1 Introduction

Since the early 21st century, significant advancements in digital technologies have revolutionized the global healthcare sector. This revolution is most notable in the transition from traditional paper records to digital ones, specifically Electronic Health Records (EHRs) [1, 2]. EHRs, structured digital repositories, store various patient data like lab results, medical history, and medication records [3]. They offer several benefits over paper records, such as reduced labor, time, and space requirements [4].

The healthcare industry, with its growing reliance on various digital applications and services, generates immense amounts of data daily. This increase in data has made robust storage and management services essential, with cloud computing emerging as a key solution [5]. Cloud Service Providers (CSPs) offer cost-effective and efficient infrastructure for data storage and processing, attracting both individual and institutional users [6, 7]. Cloud-based electronic health (eHealth) systems efficiently manage EHRs, allowing for secure data sharing across healthcare networks [8, 9].

However, the use of cloud-based eHealth systems raises significant security and privacy concerns [10]. When EHRs are stored on cloud servers, issues such as data ownership, access control, and security breaches become critical challenges. These challenges can hinder the adoption and growth of cloud-based services [11, 12].

One effective way for data security within cloud computing, encryption stands as a best technology among them. Researchers often turn to encryption technologies to secure data that is outsourced to the cloud. Among these technologies, ciphertext-policy attribute-based encryption is recognized as an effective way to implement access control on less reliable cloud storage servers. Despite the existence of numerous schemes aimed at ensuring data confidentiality and user privacy, challenges persist, particularly in the effective identification, tracking, and revocation of permissions for users who misuse their access, underscoring the need for enhanced access control systems. These systems must not only provide ease of access for authorized users but must also have stringent measures for promptly revoking access when a security breach occurs. Additionally, incorporating blockchain technology can enhance security. Blockchain’s transparent and immutable record-keeping is useful for secure and efficient data management, addressing concerns about data integrity [13–15].

Therefore, we propose a blockchain-enhanced security model for EMR, emphasizing traceable and direct access revocation. our research provides a scalable, efficient, and practical solution for secure healthcare data management. By combining blockchain’s decentralized structure with the scalability of cloud computing, our approach significantly improves EHR security protocols, ensuring robust data security and seamless performance in modern healthcare settings.

Our key contributions and innovations

1. Blockchain-Empowered Security: Our research introduces a blockchain-based method to protect privacy and secure Electronic Medical Records (EMRs). This approach leverages smart contracts to automate key management, encryption, and decryption processes, ensuring robust security and flexible access revocation.
2. Advanced Encryption with CP-ABE: We integrate Ciphertext-Policy Attribute-Based Encryption (CP-ABE) to enhance data security by embedding access control policies within the ciphertext. This ensures fine-grained access control, allowing only authorized personnel to access sensitive medical data. Additionally, the dual authentication system confirms patient keys before granting data access, with failed attempts leading to access revocation and incident logging.
3. Comprehensive Security Analysis: We performed an in-depth security analysis demonstrating the scheme’s resilience against various cyber threats and its efficiency in key management, encryption, and decryption processes.
4. Performance Analysis and Scalability: We conducted extensive performance evaluation, including detailed simulation experiments. Our results show that the scheme performs well in terms of communication and storage efficiency, handling large data volumes effectively. The assessment within a blockchain environment, focusing on throughput and delay, confirmed its effectiveness and practical applicability, making it highly suitable for modern healthcare systems.

The paper is organized as follows: Section 2 presents the related works, Section 4 details the system architecture, Section 7 covers the security model, Sections 9 and 10 present the security analysis and protocol design, while Section 11 presents simulation and experiments, and Section 14 summarizes our findings.

2 Related work

In the domain of blockchain technology for access control schemes with traceability and revocability for healthcare data management, a spectrum of research has been conducted, each contributing uniquely to the field.

T. Benil and J. Jasper’s work, investigates the utilization of blockchain for managing electronic health records (EHRs), focusing on crucial aspects of data integrity and access control. However, this study brings to light concerns regarding the scalability and computational efficiency of blockchain technologies in healthcare contexts [16]. Similarly, K. Anil and Dr. Megha Kamble’s “Health Block” presents a blockchain-based system adapted for healthcare data, prioritizing user-friendliness and data integrity. Aside from its innovative approach, the study falls short in thoroughly analyzing the scalability and transaction costs associated with such systems [17]. Maria Guzman Lizama and her team’s systematic review of the application of blockchain for medical image sharing in the cloud environment emphasizes the security benefits of this technology. But, the review underscores the necessity for further research focusing on the scalability and efficiency of blockchain in medical contexts [18].

