Search

Abstract This study aims to propose a generative deep learning model (GDLM) based on a variational autoencoder that predicts macular optical coherence tomography (OCT) images following full-thickness macular hole (FTMH) surgery and evaluate its clinical accuracy. Preoperative and 6-month postoperative swept-source OCT data were collected from 150 patients with successfully closed FTMH using 6 × 6 mm2 macular volume scan datasets. Randomly selected and augmented 120,000 training and 5000 validation pairs of OCT images were used to train the GDLM. We assessed the accuracy and F1 score of concordance for neurosensory retinal areas, performed Bland–Altman analysis of foveolar height (FH) and mean foveal thickness (MFT), and predicted postoperative external limiting membrane (ELM) and ellipsoid zone (EZ) restoration accuracy between artificial intelligence (AI)-OCT and ground truth (GT)-OCT images. Accuracy and F1 scores were 94.7% and 0.891, respectively. Average FH (228.2 vs. 233.4 μm, P = 0.587) and MFT (271.4 vs. 273.3 μm, P = 0.819) were similar between AI- and GT-OCT images, within 30.0% differences of 95% limits of agreement. ELM and EZ recovery prediction accuracy was 88.0% and 92.0%, respectively. The proposed GDLM accurately predicted macular OCT images following FTMH surgery, aiding patient and surgeon understanding of postoperative macular features.

Abstract Purpose To investigate the effect of topical nonsteroidal anti-inflammatory drugs (NSAIDs,) bromfenac on the intraretinal cystic lesions (IRC) when performing simultaneous cataract and idiopathic epiretinal membrane (iERM) surgery. Methods This study included patients with iERM who had been followed up for 6 months after vitrectomy, membrane removal, and concurrent cataract surgery. Eyes were treated with topical bromfenac or not. The baseline fluorescein angiography (FA) was obtained to assess the microvascular leakage (ML). Structural changes of macula, including IRC and central macular thickness (CMT) were assessed using optical coherence tomography (OCT). The main outcome measures were changes in IRCs and best-corrected visual acuity (BCVA) regarding FA findings. Results One hundred eighteen eyes were included. IRC and ML were observed in 51 eyes (43.2%) and 63 eyes (53.4%), respectively. The IRC did not show any association with the ML. Of total, 29 eyes (24.6%) were treated with topical bromfenac (Group A). Compared to Group B, topical bromfenac did not show beneficial effects in aspect of preventions for the newly developed IRC and treatment for pre-existed IRC. Whether the ML existed or not, topical bromfenac did not show any different effect on the changes in BCVA and IRC. Conclusion When performing simultaneous cataract and ERM surgery, topical NSAIDs, bromfenac did not show beneficial effects on the preventions and treatment of IRC in both eyes with and without the ML.

Abstract Purpose To evaluate the clinical patterns of retinopathy in patients who received cardiopulmonary resuscitation (CPR) using wide-field fundus photography and slit-lamp fundus examination. Methods The medical records of patients aged ≥ 18 years who survived after receiving CPR and underwent wide-field fundus photography and slit-lamp fundus examination within 3 months were retrospectively analyzed. Fundus findings, including retinal hemorrhage and cotton wool spots, were investigated. The subjects were categorized into the retinopathy and non-retinopathy groups based on the presence of fundus findings. Systemic and CPR-related factors were analyzed to compare the two groups. Results Twenty eyes (10 patients) and 28 eyes (14 patients) were included in the retinopathy and non-retinopathy groups, respectively. The retinopathy group had longer CPR time than the non-retinopathy group (15 ± 11 min vs. 6 ± 5 min, p = 0.027). In the retinopathy group, retinal nerve fiber layer hemorrhage was observed in all eyes, and intraretinal hemorrhage was observed in 55% of the eyes. 80% of hemorrhages were located in the peripapillary or posterior pole. There were no interval changes in visual acuity, intraocular pressure, and central retinal thickness for 6 months. The average remission periods of retinal hemorrhage and cotton wool spots were 6.8 ± 2.6 month and 5.6 ± 2.1 months, respectively. No retinopathy progression was observed. Conclusion The signs of retinopathy, such as retinal hemorrhages and cotton wool spots, which are found after CPR, mainly occur in patients who receive longer time of CPR and improve over time.

