Scholarlink

<p>As healthcare rapidly digitalizes, cloud computing has emerged as the backbone of modern medical data storage and processing, owing to its scalability, cost-efficiency, and universal accessibility. Despite these advantages, centralized cloud infrastructures inherently suffer from significant drawbacks, particularly in protecting patient privacy, enforcing transparent access control, and preventing unauthorized data manipulation—concerns that are critically amplified in the context of sensitive medical information. Following this, federated learning (FL) has grown as an open approach that enables collaborative model training across hospitals without sharing raw data without affecting the privacy of primary data. However, FL is insufficient for high-assurance environments such as hospitals due to its lack of verifiability, dynamic access control, and reliable audit trials. In response to these gaps, QuickMedBlock, a novel blockchain-based federated learning framework specifically designed for cloud-enabled hospital information systems is proposed. Blockchain smart contracts, Decentralized Identity (DID), and a hybrid Attribute- and Role-Based Access Control (ABAC + RBAC) model are seamlessly integrated in the framework to provide fine-grained, auditable, real-time authorization of healthcare entities. The framework leverages Zero-Knowledge Proofs (ZKP) for privacy-preserving verification, multi-signature authentication for critical actions, and tokens and incentivize meaningful institutional participation. Performance evaluations reveal that QuickMedBlock achieves low latency (&lt; 1300&#xa0;ms), high throughput (8.9 TPS), and gas-efficient smart contract execution—demonstrating real-world feasibility. It provides better privacy, dynamic policy enforcement, decentralized trust, and regulatory compliance than previous solutions.</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

QuickMedBlock: A framework for enhanced attribute-based access control using blockchain for EHR in cloud

  • 期刊论文

  • 发表日期：

    2025年7月31日

查看全文
  • Aarti Punia,
  • Preeti Gulia,
  • Nasib Singh Gill,
  • Umesh Kumar Lilhore,
  • Sarita Simaiya,
  • Roobaea Alroobaea,
  • Hamed Alsufyani,
  • Abdullah M. Baqasah

摘要

As healthcare rapidly digitalizes, cloud computing has emerged as the backbone of modern medical data storage and processing, owing to its scalability, cost-efficiency, and universal accessibility. Despite these advantages, centralized cloud infrastructures inherently suffer from significant drawbacks, particularly in protecting patient privacy, enforcing transparent access control, and preventing unauthorized data manipulation—concerns that are critically amplified in the context of sensitive medical information. Following this, federated learning (FL) has grown as an open approach that enables collaborative model training across hospitals without sharing raw data without affecting the privacy of primary data. However, FL is insufficient for high-assurance environments such as hospitals due to its lack of verifiability, dynamic access control, and reliable audit trials. In response to these gaps, QuickMedBlock, a novel blockchain-based federated learning framework specifically designed for cloud-enabled hospital information systems is proposed. Blockchain smart contracts, Decentralized Identity (DID), and a hybrid Attribute- and Role-Based Access Control (ABAC + RBAC) model are seamlessly integrated in the framework to provide fine-grained, auditable, real-time authorization of healthcare entities. The framework leverages Zero-Knowledge Proofs (ZKP) for privacy-preserving verification, multi-signature authentication for critical actions, and tokens and incentivize meaningful institutional participation. Performance evaluations reveal that QuickMedBlock achieves low latency (< 1300 ms), high throughput (8.9 TPS), and gas-efficient smart contract execution—demonstrating real-world feasibility. It provides better privacy, dynamic policy enforcement, decentralized trust, and regulatory compliance than previous solutions.