MP07-10 UNDERSTANDING 3D BIOMINERALIZATION OF HUMAN KIDNEY STONES WITH ARTIFICIAL INTELLIGENCE

Le résumé fourni par la source

You have accessJournal of UrologySurgical Technology & Simulation: Artificial Intelligence I (MP07)1 May 2024MP07-10 UNDERSTANDING 3D BIOMINERALIZATION OF HUMAN KIDNEY STONES WITH ARTIFICIAL INTELLIGENCE Ava Mousavi, Jeff Gelb, Tianzhu Qin, Gerard Wong, Katarzyna Matusik, Michael Lun, Sheraz Gul, Frances Su, David Vine, Wenbing Yun, and Kymora B. Scotland Ava MousaviAva Mousavi , Jeff GelbJeff Gelb , Tianzhu QinTianzhu Qin , Gerard WongGerard Wong , Katarzyna MatusikKatarzyna Matusik , Michael LunMichael Lun , Sheraz GulSheraz Gul , Frances SuFrances Su , David VineDavid Vine , Wenbing YunWenbing Yun , and Kymora B. ScotlandKymora B. Scotland View All Author Informationhttps://doi.org/10.1097/01.JU.0001008728.41882.d7.10AboutPDF ToolsAdd to favoritesDownload CitationsTrack CitationsPermissionsReprints ShareFacebookLinked InTwitterEmail Abstract INTRODUCTION AND OBJECTIVE: Kidney stones are small yet complex structures whose formation remains largely a mystery. Clarifying their structure can unveil crucial characteristics driving their genesis. However, traditional methods for microstructure analysis can be time-consuming and limited in their capabilities. Some of our preliminary work and other recent studies have indicated that bacterial species may have a role in kidney stone formation. This study introduces an innovative approach that leverages artificial intelligence (AI) to significantly enhance the visualization of kidney stone microstructure, with the goal of shedding light on the mechanisms behind their biomineralization. METHODS: Human-derived kidney stone samples were collected and scanned through micro-computed tomography (micro-CT). Dragonfly Object Research System (ORS) software's AI-based image processing was used to visualize and separate the features of one stone slice into four segments (pore, high intensity, low intensity and medium intensity). The AI model was trained to predict that segmentation twice more on different slides. Then, this training was applied to the entire dataset of stone images. RESULTS: The AI model generated- three dimensional images unveiled the intricate microstructure of kidney stones, revealing details not easily discernible through conventional methods. Similar morphological patterns were seen across all stone types: distinct areas of nucleation were visualized with surrounding organized biphasic lamellation (Figure 1). The layering pattern resembled that seen in bacteria-induced biomineralization. Porous void spaces permeated throughout the stone. CONCLUSIONS: The level of detail in this technique revealed patterns of nucleation and layering that resemble microbial mat formation observed in stromatolites, suggesting microbes may play a similar role in kidney stone formation. AI image segmentation offers promising advancements in Urological research. Download PPT Source of Funding: NIH/NIDDK K08DK13248601A NIH/NCATS KL2TR001882 © 2024 by American Urological Association Education and Research, Inc.FiguresReferencesRelatedDetails Volume 211Issue 5SMay 2024Page: e108 Advertisement Copyright & Permissions© 2024 by American Urological Association Education and Research, Inc.Metrics Author Information Ava Mousavi More articles by this author Jeff Gelb More articles by this author Tianzhu Qin More articles by this author Gerard Wong More articles by this author Katarzyna Matusik More articles by this author Michael Lun More articles by this author Sheraz Gul More articles by this author Frances Su More articles by this author David Vine More articles by this author Wenbing Yun More articles by this author Kymora B. Scotland More articles by this author Expand All Advertisement PDF downloadLoading ...

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.