HF01-21: From Stone Age to the Digital Age: The History and Future of Artificial Intelligence and ChatGPT in Urolithiasis
Friday, May 3, 2024 4:50 PM to 4:57 PM · 7 min. (US/Central)
206
Abstract
History of Urology Forum
Information
Full Abstract and Figures
Author Block
Mark N Alshak*, Isabella S Florissi, Baltimore, MD, William Du Comb, Burlington, MA
Introduction
To present the history of artificial intelligence (AI) in urology in urolithiasis and understand how AI technologies have transformed the landscape of care and the future of AI in urologic stone disease.
Methods
A comprehensive literature review via PubMed was used to find journal articles and texts describing the application of AI to urolithiasis.
Results
The first publications exploring the role of AI in the identification and treatment of urolithiasis dates to the 1990's and early 2000's. Over time, four domains of study have emerged. The first domain, machine learning (ML), which involves the development of statistical algorithms and models to analyze and draw inferences from patterns in data, has been applied to predict clinical outcomes for urolithiasis. In early studies, scientists predicted stone dimensions from computed tomography and ultrasound images, stone composition, spontaneous stone passage, and outcomes of endourological procedures. The second domain is comprised of deep learning (DL) and artificial neural networks (ANNs) through which computers learn to process data in a way that is inspired by the human brain. These algorithms have been applied to identify stones, explore genetic polymorphisms in stone disease, describe patient habits related to the development of stones, and predicting stone free status after extracorpeal shock wave lithotripsy. The third domain is computer vision (CV) through which computers derive information from the interpretation of images and videos. CV has been applied to predict kidney stone composition from digital photographs and video footage of stones. The last domain is natural language processing (NLP), through which machines interpret and respond to text or voice data. Historically, it has been used to scan the electronic medical record or radiologic reports to perform large scale urolithiasis studies. Most famously, ChatGPT, introduced in November of 2022, has yielded tremendous advancements in patient communication. ChatGPT can produce thorough and accurate patient-centric instructions. Moreover, ChatGPT has been shown to be accurate in patient-forward interactions about urolithiasis.
Conclusions
The history and future of AI in urology, particularly ChatGPT, underscore the profound impact of these technologies in urologic care. As technology continues to advance, studies have shifted from predicting clinical outcomes to having patient-forward interactions. AI in urology has a rich history and is expected to usher in a new era of precision and patient-centered care.
Source Of Funding
None.
Author Block
Mark N Alshak*, Isabella S Florissi, Baltimore, MD, William Du Comb, Burlington, MA
Introduction
To present the history of artificial intelligence (AI) in urology in urolithiasis and understand how AI technologies have transformed the landscape of care and the future of AI in urologic stone disease.
Methods
A comprehensive literature review via PubMed was used to find journal articles and texts describing the application of AI to urolithiasis.
Results
The first publications exploring the role of AI in the identification and treatment of urolithiasis dates to the 1990's and early 2000's. Over time, four domains of study have emerged. The first domain, machine learning (ML), which involves the development of statistical algorithms and models to analyze and draw inferences from patterns in data, has been applied to predict clinical outcomes for urolithiasis. In early studies, scientists predicted stone dimensions from computed tomography and ultrasound images, stone composition, spontaneous stone passage, and outcomes of endourological procedures. The second domain is comprised of deep learning (DL) and artificial neural networks (ANNs) through which computers learn to process data in a way that is inspired by the human brain. These algorithms have been applied to identify stones, explore genetic polymorphisms in stone disease, describe patient habits related to the development of stones, and predicting stone free status after extracorpeal shock wave lithotripsy. The third domain is computer vision (CV) through which computers derive information from the interpretation of images and videos. CV has been applied to predict kidney stone composition from digital photographs and video footage of stones. The last domain is natural language processing (NLP), through which machines interpret and respond to text or voice data. Historically, it has been used to scan the electronic medical record or radiologic reports to perform large scale urolithiasis studies. Most famously, ChatGPT, introduced in November of 2022, has yielded tremendous advancements in patient communication. ChatGPT can produce thorough and accurate patient-centric instructions. Moreover, ChatGPT has been shown to be accurate in patient-forward interactions about urolithiasis.
Conclusions
The history and future of AI in urology, particularly ChatGPT, underscore the profound impact of these technologies in urologic care. As technology continues to advance, studies have shifted from predicting clinical outcomes to having patient-forward interactions. AI in urology has a rich history and is expected to usher in a new era of precision and patient-centered care.
Source Of Funding
None.
Sessions
History of Urology Forum
206