OBJECTIVE: This paper aims to introduce a GAN technology for the diagnosis of eye disorders, particularly glaucoma. This paper illustrates deep adversarial learning as a potential diagnostic tool and the challenges involved in its implementation. This study describes and analyzes many of the pitfalls and problems that researchers will need to overcome to implement this kind of technology.
METHODS: To organize this review comprehensively, articles and reviews were collected using the following keywords: ("Glaucoma," "optic disc," "blood vessels") and ("receptive field," "loss function," "GAN," "Generative Adversarial Network," "Deep learning," "CNN," "convolutional neural network" OR encoder). The records were identified from 5 highly reputed databases: IEEE Xplore, Web of Science, Scopus, ScienceDirect, and PubMed. These libraries broadly cover the technical and medical literature. Publications within the last 5 years, specifically 2015-2020, were included because the target GAN technique was invented only in 2014 and the publishing date of the collected papers was not earlier than 2016. Duplicate records were removed, and irrelevant titles and abstracts were excluded. In addition, we excluded papers that used optical coherence tomography and visual field images, except for those with 2D images. A large-scale systematic analysis was performed, and then a summarized taxonomy was generated. Furthermore, the results of the collected articles were summarized and a visual representation of the results was presented on a T-shaped matrix diagram. This study was conducted between March 2020 and November 2020.
RESULTS: We found 59 articles after conducting a comprehensive survey of the literature. Among the 59 articles, 30 present actual attempts to synthesize images and provide accurate segmentation/classification using single/multiple landmarks or share certain experiences. The other 29 articles discuss the recent advances in GANs, do practical experiments, and contain analytical studies of retinal disease.
CONCLUSIONS: Recent deep learning techniques, namely GANs, have shown encouraging performance in retinal disease detection. Although this methodology involves an extensive computing budget and optimization process, it saturates the greedy nature of deep learning techniques by synthesizing images and solves major medical issues. This paper contributes to this research field by offering a thorough analysis of existing works, highlighting current limitations, and suggesting alternatives to support other researchers and participants in further improving and strengthening future work. Finally, new directions for this research have been identified.
METHODS: Review of prospectively collected data from a multicentre, randomised controlled trial comparing CLE and laser peripheral iridotomy. Eligible participants were ≥50 years old and newly diagnosed with (1) primary angle closure (PAC) with intraocular pressure above 30 mm Hg or (2) PAC glaucoma. We report the postoperative corrected distance visual acuity (CDVA) and refractive outcomes at 12 and 36 months postoperatively for those who underwent CLE.
RESULTS: Of the 419 participants, 208 were randomised to CLE. Mean baseline CDVA was 77.9 (SD 12.4) letters and did not change significantly at 36 months when mean CDVA was 79.9 (SD 10.9) letters. Mean preoperative spherical equivalents were +1.7 (SD 2.3) and +0.08 (SD 0.95) diopters (D) at 36 months. Fifty-nine per cent and 85% eyes were within ±0.5D and ±1.0D of predicted refraction, respectively, at 36 months.
CONCLUSIONS: Mean CDVA in patients undergoing CLE for angle-closure glaucoma appeared stable over the 3-year study period. Refractive error was significantly reduced with surgery but refractive predictability was suboptimal.
TRIAL REGISTRATION NUMBER:
METHODS: Retrospective analysis of prospectively collected data from adults with primary angle closure or primary angle closure glaucoma enrolled in the Effectiveness, in Angle-closure Glaucoma, of Lens Extraction study. We included data from 335 participants with patient reported visual function (VFQ-25) and health status measured by the EQ-5D-3L over 36 months. We used the recommended anchor-based methods (receiver operating characteristic (ROC), predictive modelling and mean change) to determine the MID of the VFQ-25. EQ-5D-3L anchor change was defined as none (<0.065); minimal (0.065≤EQ-5D-3L change ≤0.075 points) and greater change (>0.075 points).
RESULTS: Mean baseline VFQ-25 score was 87.6 (SD 11.8). Estimated MIDs in the change in VFQ-25 scores (95% CI) were 10.5 (1.9 to 19.2); 3.9 (-2.3 to 10.1); 5.8 (1.9 to 7.2) and 8.1 (1.7 to 14.8) for the 'within-patient', 'between-patient' change, ROC and predictive modelling anchor methods respectively. Excluding estimates from the methodologically weaker 'within-patient' method, the MID of a change in VFQ-25 composite score is 5.8 (median value).
CONCLUSIONS: Estimates of the MID using multiple methods assist in the interpretation of the VFQ scores. In the context of early glaucoma related visual disability, a change score of around six points on the VFQ-25 is likely to be important to patients. Further confirmatory research is required. Studies comparing changes in patient-reported outcome measure scores with a global measure of patients' perceived change are required.
