Scholarlink

Deep learning based convolutional neural networks are gaining much attention for medical image based classification and detection tasks. In this study, we have implemented MobileNet, ResNet50 and VGG19 for classifying Magnetic Resonance images (MRI) of the brain, Wireless Capsule Endoscopy (WCE) Images of colon and Fundus images of the eyes. The pretrained MobileNet model achieved classification accuracy of 99.84% and 99.64% on MRI training and test sets, respectively. Whereas the ResNet50 with transfer learning also did really well and Achieved classification accuracy of 97.70% and 93.57% on ACRIMA training and test sets, respectively, and classification accuracy of 96.25% and 94.75% on WCE training and test sets, respectively. These findings demonstrate how deep learning models may be used to increase medical image analysis diagnostic efficiency and accuracy.

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

Utilizing Deep Learning for the Classification of Multiple Diseases Based on Medical Images

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Anmol Garg,
  • Ravi Kumar,
  • Akshay Kanwar

摘要

Deep learning based convolutional neural networks are gaining much attention for medical image based classification and detection tasks. In this study, we have implemented MobileNet, ResNet50 and VGG19 for classifying Magnetic Resonance images (MRI) of the brain, Wireless Capsule Endoscopy (WCE) Images of colon and Fundus images of the eyes. The pretrained MobileNet model achieved classification accuracy of 99.84% and 99.64% on MRI training and test sets, respectively. Whereas the ResNet50 with transfer learning also did really well and Achieved classification accuracy of 97.70% and 93.57% on ACRIMA training and test sets, respectively, and classification accuracy of 96.25% and 94.75% on WCE training and test sets, respectively. These findings demonstrate how deep learning models may be used to increase medical image analysis diagnostic efficiency and accuracy.