Scholarlink

<p>The fusion of Artificial Intelligence (AI) and blockchain is rapidly advancing the Internet of Medical Things (IoMT) in healthcare, covering diagnosis to rehabilitation. However, centralized IoMT data transmission raises privacy concerns and struggles with diverse patient data. As IoMT data sharing grows, privacy, security, and resource limitations become the key concerns. With patient conditions varying widely, standard approaches must be revised to address customized healthcare needs. Thus, a blockchain-driven customized federated learning (FL) framework for securing IoMT is proposed to address these challenges. The framework enables clients to train customized models without exposing their data to central servers. It enhances data efficiency and security by incorporating Edge Computing gateway devices for data collection and preprocessing. The model partitioning method used in the work enables collaborative training across IoMT devices while preserving patient data privacy. The experimental evaluation is conducted on 2D Colon Pathology, Breast Tumor, and CIFAR-10 datasets, demonstrating the reliability, performance, and adherence to the security and privacy standards of the proposed framework. Security assessments validate the proposed framework’s resilience against threats, highlighting its effectiveness in safeguarding medical data in IoMT ecosystems.</p>

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 secure and efficient framework for internet of medical things through blockchain driven customized federated learning

  • 期刊论文

  • 发表日期：

    2025年2月25日

查看全文
  • Abdul Mazid,
  • Sheeraz Kirmani,
  • Manaullah Abid,
  • Vijayant Pawar

摘要

The fusion of Artificial Intelligence (AI) and blockchain is rapidly advancing the Internet of Medical Things (IoMT) in healthcare, covering diagnosis to rehabilitation. However, centralized IoMT data transmission raises privacy concerns and struggles with diverse patient data. As IoMT data sharing grows, privacy, security, and resource limitations become the key concerns. With patient conditions varying widely, standard approaches must be revised to address customized healthcare needs. Thus, a blockchain-driven customized federated learning (FL) framework for securing IoMT is proposed to address these challenges. The framework enables clients to train customized models without exposing their data to central servers. It enhances data efficiency and security by incorporating Edge Computing gateway devices for data collection and preprocessing. The model partitioning method used in the work enables collaborative training across IoMT devices while preserving patient data privacy. The experimental evaluation is conducted on 2D Colon Pathology, Breast Tumor, and CIFAR-10 datasets, demonstrating the reliability, performance, and adherence to the security and privacy standards of the proposed framework. Security assessments validate the proposed framework’s resilience against threats, highlighting its effectiveness in safeguarding medical data in IoMT ecosystems.