METHODS: This validated questionnaire-based study was conducted over 1-month during which Ngenuity 3D surgery was demonstrated. All surgeons and trainees exposed were recruited to complete a questionnaire comprising visualization, physical, ease of use, teaching and learning, and overall satisfaction.
RESULTS: All 7 surgeons and 33 postgraduate students responded. Surgeons reported no significant difference except overall (P = 0.047, paired t-test). Postgraduate trainees reported significantly better experience with 3D for illumination (P = 0.008), manoeuvrability (P = 0.01), glare (P = 0.037), eye strain (P = 0.008), neck and upper back strain (P = 0.000), lower back pain (P = 0.019), communication (P = 0.002), comfortable environment (P = 0.001), sharing of knowledge (P = 0.000), and overall (P = 0.009).
CONCLUSIONS: During early experience, surgeons and trainees reported better satisfaction with 3D overall. Trainees had better satisfaction with 3D in various subcomponents of visualization, physical, ease of use, and education.
PATIENT CONCERNS: A 64-year-old man with underlying well controlled diabetes mellitus was treated with 2 weeks' course of intravenous antifungal fluconazole for pyelonephritis as his blood culture grew Candida albicans. Concurrently, he complained of 3 months of bilateral painless progressive blurring of vision. At presentation, his visual acuity (VA) was light perception both eyes. Ocular examination revealed non granulomatous inflammation with dense vitritis of both eyes.
DIAGNOSIS: He was diagnosed with EFE but the condition responded poorly with the medications.
INTERVENTIONS: He was treated with intravitreal (IVT) amphotericin B and fluconazole was continued. Vitrectomy was performed and intraoperative findings included bilateral fungal balls in the vitreous and retina with foveal traction in the left eye. Postoperatively, vision acuity was 6/24, N8 right eye and 2/60, N unable for left eye with extensive left macular scar and hole. Vitreous cultures were negative. He received multiple IVT amphotericin B and was started on topical steroid eye drops for persistent panuveitis with systemic fluconazole. Ocular improvement was seen after switching to IVT and topical voriconazole. Despite this, his ocular condition deteriorated and he developed neovascular glaucoma requiring 3 topical antiglaucoma agents. Panretinal photocoagulation was subsequently performed.
OUTCOMES: At 3 months' follow-up, his vision acuity remained at 6/24 for right eye and 2/60 for the left eye. There was no recurrence of inflammation or infection in both eyes.
LESSONS: Voriconazole could serve as a promising broad spectrum tri-azole agent in cases of failure in first-line treatment or drug-resistant fungus.
METHODS: Seventeen POAG patients with suboptimal IOP control despite pre-existing topical medications were subjected to adjunct SLT (50 applications 180 degrees) or second line medical therapy. Current medications were continued, and patients were followed up for 6 months for degree of intraocular pressure (IOP) lowering. HRQoL was assessed using Glaucoma Quality of Life 36-item (GlauQoL-36), Assessment of Quality of Life-7D (AQoL-7D) and Vision related Quality of Life (VisQoL). Costs involved were calculated and compared to the effect (IOP reduction) achieved in each arm.
RESULTS: Ten patients were in the SLT group and 7 in the topical medication (MED) group. Mean baseline intraocular pressure (IOP) was 18.90±3.48mmHg in SLT group and 15.57±2.23mmHg in MED group. Mean reduction of IOP was 4.30±1.64mmHg in SLT group and 2.71±2.56 mmHg in MED group at 6 months which was not statistically significant (p=0.14) between two groups. All the HRQoL questionnaires did not show significant changes in the groups or between groups when compared baseline with 6-month post treatment (p-values ranging from 0.247 to 0.987). For every 1mmHg reduction in IOP, cost involved in MED group (RM53.61) was 165% of the cost involved in SLT group (RM32.56).
DISCUSSION AND CONCLUSION: This study has shown that SLT was as effective clinically and tolerable as topical anti glaucoma medications and was possibly more cost effective in the step-up treatment of patients with POAG at 6 months follow- up.
METHODS: A total of 195 5- and 6-year-old preschoolers were recruited from children attending Hospital Selayang, Selangor, Malaysia, to test the app. Uncooperative children and those with visual acuity of >logMAR 0.6 were excluded. Results from parents and the screening doctor using the app (Lea symbols) to test visual acuity were compared to each other and to gold standard vision testing by an optometrist using the Lea symbols chart.
RESULTS: Children 5 years of age represented 46.7% of the study population. The mean age of parents was 37.27 ± 7.68 years. Bland-Altman scatterplot agreement between assessors mainly was within the 95% confidence interval for bilateral eyes screening. Parents obtained a sensitivity of 86.6% (right vision) and 79.5% (left vision) and specificity of 78.9% (right vision) and 71.8% (left vision). Parents took a mean of 191.2 ± 70.82 seconds for bilateral screening. The intraclass correlation coefficient between optometrist and parents in bilateral eyes screening was good (P 0.7, indicating high internal reliability of the app. Most parents (178/195 [91.3%]) strongly agreed on the app's acceptability and ease of use.
