Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Merging RFID and Blockchain Technologies to Accelerate Big Data Medical Research Based on Physiological Signals

Merging RFID and Blockchain Technologies to Accelerate Big Data Medical Research Based on... Hindawi Journal of Healthcare Engineering Volume 2020, Article ID 2452683, 17 pages https://doi.org/10.1155/2020/2452683 Research Article Merging RFID and Blockchain Technologies to Accelerate Big Data Medical Research Based on Physiological Signals Xiuqing Chen , Hong Zhu, Deqin Geng, Wei Liu, Rui Yang, and Shoudao Li School of Medicine Information, Xuzhou Medical University, Xu Zhou 221000, China Correspondence should be addressed to Xiuqing Chen; xiuqingchen@126.com Received 4 October 2019; Revised 20 December 2019; Accepted 16 January 2020; Published 14 April 2020 Guest Editor: Liang Zou Copyright © 2020 Xiuqing Chen et al. ,is is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. ,e proliferation of physiological signals acquisition and monitoring system, has led to an explosion in physiological signals data. Additionally, RFID systems, blockchain technologies, and the fog computing mechanisms have significantly increased the availability of physiological signal information through big data research. ,e driver for the development of hybrid systems is the continuing effort in making health-care services more efficient and sustainable. Implantable medical devices (IMD) are ther- apeutic devices that are surgically implanted into patients’ body to continuously monitor their physiological parameters. Patients treat cardiac arrhythmia due to IMD therapeutic and life-saving benefits. We focus on hybrid systems developed for patient physiological signals for collection, storage protection, and monitoring in critical care and clinical practice. In order to provide medical data privacy protection and medical decision support, the hybrid systems are presented, and RFID, blockchain, and big data technologies are used to analyse physiological signals. computing, and blockchain in the medical applications 1. Introduction provide security and privacy protection for storing and ,e medical applications are continually increasing. For sharing physiological signal records. It can provide doctors handling physiological signals efficiently, specific tech- with collaboration ways through IMD [10] and RFID to help nologies, such as data gathering using RFID protocols, patients in case of emergencies mode. ,e new model based infrastructures, and distributed information storage based on blockchain can support medical background rural on blockchain frameworks, are required. ,e hospitals healthcare and analyse data for medicines and medical re- applications are adopting physiological signals to realize a search [11–15]. quicker way to visit these records. ,e physiological signals It is urgent for different research institutions to share the are responsible to offer patient care, enhance the clinical encrypted physiological signals. ,erefore, privacy and se- performances, and promote the clinical data research curity problems of physiological signals are the data owners [1–5]. and research institutions’ primary focus, when the physio- logical signals include a lot of sensitive information and the Since the fog computing solves the secure storage issues of big data in the clinical data research with minimal cost, the attackers are continually trying novel approaches to steal the fog computing technology is customizable and economical physiological signals. In order to handle these problems, the and offers infrastructure, platform, and software. Physio- medical databases adapted blockchain, and fog computing logical signals’ analysis and migration have been proposed are proposed [16, 17]. ,e medical application ecosystems for accessing and sharing physiological signal data by dif- allow the regulators to share and exchange physiological ferent research labs and health-care experts, which can signal data in Figure 1. ,e introduction of the blockchain- enable exchange of physiological signals more rapid and fog-RFID based on data ecosystems ensures that the indi- suitable by using RFID technologies and smart phone app viduals take control over physiological signal information. platforms. ,e advantages of RFID protocols [6–9], the fog ,e proposed sharing data-driven economy shares the 2 Journal of Healthcare Engineering FDA Private National health companies organization Contract Physiological research AI signal date set organisation Universities Blockchain lifedate Machine learning Insirance Hospitals companies Doctors Patients Blockchain Date Figure 1: ,e flow of data from the individuals to the companies and research institutions. physiological signals for research and commercial purposes patient management in data collection based on RFID; data in Figure 1. storage based on fog computing; and dealing with data In