Scholarlink

<p>Knee osteoarthritis, a degenerative joint disease, results in the gradual deterioration and eventual loss of knee cartilage, causing pain, stiffness, and difficulties in movement. Initiating treatment based solely on symptoms can lead to irreversible joint changes, emphasizing the importance of early detection. This study applies various machine learning models to expedite and refine the detection of knee osteoarthritis using knee X-ray images. The study analyzed several commonly utilized models, including Convolutional Neural Networks, Spiking Neural Networks, Google Teachable Machine, Support Vector Machines, and a Convolutional Neural Network enhanced with a pre-trained VGG16. The models were trained using a collection of images, each representing a different stage of osteoarthritis according to the Kellgren-Lawrence scale. Among these models, the Google Teachable Machine demonstrated the greatest capability in analyzing knee X-ray images and categorizing them into different severity grades. While the accuracies of these models are not yet sufficient for clinical application, this approach shows potential as a clinical tool. With further improvement, it could enhance treatment effectiveness by enabling proactive management of knee osteoarthritis, even before symptoms appear.</p>

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

Machine-learning-based diagnosis and progression analysis of knee osteoarthritis

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年3月13日

查看全文
  • Rejath Jose,
  • Nicholas Lewis,
  • Zain Satti,
  • Robert Steinberg,
  • Alec Toufexis,
  • Daniel Nasef,
  • Milan Toma

摘要

Knee osteoarthritis, a degenerative joint disease, results in the gradual deterioration and eventual loss of knee cartilage, causing pain, stiffness, and difficulties in movement. Initiating treatment based solely on symptoms can lead to irreversible joint changes, emphasizing the importance of early detection. This study applies various machine learning models to expedite and refine the detection of knee osteoarthritis using knee X-ray images. The study analyzed several commonly utilized models, including Convolutional Neural Networks, Spiking Neural Networks, Google Teachable Machine, Support Vector Machines, and a Convolutional Neural Network enhanced with a pre-trained VGG16. The models were trained using a collection of images, each representing a different stage of osteoarthritis according to the Kellgren-Lawrence scale. Among these models, the Google Teachable Machine demonstrated the greatest capability in analyzing knee X-ray images and categorizing them into different severity grades. While the accuracies of these models are not yet sufficient for clinical application, this approach shows potential as a clinical tool. With further improvement, it could enhance treatment effectiveness by enabling proactive management of knee osteoarthritis, even before symptoms appear.