Scholarlink

Brain metastases are tumors formed when malignant cells from other parts of the body spread to the brain. This condition is often a sign of advanced cancer, as the malignant cells have traveled to the brain through the blood or lymphatic system. Misdiagnosing the type of brain metastasis can hinder effective medical interventions and reduce the patient’s chances of survival. Therefore, this study aims to develop a method for automatically detecting brain metastases in MRI scans with high accuracy. We propose a new method for classifying brain metastases using the GoogLeNet network framework, incorporating the CA attention module, and replacing the activation function with Hard Swish. The results demonstrate that our proposed CNN network is beneficial for clinical diagnosis, achieving an ACC score of 0.894 and an AUC of 0.941.

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

Automated Diagnosis of Brain Metastases Using Deep Learning Technology

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Xuemin Fu,
  • Yong Wang,
  • Huixuan Wang,
  • Chuanbao Cheng,
  • Xu Qiao,
  • Yanwei Chen

摘要

Brain metastases are tumors formed when malignant cells from other parts of the body spread to the brain. This condition is often a sign of advanced cancer, as the malignant cells have traveled to the brain through the blood or lymphatic system. Misdiagnosing the type of brain metastasis can hinder effective medical interventions and reduce the patient’s chances of survival. Therefore, this study aims to develop a method for automatically detecting brain metastases in MRI scans with high accuracy. We propose a new method for classifying brain metastases using the GoogLeNet network framework, incorporating the CA attention module, and replacing the activation function with Hard Swish. The results demonstrate that our proposed CNN network is beneficial for clinical diagnosis, achieving an ACC score of 0.894 and an AUC of 0.941.