In an editorial written by Miernik, the growing role of artificial intelligence (AI) in urology, particularly in endourology, is critically appraised, with a focus on a novel study by Lu et al. published in this issue of BJUI. Ureteroscopy with laser lithotripsy remains a cornerstone of kidney stone management, relying heavily on the surgeon’s visual ability to identify and clear all stone fragments. Lu et al. present a deep learning-based model for automated intraoperative kidney stone segmentation, marking an important step towards AI-assisted endourological surgery. Miernik places this works within the broader evolution of AI in urology, highlighting prior successes in computer vision, intraoperative guidance, and surgical performance assessment. Examples include CystoNet for bladder tumour detection during cystoscopy, machine-learning models predicting surgical outcomes in robotic surgery, and AI-driven stone detection on CT imaging. These developments underscore the increasing credibility and maturity of AI applications in urological practice. The study by Lu et al. is notable for its scale and rigour. The model was trained on over 21,000 endoscopic image frames from ureteroscopy procedures and demonstrated excellent accuracy, achieving a Dice similarity coefficient of approximately 0.97. Importantly, its performance was consistent across different ureteroscope types and was directly benchmarked against expert urologists. The AI matched or exceeded the segmentation accuracy of most human experts, demonstrating its ability to cope with the visual complexity of live surgery. Miernik emphasises the potential clinical impact of such technology. Real-time stone segmentation could act as a “second set of eyes,” reducing residual fragments, improving stone-free rates, and enhancing trainee education. While challenges remain (such as external validation, robustness under poor visual conditions, and seamless integration into operating theatres) the editorial concludes that this work represents a significant advance towards data-driven, AI-assisted endourological surgery, with promising implications for patient outcomes and surgical quality.
The emerging role of artificial intelligence in endourology
Reviewed by Asif H Ansari
Seeing beyond the scope: the emerging role of artificial intelligence in endourology.
