To assess the performance of UrologiQ AI in the triage of 3D NCCT KUB scans in a clinical setting, focusing on the detection,
quantification, and localization of urinary calculi.

The use of artificial intelligence (AI) in the medical field has made significant progress and offers promising solutions for the diagnosis and treatment of various diseases. One area where AI shows significant potential is the treatment of kidney stone disease. Millions of people around the world are affected by kidney stones, which can cause severe pain and complications if left untreated. 

Integrating AI technologies into the healthcare system to detect and treat kidney stones promises to improve patient outcomes and reduce healthcare costs. However, it is imperative to consider the ethical implications associated with the use of AI in this context. 

AI represents a useful tool that offers numerous conveniences to urologists, which explains why it has gained importance in the search for a perfect treatment for stone diseases. By using it as an alternative or supplement to the already existing data, the effectiveness of diagnosis and therapy can be increased.

A retrospective cross-sectional analysis was conducted using 95 3D NCCT KUB scans obtained from a diagnostic center for
algorithm testing. UrologiQ employs AI algorithms to classify CT KUB scans as either calculus-positive or calculus-negative and
also detects, measures (in terms of volume), and locates urinary calculi within these scans. 

The scans were processed using the UrologiQ AI algorithm, and the results were documented. Stone masks were generated using ITK snap, and volume measurements were recorded. A radiologist with over 5 years of experience compared the UrologiQ analysis with their own interpretation of the CT KUB scans. Additionally, volume measurements were compared, and other findings such as the severity of hydronephrosis/hydroureteronephrosis and the presence of perinephric fat strandings were also assessed.

The AI algorithm demonstrated a sensitivity, specificity, and accuracy exceeding 90% in the triage and detection of urinary
calculi. The correlation between AI-generated volume measurements and those of the radiologist exceeded 93%.

In conclusion, UrologiQ AI exhibited promising performance in the triage, detection, quantification, and localization of urinary
calculi in 3D NCCT KUB scans.

The sample size was relatively small and excluded patients under the age of 18. The model may encounter challenges in
distinguishing urinary calculi from renal artery calcifications, cystic calcifications, and peripheral calculi, leading to some false
positives.
UrologiQ AI may also face limitations in analyzing cases involving ascites, polycystic kidneys, and pneumoperitoneum.
Moreover, congenital anomalies such as horseshoe kidney and duplex ureter were not included in the study.

Anastasiadis A, Koudonas A, Langas G, et al. Transforming urinary stone disease management by artificial intelligence-based methods: A comprehensive review. Asian J Urol. 2023;10(3):258-274. doi:10.1016/j.ajur.2023.02.002
Caglayan A, Horsanali MO, Kocadurdu K, Ismailoglu E, Guneyli S. Deep learning model-assisted detection of kidney stones on computed tomography. Int Braz J Urol. 2022;48(5):830-839. doi:10.1590/S1677-5538.IBJU.2022.0132
Hameed BMZ, S Dhavileswarapu AVL, Raza SZ, et al. Artificial Intelligence and Its Impact on Urological Diseases and Management: A Comprehensive Review of the Literature. J Clin Med. 2021;10(9):1864. Published 2021 Apr 26. doi:10.3390/jcm10091864