Scholarlink

The brain is one of the main organs in the human body. It controls almost all the actions of a human being. Any problem with the brain can lead to even fatal consequences. One of the most dangerous problems faced by the human brain is a brain tumor. If it is not treated properly at early stages, it may lead to severe conditions. However, finding a brain tumor is a very challenging task as the brain structure is very complex. Also, a lot of image data is required to train the model to detect brain tumors. However, the hospitals are not ready to share patient data to train the model. Hence, an approach to train the model by seeing the actual dataset or simply using decentralized data from multiple clients needs to be followed. So, a federated learning methodology is utilized to train the model. The trained federated learning model has achieved a training accuracy of 91% and a test accuracy of 88%. The federated learning and transfer learning approach stands out in preserving privacy, optimizing model performance, and leading to good collaboration among healthcare entities.

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

A Federated Learning Approach Towards a Privacy-Preserving Technique for Brain Tumor Classification

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Anurag De,
  • Gautam Pal,
  • Karnam Shyam,
  • Kalakanti Pawan Tej

摘要

The brain is one of the main organs in the human body. It controls almost all the actions of a human being. Any problem with the brain can lead to even fatal consequences. One of the most dangerous problems faced by the human brain is a brain tumor. If it is not treated properly at early stages, it may lead to severe conditions. However, finding a brain tumor is a very challenging task as the brain structure is very complex. Also, a lot of image data is required to train the model to detect brain tumors. However, the hospitals are not ready to share patient data to train the model. Hence, an approach to train the model by seeing the actual dataset or simply using decentralized data from multiple clients needs to be followed. So, a federated learning methodology is utilized to train the model. The trained federated learning model has achieved a training accuracy of 91% and a test accuracy of 88%. The federated learning and transfer learning approach stands out in preserving privacy, optimizing model performance, and leading to good collaboration among healthcare entities.