In Integration of Blockchain and Cloud Computing in Telemedicine and Healthcare by Asma Albassam et al. offer an insightful overview of the integration potential of blockchain and cloud computing. The paper identifies an urgent need for additional research in data manipulation and secure storage patterns within these systems [19]. Shujiang Xu and her colleagues introduce a blockchain-based scheme enhancing privacy in mobile health. Despite its advancements, the research highlights a gap in understanding its scalability and practical applicability in real-world scenarios [20].

Mercy Ehiwuogwu’s investigation into the integration of blockchain technology in EHR systems is a notable contribution that discusses the enhancement of security and privacy. However, this study does not examine deeply the practical implementation challenges that such systems might face [21]. The work of Insaf Boumezbeur and Karim Zarour on hybrid encryption for healthcare data sharing in cloud environments shows promising improvements in performance. Yet, it also acknowledges the necessity for more comprehensive research into the scalability and practicality of these systems [22].

The paper “Privacy and Security of Blockchain in Healthcare” [23] offers a thorough analysis of the diverse applications of blockchain in healthcare, highlighting enhanced data integrity and patient control over their data. However, the study calls for more case studies and real-world applications to demonstrate the practicality of these blockchain solutions in healthcare settings. [24] Upadrista addresses the potential of blockchain in remote health monitoring. It brings attention to significant challenges, including scalability, energy consumption, and the need for standardized health data formats. Anandarshan K Hebballi and his team focus on enhancing patient data privacy [25]. The study suggests a need for further exploration into the computational overhead and the practicality of implementing such systems in healthcare settings.

Xiaohong Zhang, Wenqi Du, and Ata Jahangir Moshayedi present a blockchain-based security model for data storage, introducing features like traceability and revocation. However, the study recognizes the need for more analysis of the scalability and applicability of these features in real-world scenarios [26]. Zhe Liu et al. [27] contribute to the field with a user revocation scheme in CP-ABE systems. However, his research calls for more exploration of practical implementation aspects of their scheme.

Hui Cui in “An Efficient and Expressive Ciphertext-Policy Attribute-Based Encryption Scheme” addresses the privacy challenges in ABE systems. The research indicates a gap in understanding the real-world application and system integration of these schemes [28]. Takeru Naruse and his team, in their paper “Attribute-Based Encryption with Attribute Revocation and Grant Function,” introduce a proxy re-encryption-based ABE scheme. The study highlights the need for scalability research, particularly in large-scale cloud environments [29].

Humera Aqeel and Syed Taqi Ali’s in [30] “Directly Revocable Attribute-Based Encryption Scheme under Ciphertext-Policy” offers an efficient user revocation method. However, the paper acknowledges the lack of in-depth exploration into scalability and computational overhead challenges.

Shangping Wang [31] in his paper proposes a traceable CP-ABE scheme for cloud storage, focusing on collusion resistance and computational efficiency. However,The main drawback is that even when a malicious user is traced, they cannot be efficiently revoked from the crypto system. The “Traceable-then-revocable ciphertext-policy attribute-based encryption scheme” paper develop a scheme with enhanced traceability and direct user revocation. However author mentioned that system can only achieve white-box traceability, which is less robust compared to black-box traceability [32].

Yong Cheng et al.’s in “Directly Revocable Attribute-Based Encryption” for CP-ABE systems in cryptographic cloud storage focuses on reducing the data owner’s workload. Still, it recognizes the potential increase in data publication and retrieval costs as a drawback [33].

Zaid Ameen Abduljabbar’s work emphasizes the need for secure and efficient key management protocols in healthcare settings, highlighting their proposed scheme’s resilience against various security threats [34]. Samir M. Umran et al. demonstrate the integration of blockchain with their work to ensure data security and privacy in industrial settings, highlighting challenges in scalability and computational efficiency [15].

Samir and his team discusses the application of blockchain to enhance the security and reliability of their work, pointing out the need for lightweight cryptographic solutions to address resource constraints [13]. On the other hand Umran and Songfeng emphasizes in their work the importance of efficient consensus algorithms and data encryption mechanisms to secure industrial IoT data while maintaining low power consumption and high performance [14].