PurposeTo evaluate the efficacy of inverted internal limiting membrane (ILM) flap technique in full-thickness macular holes (MHs) with a size of ≤400 μm compared to the ILM peeling technique.MethodsRelated literatures that compared inverted ILM flap and ILM peeling in MHs ≤ 400 μm were reviewed by searching electronic databases including Pubmed, EMbase, ClinicalTrials.gov, and Cochrane Library up to April 2023. The primary outcome measure was hole closure rate, and the secondary outcome measures were the mean postoperative best-corrected visual acuity (BCVA), retinal sensitivity, and outer status of the retinal layers, including the external limiting membrane and ellipsoid zone. The quality of the articles was assessed according to the revised version of the Cochrane risk-of-bias tool for randomized trials or the Newcastle-Ottawa scale. In the case of heterogeneity, a sensitivity analysis was conducted, and publication bias was visually evaluated using a funnel plot.ResultsThis review included six studies with 610 eyes for the primary outcome and 385 eyes for the secondary outcomes, which were two randomized control trials and four retrospective studies. Pooled data revealed that the overall MH closure rate was 99.4% in the inverted ILM flap group and 96.2% in the ILM peeling group, without significant difference between the two groups (odds ratio = 3.91; 95% confidence interval, 0.82~18.69; P = 0.09). The inverted ILM flap technique did not have a favorable effect on the BCVA, retinal sensitivity, or recovery of the outer retinal layers. These results were consistent with those of the subgroup analysis of the different follow-up periods. No significant publication bias was observed.ConclusionIn eyes with MHs of ≤400 μm, both techniques demonstrated excellent surgical outcomes without significant differences. Therefore, surgical techniques can be selected according to surgeon preferences.

Abstract To evaluate the real-world treatment outcomes in patients with neovascular age-related macular degeneration (nAMD) in Korea, focusing on retinal fluid resolution. This multi-institutional retrospective chart review study, analyzed medical records of patients with nAMD (age ≥ 50 years) who received their first anti-vascular endothelial growth factor (VEGF) treatment in ophthalmology clinics across South Korea between January 2017 and March 2019. The primary endpoint was the proportion of patients with retinal fluid after 12 months of anti-VEGF treatment. The association between fluid-free period and VA gains was also evaluated. A total of 600 patients were enrolled. At baseline, 97.16% of patients had retinal fluid; after 12 months of anti-VEGF treatment, 58.10% of patients had persistent retinal fluid. VA improvements were relatively better in patients with absence of retinal fluid compared with presence of retinal fluid (+ 12.29 letters vs. + 6.45 letters at month 12; P

Abstract We evaluated the impact of signal strength (SS) on quantitative measurements from optical coherence tomography (OCTA). Twenty healthy-volunteers were included. A neutral density filter (NDF) was attached to spectral-domain OCTA (SD-OCTA) and swept-source OCTA (SS-OCTA). All subjects were imaged with both devices three times using three different conditions: no filter, NDF0.3, and NDF0.6. For SD-OCTA, SS decreased from 10.0 to 8.2 and 4.0 with the NDF0.3 and 0.6, respectively. The vessel density (VD) and vessel length density (VLD) of the superficial capillary plexus (SCP) decreased when SS decreased from 10 to 8.2, but no further decrease when SS changed from 8.2 to 4.0. The flow metrics of the deep capillary plexus (DCP) did not change. For SS-OCTA, SS decreased from 10 to 9.5 and 7.2. The VD and VLD of the SCP and DCP decreased when SS decreased, except for the VD of the DCP when SS changed from 10 to 9.5. The choriocapillaris flow deficits significantly increased along with the decrease in SS. Quantitative flow parameters were significantly affected by a small change in SS and were most conspicuous in the SCP and choriocapillaris. These finding highlight the importance of high and consistent SS in quantitative OCTA studies.

AIM: To predict final visual acuity and analyze significant factors influencing open globe injury prognosis. METHODS: Prediction models were built using a supervised classification algorithm from Microsoft Azure Machine Learning Studio. The best algorithm was selected to analyze the predicted final visual acuity. We retrospectively reviewed the data of 171 patients with open globe injury who visited the Pusan National University Hospital between January 2010 and July 2020. We then applied cross-validation, the permutation feature importance method, and the synthetic minority over-sampling technique to enhance tool performance. RESULTS: The two-class boosted decision tree model showed the best predictive performance. The accuracy, precision, recall, F1 score, and area under the receiver operating characteristic curve were 0.925, 0.962, 0.833, 0.893, and 0.971, respectively. To increase the efficiency and efficacy of the prognostic tool, the top 14 features were finally selected using the permutation feature importance method: (listed in the order of importance) retinal detachment, location of laceration, initial visual acuity, iris damage, surgeon, past history, size of the scleral laceration, vitreous hemorrhage, trauma characteristics, age, corneal injury, primary diagnosis, wound location, and lid laceration. CONCLUSION: Here we devise a highly accurate model to predict the final visual acuity of patients with open globe injury. This tool is useful and easily accessible to doctors and patients, reducing the socioeconomic burden. With further multicenter verification using larger datasets and external validation, we expect this model to become useful worldwide.