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.
To report patients with age-related macular degeneration and atypical central retinal pigment epithelium (RPE) defects not attributable to geographic atrophy (GA) or RPE-tears with overlying preserved photoreceptor layers. Multimodal imaging case-series evaluating the course of atypical RPE- defects in patients with AMD using Color fundus images, Optical coherence tomography (OCT), OCT-Angiography, fundus autofluorescence (FAF) and fluorescein-angiography (FA). Ten patients were identified. Three patients had a prior RPE-rip and were excluded. Seven patients with a mean follow-up period of 47 ± 38 months after the occurrence of the RPE-defect were included (age range 71–87 years). Mean distance Best corrected visual acuity (BCVA) at initial presentation was 0.36 ± 0.29logMAR and at last follow-up visit 0.51 ± 0.43logMAR. Patients presented with clinically apparent GA on funduscopy and FAF, but preserved photoreceptor layers on optical coherence tomography (OCT). On FA there was early hyperfluorescence and late pooling visible. Over time, migration of RPE/drusenoid material right above the Bruch’s membrane with concomitant decrease of hypoautofluorescence was detectable in 4 cases. An enlargement of the RPE-defect was apparent in the remaining 3 cases. The majority (n = 4) showed a drusenoid pigment epithelium detachment (PED) preceding the lesion. Beside GA and characteristic RPE-tears, another atypical form of RPE-defect with overlying preserved photoreceptor layers are found in AMD. This so far disregarded subgroup of patients present with reasonable visual function and long-term survival of photoreceptors layers. Repair mechanisms such as ingrowth of RPE/drusenoid material and persistent subretinal fluid (SRF), but also a RPE-independent visual cycle for cone photopigment within the neurosensory retina may contribute to their favorable course.
18 March 2019 in European Journal of Ophthalmology
To evaluate the efficacy of switching from intravitreal ranibizumab to intravitreal aflibercept in choroidal neovascularization secondary to angioid streaks. Multicenter retrospective interventional case series. Patients previously treated with intravitreal ranibizumab with at least 12-month follow-up (M12) after switching (M0) to intravitreal aflibercept. Switch to intravitreal aflibercept was decided in cases of refractory or recurrent choroidal neovascularization. Primary endpoint: Change of best-corrected visual acuity using the Early Treatment Diabetic Retinopathy Study letters. Secondary endpoints: Mean change of central macular thickness, absence of intraretinal/subretinal fluid on spectral domain optical coherence tomography and the percentage of eyes with absence of leakage on fluorescein angiography. Fourteen eyes of 13 patients were included. Mean best-corrected visual acuity was 65.0 ± 21.03 letters at M0 and 63.5 ± 17.30 letters at M12 (p = 0.5). Secondary endpoints: Mean central macular thickness was 344 ± 194.65 µm at M0 and 268 ± 79.97 µm at M12 (p = 0.008). Absence of intraretinal/subretinal fluid was observed in 71%. Fluorescein angiography (nine eyes) showed absence of leakage in 77% (seven eyes). Switching from intravitreal ranibizumab to intravitreal aflibercept represents a therapeutic option in patients with refractory or recurrent choroidal neovascularization secondary to angioid streaks.
Purpose: To evaluate the sensitivity and specificity of ultrawide-field fundus photography (UWF-FP) for the detection and classification of sickle cell retinopathy (SCR) by ophthalmologists with varying degrees of expertise in retinal disease. Methods: Patients presenting with sickle cell disease (SCD) in the Créteil University Eye Clinic, having undergone UWF-FP and ultrawide-field fluorescein angiography (UWF-FA) on the same day, were retrospectively included. Eyes with previous retinal photocoagulation were excluded. SCR was graded independently by UWF-FP and UWF-FA using Goldberg classification by two ophthalmologists with varying expertise levels. Results: Sixty-six eyes of 33 patients were included in the study. The sensitivity of UWF-FP for the detection of proliferative SCR was 100%, (95% confidence interval [CI95%] 76.8–100) for the retinal specialist and 100% (CI95% 71.5–100) for the ophthalmology resident. The specificity of UWF-FP for the detection of proliferative SCR was 100% (CI95% 92.7–100) for the retinal specialist and 98.1% (CI95% 89.7–100) for the ophthalmology resident. Conclusions: UWF-FP is a valuable exam for proliferative SCR screening, with excellent sensitivity and specificity and a good inter-grader agreement for ophthalmologists with various degree of skills, and is easy to use in a real-life setting.
We evaluated the spectrum of choriocapillaris (CC) abnormalities in the fellow eyes of unilateral exudative age-related macular degeneration (AMD) patients using swept-source optical coherence tomography angiography (SS-OCTA). Fellow eyes of unilateral exudative AMD patients were prospectively included between May 2018 and October 2018. Patients underwent a multimodal imaging including a SS-OCTA. Demographics and clinical findings were analyzed. The estimated prevalence of macular neovascularization (MNV) was computed. Number and size of flow deficits (FDs) and percentage of flow deficits (FD%) were computed on the compensated CC flow images with the Fiji software. We included 97 eyes of 97 patients (mean age was 80 ± 7.66 years, 39 males, 58 females). The prevalence of MNV in the studied eyes was 8.25% (8/97 eyes). In the 89 non-neovascular eyes, FD% averaged 45.84% ± 11.63%, with a corresponding total area of FDs of 4.19 ± 1.12 mm2. There was a higher prevalence of drusenoid pigment epithelial detachment in eyes with subclinical neovascularization (p = 0.021). Fellow eyes with unilateral exudative AMD encompassed a series of CC abnormalities, from FDs of the aging CC to subclinical non-exudative MNV.
1 December 2016 in Ophthalmic Surgery, Lasers and Imaging Retina
HRCVs are more clearly distinguishable than other choroidal vessels on MC imaging in GA. HRCV identification is more frequent in eyes with larger areas of atrophy and reduced CT, and thus possibly represent a maker of more advanced GA. [Ophthalmic Surg Lasers Imaging Retina. 2016;47:1106-1114.].