Slider

Yang B. et al., 2020: Artificial intelligence in the diagnosis, treatment and prevention of urinary stones

Yang B, Veneziano D, Somani BK.
Royal Hampshire Hospital, Winchester, UK.
Department of Urology and Kidney Transplant, GOM, Reggio Calabria, Italy.
University Hospital Southampton NHS Trust, UK.

Abstract

Purpose of review: There has a been rapid progress in the use of artificial intelligence in all aspects of healthcare, and in urology, this is particularly astute in the overall management of urolithiasis. This article reviews advances in the use of artificial intelligence for the diagnosis, treatment and prevention of urinary stone disease over the last 2 years. Pertinent studies were identified via a nonsystematic review of the literature performed using MEDLINE and the Cochrane database.

Recent findings: Twelve articles have been published, which met the inclusion criteria. This included three articles in the detection and diagnosis of stones, six in the prediction of postprocedural outcomes including percutaneous nephrolithotomy and shock wave lithotripsy, and three in the use of artificial intelligence in prevention of stone disease by predicting patients at risk of stones, detecting the stone type via digital photographs and detecting risk factors in patients most at risk of not attending outpatient appointments.

Summary: Our knowledge of artificial intelligence in urology has greatly advanced in the last 2 years. Its role currently is to aid the endourologist as opposed to replacing them. However, the ability of artificial intelligence to efficiently process vast quantities of data, in combination with the shift towards electronic patient records provides increasingly more 'big data' sets. This will allow artificial intelligence to analyse and detect novel diagnostic and treatment patterns in the future.
Curr Opin Urol. 2020 Sep 15. doi: 10.1097/MOU.0000000000000820. Online ahead of print. PMID: 32941256

0
 

Commenti 1

Peter Alken il Venerdì, 13 Novembre 2020 08:35

This small review touches several aspects of urolithiasis. The authors focus on the most recently published 12 papers in the last two years. The enthusiasm and optimism of the authors is refreshing. What I missed was a comment or at least a hint on what hampers the artificial intelligence to enter the handling of stone disease in a way that it replaces the traditional access in the urologist’s everyday life. This may be found in the comments to one of the three reviewed papers dealing with ESWL (1-3).
Choo et al. (2) impressed by a model which based on 15 factors had “a greater than 92% accuracy and an average ROC AUC of 0.951”. They received two enthusiastic and critical editorial comments in the same issue of the Journal of Urology. A third comment - a devastating critical one - was published a year later (4) and is worth reading for those who want to get an insight into the way machine learning data should be generated and presented. Unfortunately, I could not find a rebuttal by the authors of the original paper.


1 Mannil M, von Spiczak J, Hermanns T, et al. Prediction of successful shock
wave lithotripsy with CT: a phantom study using texture analysis. Abdom
Radiol (NY) 2018; 43:1432–1438.

2 Mannil M, von Spiczak J, Hermanns T, et al. Three-dimensional texture analysis
with machine learning provides incremental predictive information for successful
shock wave lithotripsy in patients with kidney stones. J Urol 2018;
200:829–836.

3 Choo MS, Uhmn S, Kim JK, et al. A prediction model using machine learning
algorithm for assessing stone-free status after single session shock wave
lithotripsy to treat ureteral stones. J Urol 2018; 200:1371–1377.

4 Yoshioka T, Yamamoto Y, Ikenoue T. Re: A Prediction Model Using Machine Learning Algorithm for Assessing Stone-Free Status after Single Session Shock Wave Lithotripsy to Treat Ureteral Stones. J Urol. 2019 Nov;202(5):1053-1054. doi: 10.1097/JU.0000000000000400.

Peter Alken

This small review touches several aspects of urolithiasis. The authors focus on the most recently published 12 papers in the last two years. The enthusiasm and optimism of the authors is refreshing. What I missed was a comment or at least a hint on what hampers the artificial intelligence to enter the handling of stone disease in a way that it replaces the traditional access in the urologist’s everyday life. This may be found in the comments to one of the three reviewed papers dealing with ESWL (1-3). Choo et al. (2) impressed by a model which based on 15 factors had “a greater than 92% accuracy and an average ROC AUC of 0.951”. They received two enthusiastic and critical editorial comments in the same issue of the Journal of Urology. A third comment - a devastating critical one - was published a year later (4) and is worth reading for those who want to get an insight into the way machine learning data should be generated and presented. Unfortunately, I could not find a rebuttal by the authors of the original paper. 1 Mannil M, von Spiczak J, Hermanns T, et al. Prediction of successful shock wave lithotripsy with CT: a phantom study using texture analysis. Abdom Radiol (NY) 2018; 43:1432–1438. 2 Mannil M, von Spiczak J, Hermanns T, et al. Three-dimensional texture analysis with machine learning provides incremental predictive information for successful shock wave lithotripsy in patients with kidney stones. J Urol 2018; 200:829–836. 3 Choo MS, Uhmn S, Kim JK, et al. A prediction model using machine learning algorithm for assessing stone-free status after single session shock wave lithotripsy to treat ureteral stones. J Urol 2018; 200:1371–1377. 4 Yoshioka T, Yamamoto Y, Ikenoue T. Re: A Prediction Model Using Machine Learning Algorithm for Assessing Stone-Free Status after Single Session Shock Wave Lithotripsy to Treat Ureteral Stones. J Urol. 2019 Nov;202(5):1053-1054. doi: 10.1097/JU.0000000000000400. Peter Alken
Ospite
Lunedì, 19 Aprile 2021

By accepting you will be accessing a service provided by a third-party external to https://www.storzmedical.com/

Linkedin Channel Facebook Channel Instagram Channel Twitter Youtube Channel