the paper, we protect cardiac IMD against security breaches by using blockchian. In the future, we will discuss threats by presenting a security scheme. First, we verify and the method’s applications in physiological signals research: classify the IMD’s major security attacks. Second, we in- basic research; disease management; aetiology; detection troduce blockchain and the RFID systems to extend the IMD and diagnosis; health services research; treatment devel- architecture [10] and discuss the structures of the interop- opment; and treatment evaluation. ,e possibilities of the erability in the medical environment, as shown in Figure 2. blockchian-fog-RFID method for accelerating big data ,e motivation of the blockchian-fog-RFID method for medical research in physiological signals are enormous. accelerating big data medical research based on physio- ,e paper contribution consist of four parts as follows: logical signal is as follows: the method is becoming more (1) ,e security scheme is a low energy cost RFID common due to the application of powerful computers and system in IMD. ,e applied authentication protocol the availability of physiological signals from various is implemented on the RFID circuit without energy. sources. However, although the complexity of physiolog- (2) ,e applied energy harvesting scheme uses the en- ical signals makes the complex methods particularly ap- hanced WISP, which performs computational plicable, their application of physiological signals is functions and uses the harvested energy to go beyond generally considered earlier than in other fields. Big data passive RFID tags. has become a buzzword in medical innovation. Rapid advances in artificial intelligence particularly promise to (3) ,e presented authentication protocol enables the reform medical practice from the resource allocation to the authorized health-care professionals to obtain the complex diseases’ diagnosis. However, big data brings huge access permission to cardiac IMD securely in the risks and challenges, including major questions about regular and emergency model which are determined patient privacy: the importance of fairness, consent, and according to the patient’s ability to supply valid Journal of Healthcare Engineering 3 Doctor 1 creates an order, which receives an unique ID e android medical called as hash which points to TAG device records the a record in the blockchain patient’s status Reader IMD Blockchain Lab assistant queries the blockchain to access Doctor 2 may replace the the order, does the work doctor 1 during his absence and report to the record and need to observe the Reader Reader patient’s record. Figure 2: Blockchain in the medical environment. credentials, thanks to a biometric key distribution suitable for collecting, storing, and handling heterogeneous scheme implemented. physiological signal. ,e proposed model can be used for physiological signals management. (4) ,e schemes generate and share a master key se- curely based on the physiological sets of the patient 2. Related Work collected by IMD. Monitoring and ensuring data integrity during clinical trials is not always feasible in ,e industry of healthcare has changed dramatically because current research systems. Blockchain makes the data of the boom in clinical research for physiological signal data collected immutable, traceable, and probably more sharing. We summarize the healthcare studies including trustworthy during clinical trials. We also improve physiological signal data, patient information obtained by the way we currently report adverse events. fog computing, and improvements to blockchain technol- In conclusion, we argue that the blockchain can improve ogy. ,e health-care applications of physiological signal data the management of clinical trial data, enhance trust in the adopt big data and deep learning technologies and provide clinical research process, and simplify regulatory oversight with data confidentiality and identity authentication, so as to of trials. Finally, we evaluate the security solution’s security maintain patients’ privacy. In order to more conveniently and performance. serve big data medical analysis, Rajan and Rajan [1] and ,e proposed model covers the many aspects of the Faust et al. [2] proposed the importance of medical big data health industry such as doctors, patients, and pharmacies to privacy and the impact of data analysis on medical care. insurance suppliers and government. ,e paper shows the Rajan and Rajan [1] proposed a physiological signal applications of using RFID, blockchain technologies, and fog monitoring scheme by using the Internet of ,ings (IoT). computing for storing and managing the physiological signal Our schemes use IoT to improve the access method of data. A blockchain model for sharing physiological signals is physiological signals and the real-time dynamic monitoring proposed. In the next section, the combination