Scholarlink

Osteoarthritis (OA) is a degenerative condition that can impact several joints in the body, including the knee, hip, spine, and hand, with knee osteoarthritis (KOA) being the most widespread form worldwide. Deep learning methods in computer-aided medical diagnosis have established the potential to achieve diagnostic performance comparable to that of expert medical professionals in identifying knee osteoarthritis. In this study, MRI scans were employed due to their high-resolution capabilities. We applied transfer learning by fine-tuning the VGG-16, EfficientNetB0, DenseNet169, and MobileNetV3Large models, which were subsequently integrated into a weighted mean ensemble method to improve performance. This approach yielded promising results, achieving an overall accuracy of 92.5%.

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

Transfer Learning-Based Feature Extraction and Weighted Mean Ensemble Approach for Automated Knee Osteoarthritis Classification Using MRI

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Punita Panwar,
  • Sandeep Chaurasia,
  • Jayesh Gangrade,
  • Ashwani Bilandi

摘要

Osteoarthritis (OA) is a degenerative condition that can impact several joints in the body, including the knee, hip, spine, and hand, with knee osteoarthritis (KOA) being the most widespread form worldwide. Deep learning methods in computer-aided medical diagnosis have established the potential to achieve diagnostic performance comparable to that of expert medical professionals in identifying knee osteoarthritis. In this study, MRI scans were employed due to their high-resolution capabilities. We applied transfer learning by fine-tuning the VGG-16, EfficientNetB0, DenseNet169, and MobileNetV3Large models, which were subsequently integrated into a weighted mean ensemble method to improve performance. This approach yielded promising results, achieving an overall accuracy of 92.5%.