Similarly Vincent Omollo introduces an elliptic curve cryptography-based protocol to ensure data integrity and confidentiality in healthcare communications, addressing key concerns related to scalability and practical implementation in real-world scenarios [35]. Mustafa A. Al Sibahee focuses on enhancing security through biometric validation techniques, ensuring robust protection against unauthorized access in healthcare devices [36].

Lastly Umran presented the application of blockchain in industrial settings, emphasizing the need for scalable and efficient solutions to manage and secure data [37].

Each of these studies contributes valuable insights into the evolving field of blockchain and encryption technologies in healthcare data management, highlighting both their potential and the challenges that need to be addressed.

Despite the valuable contributions of these studies, several key limitations and gaps can be identified:

1. Scalability Concerns: Many existing schemes face scalability challenges when applied to large-scale healthcare systems or environments with high data volumes and numerous users. Further research is needed to ensure the practical scalability of these solutions in real-world healthcare settings.
2. Computational Overhead: Some of the proposed schemes introduce significant computational overhead, which may hinder their performance and efficiency, particularly in resource-constrained environments or with time-sensitive healthcare applications.
3. Practical Implementation Challenges: While many studies provide theoretical frameworks or simulations, there is a lack of comprehensive guidance and practical implementations to facilitate the seamless integration of these solutions into existing healthcare information systems and workflows.
4. Interoperability and Data Standardization: The integration of blockchain technology and secure access control mechanisms with diverse electronic health record (EHR) systems and healthcare data formats remains a challenge, highlighting the need for standardization and interoperability efforts.
5. User Acceptance and Adoption: The successful adoption of these solutions in healthcare settings requires addressing user acceptance factors, such as ease of use, training, and change management strategies, which have received limited attention in existing research.

Our research aims to address these gaps by proposing a scalable, efficient, and practical blockchain-based solution for secure healthcare data management, with a focus on traceability, direct access revocation, and seamless integration with existing healthcare systems and workflows.

3 Preliminaries

In this section, we provide an overview of the essential concepts and technologies used in our research, which are crucial for understanding the security model and proof of our proposed scheme.

3.5 Notation and symbols

This section introduces the important symbols, notations, and cryptographic assumptions used throughout the paper as shown in Table 1.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Notation and symbols.

https://doi.org/10.1371/journal.pone.0307039.t001

4 The system architecture of data access control

Our system design introduces a secure, privacy data exchange framework, detailed in the system architecture depicted in Fig 1. This framework comprises nine key components, each outlined below.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. System design.

https://doi.org/10.1371/journal.pone.0307039.g001

5 System flow

5.2 Case 1: Data accessing and storing

When a patient visits a hospital and hospital wants to store or access Electronic Medical Records (EMRs) below steps will be taken.

Rules of Authentication and Data Retrieval:

1. Request Submission: When hospital staff need to access a patient’s Electronic Medical Records (EMRs), they initiate the process by submitting their attributes to the Service management system (SMS) with Patients data. This is the first step in ensuring secure access to sensitive medical data.
2. Request Queue: Submitted requests undergo initial verification by the SMS, where the medical practitioner’s public key (Pk) is matched against their ID in local databases. Verified requests are then added to a queue of unprocessed requests, ensuring no request is missed.
3. Authentication and Data Retrieval:
   • Forwarding to Authentication Center (AC): The queued requests are forwarded to the AC, where a smart contract verifies the user’s details and attributes against the blockchain data.
   • Fetching Encrypted EMRs: Upon successful authentication, the smart contract fetches the encrypted Electronic Medical Records (EMRs) from the Cloud Service Provider (CSP), as illustrated in Fig 2.
   • Generating and Verifying Hashes: Before storing or retrieving the data, a cryptographic hash of the data is generated using the hash function H : {0, 1}* → G. This hash is stored along with the encrypted data. When the data is retrieved, the hash is recalculated and compared to the stored hash to ensure the data has not been tampered with. This process ensures data integrity and authenticity, protecting the stored data from potential manipulation and unauthorized changes.
   • Transmitting Decryption Parameters: The smart contract transmits the decryption parameters to the authorized user, ensuring only authenticated users can access the sensitive data.
4. Data Decryption Request: Simultaneously, the AC sends intermediary parameters back to the SMS to monitor and control the decryption activities, enhancing overall data access security.
5. Data Decryption and Access: With the received decryption parameters and their private key, users can decrypt and access the EMRs, ensuring the confidentiality of patient records during access.
6. Access Denial and Logging: In the event of a mismatch or unauthorized access attempt, the SMS denies access and logs the incident. These logs, stored on the CSP, are crucial for tracking potential breaches and maintaining data security.
7. Data Storage Process:EMRs created and stored by the hospital are encrypted using Attribute-Based Encryption (ABE) by the AC after verification with the patient’s master key (MK). The encrypted records are then hashed, and both the encrypted data and its hash are securely stored on the CSP. This process ensures that any unauthorized changes to the data can be detected.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. UML diagram of case 1 and case 2.

