Citations

ABSTRACT Artificial intelligence (AI) has emerged as a transformative technology in the fields of medicine and ophthalmology. Recent advances in machine learning, deep learning and convolutional neural networks have significantly improved cataract detection and grading, intraocular lens power calculations, surgical decision‐making, risk prediction and patient education. Intraoperative AI‐assisted surgical phase recognition, workflow analysis, real‐time guidance, robotic technologies and femtosecond laser integration can improve surgical precision, efficiency and safety. AI has shown promise in predicting visual outcomes, detecting postoperative complications and facilitating telemedicine follow‐up. Though challenges with AI include transparency, data diversity, regulatory oversight and clinical validation, continued advances are expected to further personalise cataract surgery, support clinical decision‐making, optimise surgical performance and improve patient outcomes. This narrative summarises the current and emerging applications of AI in preoperative, intraoperative and postoperative management of cataracts.

Abstract Introduction Cystoid macular edema (CME) is a recognized complication associated with penetrating keratoplasty (PK) and Endothelial Keratoplasty (EK). The purpose of this study was to determine the incidence of acute CME within the first 6 months after PK, Descemet stripping automated EK, or Descemet’s membrane EK. Methods A retrospective cohort analysis of the TriNetX United States collaborative database was conducted. Patients with a diagnosis of Fuchs dystrophy or keratoconus who underwent PK or EK between January 1, 2004 and August 30, 2024 were included. Those with a history of uveitis, retinal vascular occlusion, diabetic macular edema, or prior vitrectomy were excluded. Outcomes and diagnoses were defined using current procedural terminology and International Classification of Diseases, Tenth edition (ICD-10) codes. Cohorts were stratified based on a history of prior PK or EK with >6 months follow-up. Propensity score matching was performed on both cohorts to control for age at index, sex, race, ethnicity, and the diagnosis of Fuchs dystrophy and keratoconus. The incidence of acute CME during a 6-month postoperative follow-up period was assessed. Odds ratios were calculated between cohorts. Results A total of 2516 eyes were included in the study (PK: 1258; EK: 1258). At six months postoperatively, the incidence of acute CME in the EK cohort (46 of 1258) was significantly higher than in the PK cohort (20 of 1258) (P = 0.0012) (OR 0.426, 95% confidence interval 0.25–0.72). Conclusions Our findings suggest that the incidence of acute CME is higher in patients undergoing EK compared to PK.

Purpose: To determine the factors associated with compliance with recommended intravitreal injections (IVIs) for retinal diseases among out-of-pocket, insured, and fully subsidized (non-paying) patients. Methods: This retrospective observational hospital-based study included patients recommended for IVI in at least one eye between November 2011 and May 2022. Three cohorts of patients were assessed: out-of-pocket payers, insured, and those fully subsidized (non-paying) by the hospital. The main outcome measure was the compliance rate, defined as the proportion of patients receiving an injection after one was recommended. The impact on the visual outcome in each cohort was a secondary outcome. Results: Overall, 50,408 IVI were recommended for at least one eye and were included in the analyses. Male patients were more common (66.72%), and the mean age was 55.36 ± 16.39 years. The overall compliance rate with the first recommended injection was 51.37%, which increased and eventually stabilized with subsequent recommendations to approximately 80% (range: 77.48%–82.93%). Insured patients had a higher compliance rate with the first recommendation compared to the non-insured (100% vs. 49.84%; P < 0.000 01 ), and non-paying patients had a higher compliance rate after the first recommendation compared to paying patients (62.51% vs. 48.66%; P < 0.000 01 ). Patients with insurance achieved a better visual outcome at the last follow-up compared to other cohorts. Conclusion: Insured patients have a much higher compliance rate with recommended IVIs compared to out-of-pocket or non-paying patients. The high non-compliance rate among non-paying patients suggests that factors other than the direct cost of the injections may impact compliance with recommended treatment.

A systematic review was conducted of studies published up to 30 August 2024. Studies comparing conventional visual field (VF) indices, ability to detect central visual field defects (CVFDs), structure–function (S-F) concordance, and test characteristics across the HVF 24-2C SITA-Faster, 24-2 SITA-Standard/Faster, and 10-2 SITA-Standard/Fast tests were included. Eight studies with 1239 subjects (49.1% male; mean age, 54.8–66.9 years) were analyzed. The 24-2C produced similar global VF indices compared to the 10-2 and 24-2 (Standard/Faster) (ICC = 0.95 and 0.80, p < 0.001), detected more VF defects and CVFD clusters, and demonstrated greater S-F concordance than the 24-2 (Standard/Faster). Although the 10-2 (Standard/Fast) detected significantly more CVFDs and had a greater S-F concordance, the agreement between the 24-2C and 10-2 grids was substantial (ĸ = 0.488 to 0.708). The 24-2C was also faster compared to the 24-2 Standard and 10-2 (Standard/Fast), with comparable false positives, higher false negatives, and fewer fixation losses than the 24-2 Standard. In conclusion, the HVF 24-2C is quicker and identifies more CVFDs than the 24-2 grid, demonstrates high agreement with the 10-2 grid, and aids in CVFD screening.

