Home / Current Issue / Paper 1712231
SmartUro: An Advanced Deep Learning Framework for Automated Kidney Stone Detection and Classification in Medical Imaging
Subject area: Science,Engineering and Technology · Area of research: Machine Learning
DOI: https://doi.org/10.64388/IREV9I5-1712231
Abstract
Kidney stone disease affects 10-15% of the global population, yet traditional diagnostic methods are time-intensive and error-prone. This paper presents SmartUro, an intelligent diagnostic system leveraging YOLOv8 deep learning architecture for automated kidney stone detection across CT, MRI, X-ray, and ultrasound imaging. Our system achieves 93.0% mean Average Precision (mAP50), with 93.8% precision and 92.9% recall, processing images in under 3 seconds. Through multi-dataset integration and systematic optimization, SmartUro demonstrates clinical-grade accuracy suitable for deployment in both well-resourced centers and underserved facilities. A Streamlit-based web interface enables real-time clinical integration with comprehensive diagnostic reporting.
Keywords
Kidney Stone Detection, YOLOv8, Deep Learning, Medical Image Analysis, Automated Diagnosis
How to cite this paper
@article{1712231,
author = {Swaroop M, Nandan Gowda H M, Rakshitha P, Praveen Kamalappanavar, Satisha T},
title = {SmartUro: An Advanced Deep Learning Framework for Automated Kidney Stone Detection and Classification in Medical Imaging},
journal = {Iconic Research And Engineering Journals},
year = {2025},
volume = {9},
number = {5},
pages = {2214-2219},
issn = {2456-8880},
url = {https://www.irejournals.com/formatedpaper/1712231.pdf},
abstract = {Kidney stone disease affects 10-15% of the global population, yet traditional diagnostic methods are time-intensive and error-prone. This paper presents SmartUro, an intelligent diagnostic system leveraging YOLOv8 deep learning architecture for automated kidney stone detection across CT, MRI, X-ray, and ultrasound imaging. Our system achieves 93.0% mean Average Precision (mAP50), with 93.8% precision and 92.9% recall, processing images in under 3 seconds. Through multi-dataset integration and systematic optimization, SmartUro demonstrates clinical-grade accuracy suitable for deployment in both well-resourced centers and underserved facilities. A Streamlit-based web interface enables real-time clinical integration with comprehensive diagnostic reporting.},
keywords = {Kidney Stone Detection, YOLOv8, Deep Learning, Medical Image Analysis, Automated Diagnosis},
month = {November},
doi = {https://doi.org/10.64388/IREV9I5-1712231}
}