https://doi.org/10.1371/journal.pone.0307039.g002

6 Protocol illustration

We employ a robust and secure encryption algorithm, Ciphertext-Policy Attribute-Based Encryption (CP-ABE), integrated with blockchain technology, to ensure the confidentiality, integrity, and traceability of Electronic Health Records (EHRs). The Security Algorithm Framework consists of cryptographic operations for secure and traceable EHR management in cloud systems (as shown in Table 2). Our implementation includes the following key components:

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Security algorithm framework.

https://doi.org/10.1371/journal.pone.0307039.t002

7 Security model

In this section, we outline the IND-CPA (Indistinguishability under Chosen Plaintext Attack) cryptographic security model. The model involves a protocol between an adversary, denoted as , and a challenger, denoted as . The steps of the protocol are as follows:

1. Initial Step: The opponent selects a specific access structure T* and compiles a group of revoked users R*. This data is then passed to the challenger .
2. Setup Phase: The challenger generates a MK and a PK. The PK is shared with , while the MK is kept secret.
3. Phase 1: requests multiple secret keys SK_u that correspond to a predefined list L*. produces these keys using a designated key generation algorithm.
4. Challenge Step: presents two messages of equal size, M₁ and M₂, to . chooses a random bit u from {0, 1} and encrypts one of the messages using T* and R*. The ciphertext C_p is then given to .
5. Phase 2: This phase is similar to the First Phase, with continuing to request secret keys.
6. Prediction Step: predicts the value of u as u′. The adversary’s success is defined if u′ = u, with the probability given by .

The security of our model is considered robust if adversaries constrained to polynomial time have but a negligible success probability in the aforementioned game, thereby affirming the system’s resilience against plain text disclosure attacks.

8 Security proof

Consider a scenario where an adversary, denoted as , successfully compromises the robustness of an encryption protocol within the IND-CPA security. In response, we formulate an algorithm, , that utilizes to solve the DBDH problem. This is achieved by verifying whether e(g, g)^z is identical to e(g, g)^abc.

The process is structured as follows:

1. Initial Step: Adversary picks a specific access structure, , and a group of revoked users, . These selections are forwarded to the challenger, .
2. Preparation Phase: begins by executing the initial setup algorithm. It chooses a random element x from and calculates a as ab+x, which leads to y = e(g, g)^a. Subsequently, generates a series of random elements , ensuring their collective sum is b, and computes . This results in the creation of both the public key PK and the master key MK, with PK being revealed to .
3. Phase 1: Here, requests private keys for users who are neither part of nor among the revoked. In response, generates and provides the necessary keys to .
4. Challenge Stage: proposes two messages of equal length, M₁ and M₂. randomly chooses u from {0, 1} and encrypts M_u considering both and , using a random s. The resultant ciphertext, , is then handed over to .
5. Phase 2: Similar like First Phase, continues to make further private key requests, to which responds as required.
6. Concluding Guess: then attempts to deduce u and presents its guess as u′. If u′ matches u, concludes that e(g, g)^z and e(g, g)^abc are equivalent. Otherwise, infers that e(g, g)^z is a random value.

9 Security analysis of the proposed method

The security of electronic health records (EHRs) in cloud environments is of paramount importance. Our proposed algorithm integrates blockchain and smart contracts with cloud-based healthcare Service Management Systems (SMS) to ensure robust security measures. This section presents a detailed security analysis, demonstrating the resilience of our scheme against various cyber threats.

10 Protocol design

11 Simulation experiment

This part of the document evaluates the performance of the suggested framework. Performance metrics were gathered by conducting 700 iterations of linear pairing and exponentiation calculations on a system equipped with an Intel Core i7 processor clocked at 2.7 GHz, running a 64-bit version of Windows 10, and 32 GB of RAM. The selection of Ethereum(Ganache) and Truffle Node for the data environment and for implementation was motivated by its simplified verification process.

