Abstract
Purpose
To summarize recent advances in artificial intelligence (AI) applications across the entire cataract management pathway and to discuss current challenges and future directions.Methods
We conducted a narrative mini review of studies published since 2021 on AI applications in cataract diagnosis and screening, intraocular lens (IOL) power calculation, intraoperative guidance, and postoperative management, with emphasis on deep learning and emerging technologies.Results
AI models have achieved expert-level performance in automated cataract detection and grading from slit-lamp images, with pooled AUC values approaching 0.99 in meta-analyses. AI-based IOL power calculation formulas have shown superior accuracy compared with traditional formulas, especially in challenging cases such as highly myopic eyes. Intraoperatively, AI systems can recognize surgical phases and track instruments with high accuracy, supporting surgical training and quality assessment. Postoperatively, multimodal AI models integrating imaging and clinical data have improved the prediction of visual outcomes and complications. Emerging technologies, including large language models, portable smartphone-based screening devices, and multimodal prediction models, represent promising new directions, although challenges related to data generalizability, algorithm interpretability, and clinical translation remain to be addressed.Conclusion
AI has the potential to significantly improve the efficiency, accuracy, and accessibility of cataract care, but should serve as an assistive tool rather than a replacement for clinical judgment.