Scholarlink

The increasing demand for a precise and timely diagnosis of kidney abnormalities has caused a significant increase in research on counterfeit intelligence-driven frameworks. Recent advances in profound learning have opened up modern avenues for examining different kidney diseases. The present research evaluates the performance of four popular convolutional neural network (CNN) architectures—MobileNet, VGG-16, VGG-19, and ResNet50—to distinguish between normal cases and abnormal conditions such as kidney stones, tumors, and cysts. Furthermore YOLO-NAS, is employed to precisely locate kidney stones within images. The results of this study show that MobileNet outperforms the other pre-trained models, achieving an accuracy of 99.60%. Additionally, YOLO-NAS detects kidney stones with a high mean Average Precision (mAP) score. This study illustrates deep learning’s potential to facilitate clinicians for informed decision-making and improve the ongoing management of patient health.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Comparative Evaluation of Deep Learning Architectures for Multi-class Classification of Kidney Pathologies and Kidney Stone Detection

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Abhiraj Singh Thakur,
  • Nitin Gupta,
  • Preeti Soni,
  • Kuldeep Singh Jadon

摘要

The increasing demand for a precise and timely diagnosis of kidney abnormalities has caused a significant increase in research on counterfeit intelligence-driven frameworks. Recent advances in profound learning have opened up modern avenues for examining different kidney diseases. The present research evaluates the performance of four popular convolutional neural network (CNN) architectures—MobileNet, VGG-16, VGG-19, and ResNet50—to distinguish between normal cases and abnormal conditions such as kidney stones, tumors, and cysts. Furthermore YOLO-NAS, is employed to precisely locate kidney stones within images. The results of this study show that MobileNet outperforms the other pre-trained models, achieving an accuracy of 99.60%. Additionally, YOLO-NAS detects kidney stones with a high mean Average Precision (mAP) score. This study illustrates deep learning’s potential to facilitate clinicians for informed decision-making and improve the ongoing management of patient health.