of block- method of the remote monitoring system, which enhances chain, RFID, and artificial intelligence (AI) technologies is the efficiency of the remote monitoring systems. Faust et al. 4 Journal of Healthcare Engineering attackers are continually trying novel approaches to steal [2] summarized the application of deep learning algorithms in physiological signals and pointed out that deep learning information. In order to meet the privacy needs and deal with the security problems, medical databases which use methods performed better than classical analysis and ma- chine classification methods for large and diverse datasets. blockchain and fog computing technology are proposed. Shanthapriya and Vaithianathan [3] proposed the health ,e enhanced trusted sharing physiological signals monitoring system for human regional network. ,e steg- model features highly secured data encryption and de- anography technologies monitor patients’ health safety and cryption schemes. ,e model requires permission from the provide patients with data confidentiality and identity au- blockchain network to share patient information among thentication. Orphanidou [4] reviewed big data applications medical staff. ,e proposed model encrypts and analyzes the physiological signals through the blockchain network, big of physiological signals, pointed out how the applications use physiological signals to provide real-time support for data analysis technology, and AI technologies. Kamel et al. [16] pointed out that blockchain technology is becoming medical decision making in both clinical and family settings, and need to be overcome in clinical practice. Tartan et al. [5] more and more important in the research of medicine and medical care, proposed eight solutions of blockchain ap- proposed a heart rate monitoring system based on mobile devices and geographical location, which can monitor plication in medical care, and predicted that blockchain and physiological signals and send alarm information when AI solve various medical problems in the future. Jen Hung abnormal heart rate changes. et al. [17] used blockchain in the drug supply chain to create ,e health-care systems [6–9] are data-distribution transparent drug transaction data, prevent counterfeit drugs, domains where many physiological signals are generated, and protect public health. stored, scattered, and accessed daily by using RFID. Yuri ,e abovementioned research findings do not apply blockchain to RFID systems. However, the protocol [18] alvarez ´ et al. [6] described that the contribution of RFID technology can improve medical services, can offer hospital proposed the RFID system based on blockchain and did not apply fog computing to medical fields. It is our innovative tracking of patients, drugs, and medical assets, and can improve the efficiency and safety of electronic medical work to propose RFID protocol based on fog computing and ´ block chain technology in medical systems. applications. Martinez Perez et al. [7] used RFID technology in the ICU (information management system) to track ICU RFID protocol framework based on fog computing and patients’ admission, nursing plan, life monitoring, pre- blockchain is used for medical big data collection and data scription, and drug management process, improving the privacy protection [19–21]. Gu et al. [19] proposed a security quality of patients’ care during hospitalization. Adame et al. and privacy protection solution for fog computing, which [8] proposed the monitoring systems for intelligent designs a framework for security and privacy protection healthcare which provides location status and tracks patients using fog computing and a privacy leakage based on context- based dynamic and static information to improve health and and health-care assets. Omar et al. [9] proposed the reliable, secure, and privacy-based medical automation and orga- medicine infrastructure. Silva et al. [20] proposed a medical records management architecture based on fog computing. nizational information management system that can provide real-time monitoring of vital signs of patients during hos- ,e architecture used blockchain technology to provide pitalization for intelligent patient management. necessary privacy protection and to allow fog nodes to ,e literatures [11–15] have been tremendous concen- execute authorization processes in a distributed manner. tration in blockchain applications. Xu et al. [11] provided a Guan et al. [21] discussed data security and privacy issues in decentralized resource management framework based on fog computing. ,ey pointed out that the data security and blockchain by studying resource management issues. Aiqing privacy challenges posed by fog layers and data protection and Xiaodong [12] proposed a blockchain-based security technologies in cloud computing cannot be directly applied and privacy protection sharing protocol to improve the to fog computing. Patel