Scholarlink

The study uses the Kaggle Multi Cancer Dataset to develop a machine-learning method for identifying various cancer classifications, including cervical cancer. The dataset includes labels of cervical cancer images classified as ascervix_dyk, cervix_koc, cervix_mep, cervix_pab, and cervix_sfi. The study uses two techniques for image enhancement: weak enhancement and strong augmentation. The model is trained using algorithms like Resnet34, CNN, and deep learning. The findings show the effectiveness of deep learning in interpreting medical images, improving patient care and diagnostic precision in cervical cancer treatment. The training accuracy of pretrained CNN models and the FixMatch Resnet34 pre-trained model is 93.6% and 99.8%, respectively. The validation accuracy is 94.0 and 99.4% for classification. The project aims to create trustworthy automated methods for classifying and detecting cervical cancer using pre-trained MRI.

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

Cervical Cancer Classification: Deep Learning Techniques Using FixMatch Semi-supervised Approach

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Md. Anikur Rahaman,
  • Zeshan Ahmed,
  • K. M. Safin Kamal,
  • Ahmed Wasif Reza

摘要

The study uses the Kaggle Multi Cancer Dataset to develop a machine-learning method for identifying various cancer classifications, including cervical cancer. The dataset includes labels of cervical cancer images classified as ascervix_dyk, cervix_koc, cervix_mep, cervix_pab, and cervix_sfi. The study uses two techniques for image enhancement: weak enhancement and strong augmentation. The model is trained using algorithms like Resnet34, CNN, and deep learning. The findings show the effectiveness of deep learning in interpreting medical images, improving patient care and diagnostic precision in cervical cancer treatment. The training accuracy of pretrained CNN models and the FixMatch Resnet34 pre-trained model is 93.6% and 99.8%, respectively. The validation accuracy is 94.0 and 99.4% for classification. The project aims to create trustworthy automated methods for classifying and detecting cervical cancer using pre-trained MRI.