Scholarlink

Cervical cancer is a serious problem that affects more than half a million women around the world. For the treatment of this disease, the prediction of the prognosis or the optimization of medical intervention, detection at an early stage is essential. This research proposes a solution based on the detection of this disease through the processing of images taken from the cervix for diagnosis using machine learning architectures, specifically convolutional neural networks to help experts in its early detection. We worked with images from the Intel ODT for Cervix Images dataset, which were improved through corresponding preprocessing. Finally, the precision and performance of the implemented architectures were compared. These were VGG16, Inception V3 and XceptionV3, analyzing the efficiency of the results with which it was possible to observe that the best base to the results in terms of precision is InceptionV3.

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

Classifier for the Detection and Classification of Cervical Cancer Using Machine Learning Techniques

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • José Guadalupe Domínguez García,
  • Julio Cesar Ponce Gallegos,
  • Ángel Eduardo Muñoz Zavala,
  • Alejandro Padilla Diaz

摘要

Cervical cancer is a serious problem that affects more than half a million women around the world. For the treatment of this disease, the prediction of the prognosis or the optimization of medical intervention, detection at an early stage is essential. This research proposes a solution based on the detection of this disease through the processing of images taken from the cervix for diagnosis using machine learning architectures, specifically convolutional neural networks to help experts in its early detection. We worked with images from the Intel ODT for Cervix Images dataset, which were improved through corresponding preprocessing. Finally, the precision and performance of the implemented architectures were compared. These were VGG16, Inception V3 and XceptionV3, analyzing the efficiency of the results with which it was possible to observe that the best base to the results in terms of precision is InceptionV3.