Background/Objectives: Subjective refraction is the clinical gold standard but is time-consuming and examiner-dependent. Most artificial intelligence (AI)-based approaches rely on specialized imaging or biometric data not routinely available. This study aimed to predict subjective refraction using only routine, non-cycloplegic autorefraction and keratometric data and to identify factors associated with reduced prediction accuracy. Methods: This retrospective study included 1856 eyes from 1006 patients. A multi-output histogram gradient-boosting model predicted subjective spherical equivalent, cylindrical power, and astigmatic axis. Performance was evaluated on an independent test dataset using R² and mean absolute error, with circular statistics for axis prediction. Prediction failure was assessed using clinically relevant tolerance thresholds (sphere/cylinder ≤ 0.50 D; axis ≤ 10°) and multivariable logistic regression. Results: The model achieved high accuracy for spherical and cylindrical prediction (R² = 0.987 and 0.933; MAE = 0.126 D and 0.137 D). Astigmatic axis prediction demonstrated strong circular agreement (ρ = 0.898), with a mean absolute angular error of 4.65° (median, 0.96°). Axis errors were higher in eyes with low cylinder magnitude (<0.75 D) and oblique astigmatism. In multivariable analysis, steeper keratometry (K2; OR = 7.25, 95% CI 1.62–32.46, p = 0.010) and greater objective cylindrical power (OR = 2.79, 95% CI 1.87–8.94, p = 0.032) were independently associated with poor prediction. Conclusions: A machine-learning model based solely on routine, non-cycloplegic autorefractor and keratometric measurements can accurately estimate subjective refraction, supporting AI as a complementary decision-support tool rather than a replacement for conventional subjective refraction.

Prevalence of Retinal Vascular Diseases in a Tertiary Care Hospital in Türkiye: A Hospital-Based Epidemiologic Study - Turkish Journal of Ophthalmology

AIM: To analyze the ocular findings of patients who received lifelong leptin therapy due to congenital leptin deficiency (CLD), an extremely rare condition. METHODS: A prospective, cross-sectional comparative study was performed on six patients with CLD and 13 healthy age- and sex-matched controls. The central corneal thickness (CCT), anterior chamber depth (ACD), axial length (AL), keratometry (K1, K2), optical coherence tomography (OCT), and OCT angiography parameters were compared between the leptin and control groups at the baseline visit. The change in these measurements in leptin patients over a two-year period was analyzed. RESULTS: CLD patients had lower mean AL, ACD, and CCT (P≤0.012 for all). Mean K1, K2 (P≤0.047 for both), choroidal thickness (P≤0.001), and central ganglion cell layer (GCL) thickness (P=0.029) were higher in the leptin group. Perifoveal superficial capillary plexus (SCP) density was decreased in all quadrants except the temporal region (P<0.05), and parafoveal deep capillary plexus (DCP) density was decreased in the superior hemisphere, temporal quadrant (P≤0.036 for both) and nasal quadrant (P=0.048) in the leptin group. During the two-year follow-up, no changes in anterior and posterior segment measurements were observed in the leptin patients, except for subfoveal choroidal thickness (P<0.001). CONCLUSION: CLD patients exhibit structural alterations in both the anterior and posterior segments of the eye, including notable changes in retinal and choroidal vasculature. However, there is limited evidence concerning the influence of leptin therapy on the eye.

Background/Objectives : To evaluate the associations between corneal topographic irregularity, higher-order aberrations (HOAs), and posterior segment structural and microvascular parameters in keratoconus using optical coherence tomography (OCT) and OCT angiography (OCTA). Methods: In this cross-sectional study, 81 eyes with keratoconus and 60 healthy control eyes underwent corneal topography and wavefront analysis, spectral-domain OCT with enhanced depth imaging, and OCTA. Retinal layer thicknesses, choroidal thickness and area metrics, choroidal vascularity index (CVI), and OCTA-derived vascular parameters were analyzed. Associations were assessed using Spearman correlation analysis with false discovery rate (FDR) correction. Results: Compared with controls, keratoconus eyes showed significantly increased corneal curvature, corneal irregularity indices, and HOAs (all p < 0.001). Structural OCT analysis demonstrated preserved inner retinal layers, whereas outer nuclear layer thickness was reduced (p < 0.001) and overall outer retinal layer thickness was increased (p = 0.005). Choroidal thickness and both total and luminal choroidal areas were significantly greater in keratoconus eyes (all p ≤ 0.011), while CVI did not differ between groups (p > 0.05). OCTA revealed reduced superficial capillary plexus vessel density at the whole image and perifoveal regions (all p < 0.001), whereas deep capillary plexus and foveal avascular zone metrics were largely preserved. Correlation analyses identified only weak and inconsistent associations between corneal parameters, HOAs, and posterior segment measurements, none of which remained statistically significant after FDR correction. Conclusions: Despite pronounced anterior segment deformation and optical degradation, posterior segment structural and microvascular alterations in keratoconus are limited and weakly related to corneal disease severity. These findings support a predominantly anterior segment centered pathophysiology of keratoconus and highlight the importance of stringent multiple-comparison control in multimodal imaging studies.