PURPOSE
To assess long-term functional and structural outcomes in eyes treated with laser photocoagulation for retinopathy of prematurity (ROP).
METHODS
Patients treated for ROP (1997-2004) underwent best-corrected visual acuity (BCVA) assessment, autorefraction, biometry, OCT, and OCT angiography of the macula and optic disc. Patient characteristics were retrieved from neonatal medical records.
RESULTS
Sixty-three eyes of 33 patients with a mean follow-up of 19 years, and 22 eyes of 11 age-matched controls were included. The laser-treated group had a mean BCVA of 0.133 (±0.17) logMAR. Poor anatomical outcome occurred in 6.3% (n = 4), with 2 eyes requiring surgery. Mean spherical equivalent was -5.04 (±5.13) D. Corneal curvature and lens thickness were significantly higher in the study group (p < .001). Central macular thickness was greater in ROP eyes (311.97 μm vs 252.57 μm; p < .001). Vascular density was increased in the foveal region and decreased in other regions (p < .001), and the foveal avascular zone was smaller (p < .001). Optic disc analysis showed a reduction in cup-to-disc ratios and peripapillary capillary density (p < .001).
CONCLUSIONS
Two decades after laser treatment, our patients demonstrated excellent visual acuity and favourable anatomic outcomes. Myopia was linked to increased corneal curvature and lens thickness. Significant alterations in the vascular plexuses of the macula and optic disc occur and may impact visual function in adulthood.
Dolev Dollberg, Karny Shouchane-Blum, Orly Gal-Or et al.· Canadian Journal of Ophthalm...· 0 citations
PURPOSE
Myopia is rapidly increasing worldwide, with macular neovascularization (MNV) representing a major cause of vision loss. Early diagnosis remains challenging. Our study aims to incorporate a deep learning (DL) model to identify and localize myopic MNV using OCT images.
METHODS
A total of 6,215 OCT B-scan images of patients with pathologic myopia were included, of which 1,664 contained MNV. MNV lesions were manually annotated and validated by retina specialists. The dataset included additional myopia-related pathologies to account for factors affecting model robustness. Data were split into training and validation sets using patient-level 5-fold cross-validation. A YOLOv11 (You Only Look Once) object detection model was trained to localize MNV. Two postprocessing pipelines using either non-maximum suppression (NMS) or weighted box fusion (WBF) were tested. Model performance was evaluated at 25% intersection over union (IoU) using precision, recall, and mean average precision (mAP25).
RESULTS
The YOLOv11 models achieved mean precision ranging from 0.73 to 0.79 and recall from 0.76 to 0.80. The highest mAP25 reached 0.84. Adapting the model's output into an image-level binary classification yielded precision up to 0.85, recall up to 0.88, and F1-score up to 0.85.
CONCLUSION
A YOLOv11-based deep learning model reliably detects and localizes myopic MNV on OCT imaging, overcoming the challenge of structurally complex eyes with coexisting myopic pathologies. This tool may serve as an automated reader to assist with earlier diagnosis and timely anti-VEGF intervention in patients with pathologic myopia.
Hanna Moalem, Einav Baharav Shlezinger, A. Tiosano et al.· Retina· 0 citations
