focuses on the management of d-AMD and especially on current data about potential targets for therapies evaluated in clinical trials. Numerous examinations are available in clinics to monitor morphological changes in the retina, RPE and choroid of d-AMD patients. Fundus autofluorescence and optical coherence tomography (OCT) are considered the most useful tools in the diagnosis and follow-up of d-AMD alterations, including the monitoring of atrophy area progression. Instead, OCT-angiography is a novel imaging tool that may add further information in patients affected by d-AMD. Several pathways, including oxidative stress, deposits of lipofuscin, chronic inflammation and choroidal blood flow insufficiency, seem to play an important role in the pathogenesis of d-AMD and represent possible targets for new therapies. A great number of treatments for d-AMD are under investigation with promising results in preliminary studies. However, only few of these drugs will enter the market, offering a therapeutic chance to patients affected by the dry form of AMD and help them to preserve a good visual acuity. Further studies with a long-term follow-up would be important to test the real safety and efficacy of drugs under investigation.

Sign in / Sign up

Export Citation Format

Share Document