added the fog computing in the original blockchain medical data sharing sequence model diagnosis of electronic health systems. ,e private block- chain is responsible for storing personal medical informa- [22]. Tang et al. [23] proposed a new game theory framework to improve the mining efficiency of blockchain network and tion (PHI), while alliance blockchain keeps the secure index record of PHI. Dubovitskaya et al. [13] proposed a frame- maximize the total benefits of blockchain network. In order work for sharing EMR data for cancer patients based on the to improve the diagnosis of an electronic medical system, blockchain and implemented. Lebech et al. [14] used mul- Zhang and Lin [12] proposed a security and privacy pro- tisignature blockchain protocol for diabetes data manage- tection based on the blockchain PHI sharing (BSPP) scheme. ment and access control, as well as sharing and encryption. ,e consensus mechanism (private blockchain and joint ,e new approach helps to share diabetes data more ef- blockchain) is constructed by designing a blockchain data structure. fectively in different institutions. Yue et al. [15] proposed the medical data gateway (HGD) architecture based on block- chain, which enabled patients to safely own, control, and 3. Mutual Authentication Protocol Using IMDs share the data without infringing privacy. When different research institutions share the physio- ,e presented mutual authentication protocols for the WISP logical signals, the issues of privacy and security are the have two modes: the regular mode shares the IMD and the primary focus of research institutions because the physio- same credentials; the emergency mode is initiated when one logical signals include the sensitive information, and the of the following status appear. ,e IMD credentials are not Journal of Healthcare Engineering 5 shared by the programmer; the patients cannot communi- K′ � H (Q), and verifies K′ ? � K . If the key is bio bio bio cate with the shared credentials; and the credentials con- successfully confirmed, WISP generates N and com- figured are expired. putes K � H (K | N ) and K′ � H (K | N ). WISP ad- bio W W mits the reader by transmitting M � ((N , ID ) , 4 W W Kbio HMAC (K , N |N |ID )). bio R W W 3.1. "e "reats and Its Influence on the Medical Record. Step5: in order to determine (N , ID ), the reader W W ,e threats and its influence on physiological signals are as decodes the message’s first part using K . After that, it bio follows: privacy, equity, consent, and patient governance in verifies the authenticity of (N , ID ) by employing W W health information collection; discrimination in information HMAC function and comparing the result to the re- applications; and handling data breaches. ceived message’s second section. If they are equal, the Because of newly developing data collection and storage reader calculates K � H (K |N ) and K′ � H (K | N ) bio W W technologies to collect and analyse vast amounts of data, the and then sends M � (Seq , HMAC (K′, N |Seq )). ,e 5 1 W 1 technologies (RFID, blockchain, and artificial intelligence) reader sends messages (K′, Seq ) to the programmer. enable more human experience. While strict clinical testing Step 6: WISP verifies the session keys’ equality. IMD is still required for handling data breaches, the technologies collects the key of session and the relevant sequence will fuel a new age of precision medicine in various methods, number. as shown in Table 1. Two modes (emergency mode and regular mode) have the same shortcomings. First, neither model talks about how 3.2. Physiological Signals Data Privacy Rules. While physi- to store large amounts of data on the database. Second, both ological signals are the lifeblood of today’s digital society, models have secret key leakage attacks and tracking attacks. numerous people are not fully aware of appropriate data ,ird, neither model uses cloud storage technology or collection and processing. ,e privacy issues are the con- blockchain technology. cerns in the process of generating data. It is more significant to be considered privacy protection in healthcare, where personal physiological signals consist of a large percentage of 3.5. Attacks for Mutual Authentication Protocol in the the data. ,e rules and regulations guide the process of data Emergency Mode generation, transmission, access, and exchange. ,e privacy storage rules are as follows: entitles patients more control 3.5.1. "e Reader Impersonation Attacks. ,e reader com- over physiological signals; establishes boundaries of physi- putes K � H (Q) and then sends M � (ID , I, HMAC bio 3 R ological signals’ use and release; protects the privacy of (K , I|Q|ID )) to WISP. bio R physiological signal; enables patients to make choices wisely; In order to simplify the analysis steps, the steps 3–6 in and enables patients to be aware of methods for preventing Figure 4 are omitted here. ,e tracing attacks in the data leakage. It is completely important to maintain the emergency mode have three phases. security and privacy of physiological signals by using RFID, (1) ,e testing phase: the attacker chooses the target tag fog computing, and blockchain. 