In evaluating the performance of networked systems, particularly those based on blockchain technology, throughput and delay serve as crucial metrics. Throughput is the measure of how many transactions a system can handle per second, while delay denotes the time taken for data to be transmitted between nodes, often described as latency. We assessed the system’s read and write performance by varying the transaction load from 100 to 700 transactions per second as shown in Table 3.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Selected transaction data.

https://doi.org/10.1371/journal.pone.0307039.t003

The resulting graphs Figs 3 and 4 show the read and write performances. In the read performance graph, throughput is represented by the blue line, which shows an increase with the number of transactions. The cyan dashed line indicates the read delay, showing a trend that suggests an increase in latency as transaction volume rises. For write performance, represented by the green line, throughput initially increases with the number of transactions, peaks, and then slightly decreases. The orange dashed line reflects write delay, illustrating a rise in latency with increased transaction volume.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Read performance and throughput.

https://doi.org/10.1371/journal.pone.0307039.g003

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Write performance and throughput.

https://doi.org/10.1371/journal.pone.0307039.g004

On the other hand The comparison results for time overhead between our proposal and other works [26, 31, 32] are shown in Figs 5 and 6 which shows the system performance in terms of encryption and decryption times. Notably, both the encryption and decryption times exhibit an upward trend as the number of transactions increases but still our system performance is better then other compared systems. Fig 7 shows our smart contract cost in wei which we have implemented by using solidity truffle environment and ganache Ethereum simulator.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. Encryption performance.

https://doi.org/10.1371/journal.pone.0307039.g005

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 6. Decryption performance.

https://doi.org/10.1371/journal.pone.0307039.g006

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 7. Smart contract cost in wei.

https://doi.org/10.1371/journal.pone.0307039.g007

These graphs show that our blockchain system offers higher throughput and lower latency during read operations compared to write operations. The read operations are handled efficiently by the system, while the write operations introduce more latency, likely due to the complexities of data validation and recording on the blockchain and CSP. The overall results point to a blockchain solution that is effective in managing high throughput and maintaining low latency, which aligns with the operational requirements of hospital data management systems.

12 Comparison

Our work introduce an integrated security framework that significantly enhances the protection and accessibility of Electronic Health Records (EHRs) through the synergy of blockchain, smart contract and cloud computing technologies. This work transcends previous research by delivering a more scalable, efficient, and practical solution for healthcare data management, incorporating advanced traceability and direct access revocation mechanisms. The proposed model not only addresses current challenges in data security and privacy but also demonstrates a clear, implementable pathway towards improving healthcare information systems, marking a pivotal advancement in secure digital health records management, as shown in Table 4.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Comparison of our system with others.

https://doi.org/10.1371/journal.pone.0307039.t004

13 Future works

In future work, we aim to explore the integration of advanced machine learning algorithms with our blockchain-cloud-based framework to enhance the predictive analysis and personalization of healthcare services. This will involve developing AI-driven models for real-time health data analysis, leading to more accurate and timely disease diagnosis and treatment recommendations. Additionally, we plan to expand the scope of our platform to include interoperability with a wider range of electronic health record systems and medical devices, ensuring seamless data exchange across different healthcare providers and platforms. Another key area of focus will be enhancing the security features of our system by incorporating newer cryptographic techniques and exploring quantum-resistant algorithms to safeguard against evolving cyber threats. Furthermore, we intend to conduct extensive field trials to validate the efficacy of our proposed system in diverse healthcare settings, ranging from urban hospitals to remote healthcare units, thus broadening its applicability and impact.

14 Conclusion

It is a huge challenge to take secure storage and sharing of Medical data among different cloud-based hospital systems. We proposed a robust algorithm that integrates blockchain technology and smart contracts with cloud computing to ensure secure and authenticated access to health data. The proposed solution addresses critical concerns in the healthcare sector, including data privacy, security, and integrity, by leveraging the decentralized nature of blockchain and the scalability of cloud computing. The results demonstrate that this integrated approach not only enhances the security of sensitive health data but also ensures seamless and efficient data access for authorized users. This algorithm sets a new standard in healthcare data management, offering a promising path forward for healthcare providers and patients alike in the age of digital health information.