Artificial intelligence in optic disc imaging: implications for future neurology

Artificial intelligence (AI) has emerged as a promising tool for quantitative analysis of optic disc imaging, transforming evaluation from a subjective clinical assessment into an objective framework capable of extracting innovative and meaningful structural biomarkers. A structured literature search was performed using PubMed/MEDLINE, Scopus, and Google Scholar for studies published through May 2026, supplemented by manual review of reference lists. This review describes the application of deep learning to optic disc segmentation and quantification of features such as disc edema and pallor from fundus photographs, enabling accurate classification of optic nerve abnormalities. Several AI models have demonstrated performance comparable to expert assessment in selected validation studies, supporting future clinical triage where direct ophthalmoscopy is performed inconsistently. In papilledema and optic neuritis, AI-assisted fundus image analysis may facilitate early diagnosis, particularly when access to magnetic resonance imaging (MRI) or expert neuroophthalmological assessment is delayed. However, clinical implementation in neurology remains at an early stage because of limited neurology-specific datasets, variability across imaging devices, and the lack of large-scale prospective validation linked to neurological outcomes. Future development needs to focus on multimodal integration with optical coherence tomography (OCT), MRI, and serum biomarkers, together with decentralized image acquisition using smartphone-based fundus imaging platforms.