Artificial intelligence in cataract surgery: Smart screening, precision planning, and cataract coaching

Artificial intelligence (AI) is rapidly transforming cataract surgery, the world’s most frequently performed procedure, by enhancing precision, efficiency, and patient outcomes across the entire care pathway. In preoperative planning, machine learning algorithms for intraocular lens (IOL) power calculation, such as the Hill-RBF and Kane formulas, have demonstrated superior accuracy over traditional methods, achieving mean absolute errors below 0.30 D, particularly in complex cases. Deep learning models are also enabling automated cataract screening and grading from various imaging sources with high accuracy (up to 98%), and AI assistants are being developed to aid in complex surgical decision-making with high consistency (96.26%) compared to experts. Intraoperatively, AI analyzes surgical videos for workflow recognition, instrument tracking, and objective skill assessment, distinguishing between junior and senior surgeons and providing real-time risk prediction for complications with probabilities as high as 92%. Postoperatively, AI facilitates telemedicine follow-ups through automated clinical assistants like Dora and predicts late complications such as posterior capsule opacification with up to 92.2% accuracy. Despite this promise, significant challenges remain, including data scarcity (especially for MSICS), the need for external validation, and critical medicolegal and ethical concerns surrounding data privacy, algorithmic bias from nondiverse datasets, liability, and the “black-box” nature of many models. The future lies in leveraging AI for enhanced surgical training with real-time feedback, bridging global healthcare disparities through accessible screening tools, and developing advanced multimodal models. Realizing AI’s full potential will require robust collaboration to address these limitations and build frameworks that ensure accountability and trust.