1 1 1 ∗ ∗ R , monitors the first round ( M , M , M ) to R , 1 2 3 and obtains the outputs keys K � H (Q), and the bio reader applies M � (ID , I, HMAC (K , I|Q|ID )) 3.3. Security Attacks and Requirements for IMDs. ,is part 3 R bio R to WISP. shows IMDs’ main security attacks [10] and discusses the security requirements in Figure 3. Table 2 explains the (2) ,e reader impersonation attacks phase: the attacker symbols and definitions of all the authentication protocols. (the counterfeit reader R′) chooses the monitored information M . ,e attacker monitors the output 2 2 information ( K � H (Q), M � (ID , I, HMAC bio 3 R 3.4. Mutual Authentication Scheme in the Emergency Mode. (K , I|Q|ID ))) in the second round. bio R ,e IMD and programmer can securely produce and offer (3) ,e decision phase: the adversary obtained the values the major key which is extracted from the patient’s data by 1 1 2 2 1 1 2 ( K , M ) and ( K , M ). If ( K , M )≠ ( K , bio 3 bio 3 bio 3 bio executing the presented mutual authentication protocol’s M ), and the attacker confirms that R is not R′ with emergency mode in Figure 4. 1 1 2 2 the probability 1; if ( K , M ) � ( K , M ), the bio 3 bio 3 Step1: the reader initiates the presented mutual au- attacker makes sure that R is the counterfeit R′. thentication protocol’s emergency mode by transmit- ,erefore, the protocol does not meet the weak ting the synchronization request M � (ID , N , and 1 R R indistinguishability property and suffers from the flag) to the IMD. reader impersonation attacks. Step2: WISP computes features V � RandPermute (F ∪ F′ W) and sends V to the reader. 3.5.2. Reducing the Calculation Cost of Reader and WISP. Step3: the reader computes K � H (Q) and sends bio In order to reduce the computation of the whole systems, the M � (ID , I, HMAC (K , I|Q|ID )) to WISP. 3 R bio R HASH computational expense of the reader and WISP are Step4: if the number of matching characters is greater high, the proposed protocol uses the PRNG function to than the predefined threshold, the WISP calculates replace HASH function. 6 Journal of Healthcare Engineering Table 1: Audiences and influence functions of medical record. Audiences Influence functions Patients Promote diagnoses and identification of physiological signals, facilitate preventive care, and reduce costs Doctors ,e rigorous diagnosis, treatment choices, monitoring disease progression, therapy response, and patient susceptibility Researchers Perform large-scale disease modelling and efficacious therapies Risk estimation, forecasting relapse possibility, designing criteria for discharge/readmission, predicting mortality, and Clinics conveying potential crisis episodes Battery Data reliability depletion attack Unsecure access Relay Renewable Energy during attack credentials emergency preservation situations Attacks Perfect forward secrecy Secure access in Attacks Traffic (PFS) property emergency situations targeting capture and authentication analysis protocl De Protection against battery synchronization depletion attacks attacks Figure 3: Security attacks and requirements for secure IMDs. Table 2: Symbols and definitions of the enhanced RFID system privacy protection authentication protocol. Symbols Definitions C ; TID ; COUNT Challenge from the DB to reader; temporary identity; count R ; R ; N Response for the reader; R ⊕ N ; random number generated i s i s th th Res ; CRP (C R ); K h(COUNT + 1‖Ri‖R )); i challenge-response; i key s i i i i PUF ;h (.); ⊕; ‖(|) PUF for the tag T; one-way hash function; XOR; concatenation K ; K ; K Hospital; patient; doctor H P D 3.6. Mutual Authentication in the Regular Mode. ,e regular Step6: WISP can confirm the message’s freshness and mode ensures the secure data exchange, as shown in the keys’ equality computed on both sides. WISP in- Figure 5. crements the Nbr parameter which represents the total number of session keys which originated from the Step1: the reader sends M � (N , ID , flag, HMAC (K, R R primary key. N |ID )) in the regular mode. R R Step7: WISP delivers the messages (K′, Seq , Nbr) to Step2: WISP can confirm the received request’s awaken IMD antenna. freshness and the reader’s authenticity. If the organized primary key has not run out, the received request from ,e attacks for mutual authentication protocol in the the keys is authenticated by the WISP. By contrary, the regular mode. WISP rejects access by sending the denial message. Step3: WISP computes K′ � H (K | N ), and sends 3.6.1. Secret Key Disclosure Attacks. ,e attackers monitor M � ((Nbr, N , ID ) , HMAC (K, N | N |ID )) to W W K R W W the delivery messages and reveal the secret keys as follows: reader. In Step1, M � (N , ID , flag, HMAC (K, N |ID )), the Step4: when receiving the messages, the reader decodes the 1 R R R R attacker discloses ID first part of the messages to obtain (Nbr, N , and ID ). R W W In Step3, M � ((Nbr, NW, IDW) K, HMAC (K, NR| Step5: after verifying successfully, the reader calculates 2 NW|IDW)), the attacker discloses IDW the key value K′ using N and sends the messages M � (Seq , HMAC (K′, N | Seq )). 