Scholarlink

<p>Living with cancer profoundly impacts every aspect of a woman’s life. According to the World Health Organization, ovarian cancer (OC) has recently become one of the most prevalent gynaecologic cancers affecting women worldwide. Since OC is preventable, early detection plays a critical role in saving women’s lives. Currently, artificial intelligence (AI)-based systems are increasingly being utilized in medical decision-making. In this context, machine learning techniques offer the potential for rapid and efficient automated detection of ovarian cancer. This research focuses on leveraging transfer learning models and the advanced YOLOv8 model to develop an intelligent system for the accurate detection of ovarian cancer using medical images. The performance of the models was further enhanced using filtering techniques such as sharpening and clipping with gamma correction. Additionally, these filtered images were processed using other data preprocessing steps, including resizing and normalization. Among the various models evaluated, YOLOv8l demonstrated superior performance, achieving an accuracy of 97.20%, precision of 97.20%, F1 score of 97.19%, recall of 97.20%, and an MCC of 96.50%. The proposed model has significant potential to assist clinicians and medical practitioners in the early detection of ovarian cancer through real-time image analysis, thereby improving patient outcomes.</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

Optimizing ovarian cancer screening using YOLOv8 and transfer learning

  • 期刊论文

  • 发表日期：

    2025年6月2日

查看全文
  • Rajwinder Singh,
  • Hardeep Kaur,
  • Jyoteesh Malhotra

摘要

Living with cancer profoundly impacts every aspect of a woman’s life. According to the World Health Organization, ovarian cancer (OC) has recently become one of the most prevalent gynaecologic cancers affecting women worldwide. Since OC is preventable, early detection plays a critical role in saving women’s lives. Currently, artificial intelligence (AI)-based systems are increasingly being utilized in medical decision-making. In this context, machine learning techniques offer the potential for rapid and efficient automated detection of ovarian cancer. This research focuses on leveraging transfer learning models and the advanced YOLOv8 model to develop an intelligent system for the accurate detection of ovarian cancer using medical images. The performance of the models was further enhanced using filtering techniques such as sharpening and clipping with gamma correction. Additionally, these filtered images were processed using other data preprocessing steps, including resizing and normalization. Among the various models evaluated, YOLOv8l demonstrated superior performance, achieving an accuracy of 97.20%, precision of 97.20%, F1 score of 97.19%, recall of 97.20%, and an MCC of 96.50%. The proposed model has significant potential to assist clinicians and medical practitioners in the early detection of ovarian cancer through real-time image analysis, thereby improving patient outcomes.