Abstract Background: Chronic dry eye disease (CDED) is a condition of the ocular surface that causes discomfort, visual disturbance, conjunctival injection, and photophobia. A debilitating form of CDED in patients undergoing allogeneic bone marrow transplant results from ocular graft versus host disease (oGVHD). Autologous serum eye drops (ASED) have beneficial effects on the ocular surface. The aim of this study is to compare improvements in symptoms and function in a cohort of CDED patients with ocular GVHD treated with ASED and a cohort of CDED without GVHD (non-GVHD). We additionally report on some of the biologic characteristics of one specific ASED formulation. Materials and Methods: A total of 71 patients seen in a multi-specialty ophthalmology practice in a university hospital setting receiving ASEDs underwent ocular surface disease index (OSDI) and visual function questionnaire measurements at baseline, 3 months, and 6 months. Twenty-four ASED samples were studied for biomarkers. Results: Compared to baseline, OSDI scores significantly decreased after 3 and 6 months of ASED treatment in both oGVHD and non-GVHD CDED groups. Additionally, visual function questionnaire-25 (VFQ-25) subscales (ocular pain, distance activity, driving, role difficulties, dependency, and mental health) were significantly improved at 3 months and 6 months, and social functioning was improved at 6 months. The majority of patients in both groups had severe disease. Improvements in patient-reported outcomes were greater in the oGHD CDED group than in the non-GVHD CDED group There were no reported complications. Conclusion: ASED is safe and effective in patients with CDED due to oGVHD and non-GVHD. Those with oGVHD had greater improvements. These findings potentiate the wider adoption of this therapy in CDED patients and warrant additional studies into the biochemical makeup of ASED.

Purpose: The purpose of this study was to evaluate and compare the anatomic characteristics of eyes with Fuchs endothelial corneal dystrophy (FECD) with eyes without FECD. Methods: This study was a retrospective chart review performed at an academic medical center. Patients with FECD were identified through a search of the electronic medical records. Eligible patients underwent Scheimpflug imaging and optical biometry and were compared with age and sex-matched control subjects who underwent similar testing in preparation for cataract surgery. Several measurements of the cornea, anterior chamber, and eyes were evaluated using multivariable linear regression models and multivariable logistic regression models. Results: A total of 404 eyes (202 eyes with FECD and 202 control eyes) were included in this study. Compared with controls, eyes with FECD had shallower AC depths, lower AC volumes, and narrower angles. Conversely, the spherical equivalent before cataract surgery, corneal pachymetry, and corneal volume were higher in eyes with FECD. On Scheimpflug imaging analysis, these anatomical differences were present in FECD eyes with and without corneal edema. After adjusting for sex, these differences remained statistically significant. Shorter axial length was found to be statistically significant in male eyes but not in female eyes with FECD. Conclusions: This study reports new ocular characteristics in FECD eyes with and without edema. Optical biometry and Scheimpflug imaging established that the anatomic findings in eyes with FECD were not simply due to the larger volume of an edematous cornea but rather unique to eyes with FECD. These findings will provide reliable, normative data for future studies examining surgical, medical, and anatomical factors in FECD.

Objective To report a fatal case of Susac syndrome in a congenitally deaf patient with a cochlear implant and a history of migraines, emphasizing the diagnostic challenges in patients with preexisting conditions. Patient A 33-year-old male with congenital hearing loss, a cochlear implant, and chronic migraines who presented with mild subacute auditory disturbance and headaches that later progressed to severe encephalopathy. Intervention Explantation of a non–magnetic resonance imaging (MRI) compatible cochlear implant followed by MRI, fundoscopy, and the administration of immunosuppressive medications. Main Outcome Measures Diagnosis was confirmed by characteristic MRI appearance and the presence of a hemi-retinal artery occlusion. Results After weeks of immunosuppressive treatment, the patient died of a global cerebral ischemic event of unknown origin. Conclusions For patients with preexisting sensorineural hearing loss and cochlear implants, Susac syndrome poses a diagnostic challenge. Auditory disturbances in the absence of cochlear implant failure should prompt further evaluation for visual disturbances and encephalopathy. MRI and fundoscopy should be performed to detect other features of the disease.