The Role of Artificial Intelligence in Kidney Calculi Management: A Systematic Review
Code: G-1457
Authors: Zahra Abdollahi ℗, Narges Mohamadi, Hossein Zamani *
Schedule: Not Scheduled!
Tag: Intelligent Virtual Assistant
Download: Download Poster
Abstract:
Abstract
Background and aims: Kidney Stone are one of the most common urinary tract diseases, with a prevalence of 1-15% in all age groups. Recently, AI has presented innovative solutions in the management of kidney Stone, but the evidence is sparse. So, this systematic review was conducted to analyze the impact of AI in areas such as diagnosis, treatment, and prevention of kidney Stone. Method: In this systematic review, extensive searches were performed with keywords related to 'artificial intelligence' and 'kidney Calculi' in the international databases of Web of Science Core Collection, PubMed/Medline Scopus, Google_Scholar search engine, and national databases of Irandoc, SID, and Magiran. The initial search yielded 39 studies. Inclusion criteria were publication in Persian/English. Grey literature came out. The time limit wasn't applied. After eliminating duplicates and critiquing with relevant tools, at least seven studies were analyzed. Ethical considerations of non-bias were observed at the stages of selection, extraction, analysis, and classification of evidence, and the abstract was reported according to PRISMA. Results: Studies have shown that using AI algorithms can increase the accuracy of detection of kidney Stone composition to more than 90% and improve the accuracy of Fourier transform infrared spectroscopy(FTIR) analyses to detect laboratory errors in determining Calculi composition. A transferential learning model using VGG16 with Explainable AI(XAI) helps doctors accurately diagnose kidney Stone. ChatGPT is also useful in urological diagnoses but requires professional supervision. Also, using a Decision Support System(DSS) with 94.8% accuracy is effective in predicting the results of a specific treatment for kidney Stone. The UroGPT AI bot can also assist patients with kidney Stone in accessing information and managing self-care. When comparing chatbots' accuracy in categorizing foods based on oxalate content, Bard AI performed best with 84% accuracy, followed by GPT-4, Bing Chat, and GPT-3.5. Conclusion: It seems that the use of AI can help urologists provide advice, choose the type of treatment, and predict the results of the treatment method used to remove kidney Stone. It is suggested that future studies compare the accuracy of different AI models in analyzing data in this field.
Keywords
Artificial Intelligence, Kidney Calculi, Systematic