CONCLUSIONS: The AAPOS Vision Screening App used by parents is a promising tool for visual acuity screening among Malaysian preschool children and a reliable app for vision screening.
PURPOSE: This study aimed to determine the prevalence of depression among patients with POAG and examine the relationship between depression and the severity of POAG in older adults.
MATERIALS AND METHODS: Three hundred and sixty patients with POAG aged 60 years or above were recruited from 2 tertiary centers located in an urban and suburban area. The participants were stratified according to the severity of their glaucoma based on the scores from the modified Advanced Glaucoma Intervention Study (AGIS) to mild, moderate, severe, and end stage. Face-to-face interviews were performed using the Malay Version Geriatric Depression Scale 14 (mGDS-14) questionnaire. Depression is diagnosed when the score is ≥8. One-way analysis of variance was used to compare the subscores between the groups. Multifactorial analysis of variance was also applied with relevant confounding factors.
RESULTS: Depression was detected in 16% of older adults with POAG; a higher percentage of depression was seen in those with end stage disease. There was a significant increase in the mean score of mGDS-14 according to the severity of POAG. There was evidence of an association between depression and severity of visual field defect (P<0.001). There was a significant difference in mGDS-14 score between the pairing of severity of POAG [mild-severe (P=0.003), mild-end stage (P<0.001), moderate-severe (P<0.001), and moderate-end stage (P<0.001)] after adjustment to living conditions, systemic disease, and visual acuity.
CONCLUSION: Ophthalmologists should be aware that older adults with advanced visual field defects in POAG may have depression. The detection of depression is important to ensure adherence and persistence to the treatment of glaucoma.
PATIENTS AND METHODS: In total, 100 eyes from 50 patients on long-term intranasal steroids (>2 y) for allergic rhinitis and 90 eyes from 45 controls were included in this study. Patients on other forms of steroids and risk factors for glaucoma were excluded. IOP was measured and nonmydriatic stereoscopic optic disc photos were taken for each eye. The vertical cup-to-disc ratio and the status of the optic disc were evaluated.
RESULTS: The mean IOP for intranasal steroids group was significantly higher (15.24±2.31 mm Hg) compared to the control group (13.91±1.86 mm Hg; P=0.000). However, there were no significant differences in the vertical cup-to-disc ratio and the status of glaucomatous optic disc changes between the groups.
CONCLUSIONS: Prolonged use of intranasal steroids cause statistical significant increase in IOP in patients with allergic rhinitis although no significant glaucomatous disc changes were seen. We suggest patients on long-term use of intranasal steroid have a yearly eye examination to be monitored for IOP elevation and those with additional risk factors for glaucoma is closely monitored for glaucoma.
METHODS: This was a prospective observational study carried out at a tertiary referral centre. POAG patients on topical antiglaucoma medications and planned for phaco-ECP were recruited. WDT was performed before surgery and 6 weeks postoperatively by drinking 10 mL/kg of water in 5 min followed by serial IOP by Goldmann applanation tonometry measurements at 15, 30, 45, and 60 min. Mean IOP, IOP fluctuation (difference between highest and lowest IOP), IOP reduction, and factors affecting IOP fluctuation were analysed.
RESULTS: Twenty eyes from 17 patients were included. Baseline IOP was similar before (14.7 ± 2.7 mm Hg) and after (14.8 ± 3.4 mm Hg, p = 0.90) surgery. There was no difference in mean IOP (17.6 ± 3.4 mm Hg vs. 19.3 ± 4.7 mm Hg pre- and postoperative, respectively, p = 0.26) or peak IOP (19.37 ± 3.74 mm Hg vs. 21.23 ± 5.29 mm Hg, p = 0.25), albeit a significant reduction in IOP-lowering medications (2.2 ± 1.15 vs. 0.35 ± 0.93, p < 0.001) postoperatively. IOP fluctuation was significantly greater (6.4 ± 3.2 mm Hg vs. 4.6 ± 2.1 mm Hg, p = 0.015) with more eyes having significant IOP fluctuation of ≥6 mm Hg (11 eyes [55%] vs. 4 eyes [20%], p < 0.001) postoperatively. Factors that were significantly associated with increased postoperative IOP fluctuations were higher preoperative IOP fluctuation (β = 0.69, 95% CI 0.379-1.582, p = 0.004) and more number of postoperative antiglaucoma medications (β = 0.627, 95% CI 0.614-3.322, p = 0.008).
CONCLUSION: Reducing aqueous production with phaco-ECP does not eliminate IOP fluctuation in POAG patients. The increase in postoperative IOP fluctuation suggests increased outflow resistance after phaco-ECP.