AI and Ultrasound in Urolithiasis: From Stone Age to Digital Age

The integration of artificial intelligence (AI) in urology, particularly in managing urolithiasis, has significantly evolved since the 1990s. This review examines the history and future of AI in this field, emphasizing how these technologies have transformed patient care. Four primary domains have emerged in AI research related to urolithiasis. The first is machine learning (ML), which uses statistical algorithms to predict clinical outcomes, such as stone dimensions, composition, and the likelihood of spontaneous stone passage based on data from CT and ultrasound images. The second domain, deep learning (DL) and artificial neural networks (ANNs), enables computers to mimic human cognitive functions, allowing for the identification of stones and understanding genetic factors influencing stone formation. The third domain, computer vision (CV), involves analyzing images to determine kidney stone composition through digital photographs and videos. Finally, natural language processing (NLP) has been instrumental in scanning electronic medical records and radiologic reports, facilitating large-scale studies in urolithiasis. Notably, the introduction of ChatGPT in November 2022 has revolutionized patient communication, providing accurate and comprehensive patient-centric instructions and enhancing interactions regarding urolithiasis. As AI technology continues to progress, the focus is shifting from merely predicting clinical outcomes to fostering meaningful patient engagement. This evolution signifies a rich history of AI in urology, heralding a future where precision and patient-centered care will be at the forefront of treatment for urolithiasis.