Prediction of Postoperative Macular Hole Status by Automated Preoperative Retinal OCT Analysis: A Narrative Review.

Abstract

Optical coherence tomography (OCT) is a non-invasive imaging modality essential for macular hole (MH) management. Artificial intelligence (AI) algorithms could be applied to OCT to garner insights for MH prognosis and outcomes. The objective was to review literature assessing automated image analysis algorithms in predicting postoperative outcomes for MH patients based on OCT images. A narrative search of all available published studies in peer-reviewed journals was conducted up to June 2023 following PRISMA guidelines. Three hundred sixty-eight publications underwent screening, with 14 selected for full-text review and seven determined as relevant. In MH status prediction, AI models achieved an area under the curve (AUC) of 83.6% to 98.4%. For postoperative visual acuity prediction, algorithm performance ranged from AUCs of 57% to 85%. In conclusion, novel AI algorithms were found to be predictive for postoperative MH status and postoperative visual acuity. More research in larger populations should be conducted to gauge the value of these novel algorithms in a real-world setting. [Ophthalmic Surg Lasers Imaging Retina 2025;56:XX-XX.].