Breast cancer is one of the leading causes of death among women in India as well as other countries which stresses the importance of early diagnosis. This study also presents an automated method to identify breast cancer using histopathological images and applying deep learning strategies. We fine-tuned ResNet50, AlexNet, and EfficientNet_B0 on the BreakHis dataset and took their predictions through an ensemble model to increase accuracy. The ensemble model gave a training accuracy of 94.20% while the testing accuracy was 92.3%, better than the individual models. This approach shows how deep learning can assist medical personnel in India to make accurate and timely diagnosis. Future research will involve the implementation of the ensemble model in low-power devices to facilitate real-time diagnosis on the patient’s side or in the primary healthcare unit.

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 Breast Cancer Detection Using Ensemble Method with Histopathological Images

  • Appu. S. Chalawadi,
  • Tushar Kattishettar,
  • Gautam Narajji,
  • Daksh Porwal,
  • Uday Kulkarni,
  • Sunil Gurlahosur

摘要

Breast cancer is one of the leading causes of death among women in India as well as other countries which stresses the importance of early diagnosis. This study also presents an automated method to identify breast cancer using histopathological images and applying deep learning strategies. We fine-tuned ResNet50, AlexNet, and EfficientNet_B0 on the BreakHis dataset and took their predictions through an ensemble model to increase accuracy. The ensemble model gave a training accuracy of 94.20% while the testing accuracy was 92.3%, better than the individual models. This approach shows how deep learning can assist medical personnel in India to make accurate and timely diagnosis. Future research will involve the implementation of the ensemble model in low-power devices to facilitate real-time diagnosis on the patient’s side or in the primary healthcare unit.