1 W 1 In Step7, (K′, Seq1, Nbr), the attacker discloses K′ 3 Journal of Healthcare Engineering 7 IMD WISP RFID reader Programmer M : ⟨N , ID , flag⟩ 1 R R Record ECG Record ECG Create vault M : ⟨Vault V⟩ Identify common features Q compute key: K = H (Q) bio M : ⟨ID , I, HAMC(K , |Q| ID )⟩ 3 R bio R Compute key: K = H (Q) bio Select N M : ⟨N , ID  , HAMC (K ,N | N | ID )⟩ 4 W W K bio R W W bio Compute key: Compute key: K = H (K | N ) K = H (K | N ) bio W bio W K′ = H (K | N ) K′ = H (K | N ) bio W bio W M : ⟨Seq , HMAC, (K′, N | Seq )⟩ 5 1 W 1 ⟨K′, Seq ⟩ Verify K′ 1 ⟨K′, Seq ⟩ Secure communication using K′ Figure 4: Mutual authentication protocol in the emergency mode (protocol 1). 3.6.2. "e Tracing Attacks. In order to simplify the analysis 3.6.3. Medical Framework Based on RFID, Blockchain, and process, the steps 3–6 in Figure 5 are omitted here. ,e Artificial Intelligence. At present, amounts of patients have tracing attacks have three phases. the comprehensive datasets which consist of clinical history (the genetic, lifestyle data, drug, and blood biochemistry). In (1) ,e testing phase: the attacker chooses the target tag T . addition, the consumer companies and the pharmaceutical 1 1 1 ,en, she/he monitors the first round ( M , M , M , 1 2 3 are willing to pay much money for the vast personal 1 1 1 M ) to T and obtains the outputs keys ( ID , ID ). 4 R W physiological signal data applied to train their AI model via (2) ,e tracing attacks phase: we assume that the tag set using the machine learning. We proposed the medical 0 4 i ∗ (T , T ,. . . T ) includes T and the counterfeit tag T′. framework based on RFID, blockchain, and artificial in- 2 2 ,e attacker monitors the keys ( ID , ID ) in the R W telligence, as in Figure 6. second round. Previous researches based on RFID, blockchain, and artificial intelligence mainly focused on the medical ap- (3) ,e decision phase: the adversary obtained the values 1 1 2 2 1 ( ID , ID ) and ( ID , ID ). If ( ID , plication, respectively. ,e studies improve the time R W R W R 1 2 2 ID )≠ ( ID , ID ), the attacker confirms that T′ is proficiency of physiological signal data processing and W R W ∗ 1 1 2 not T with the probability 1; if ( ID , ID ) � ( ID , contribute to medical data management by combining R W R i ∗ ID ), the attacker makes sure that T is T (the three technologies. ,e effectiveness of the medical counterfeit tag T′). ,erefore, the original protocol in framework involves low resource usage, large computation the regular mode does not meet the weak indistin- time, more energy, less power, and low memory con- sumption (Algorithm 1). guishability property and suffers from the tracing attacks. 8 Journal of Healthcare Engineering IMD RFID reader Programmer WISP M′ : <N , ID , flag, HAMC (K, N | ID > Key K 1 R R R R Request Check remaining Nbr key lifetime [Invalid key] ⟨Deny, flag⟩ [Valid key] Select random M′ : <{Nbr, N , ID } , HAMC (K, N | N | ID )> 2 W W K R W W Compute session key Compute session key K′ = H (K | N ) K′ = H (K | N ) W W M′ : <Seq , HMAC, (K′, N | Seq )> 3 1 W 1 <K′, Seq > Verify K′ update Nbr <K′, Seq , Nbr> Secure communication using K′ Figure 5: Mutual authentication in the regular mode (protocol 2). Application Machine Feature reverse learning extraction engineering classification TAG Patient Data collection Reader Blockchain Figure 6: ,e medical framework based on RFID, blockchain, and artificial intelligence. Journal of Healthcare Engineering 9 ,e proposed protocol in the emergency mode (Figure 7) is as follows: (1) Step 1 (2) ,e reader initially generates the random numbers (N , ID , flag � 1); R R (3) Calculate A � N ⊕ ID ; R R (4) Broadcast M ; (5) Step 2 (6) Compare ID ; (7) if ID ≠ A⊕ N then R R (8) Process termination; (9) else (10) { (11) set up V�RandPermute (F ∪ F ); W W (12) Send M to reader (13) M in V; (14) Step 3 (15) for each f do (16) if f � F then r R (17) ,e reader and the tag match each other; (18) Calculate K � H (Q || N ); bio R (19) Send the message M � (I HMAC(K , I|Q|ID )); 3 bio R (20) Step 4 (21) If the number of matched characteristics is greater than the predetermined threshold in WISP (22) Calculate K � H (Q||N ); bio (23) if K � Kbio then bio (24) if HMAC(K , I|Q|ID ) � HMAC(K , I|Q|ID ); bio R bio R (25) Verify success, generate random number N , Calculate B � N ⊕ ID ; W W W (26) Calculate K � H (K | N ), and new key K’ � H (K | N ); bio W W (27) Send S1 � HMAC(K , N | N |ID ) M � <􏼈 N , ID 􏼉 , HMAC(K , N |N |ID )> bio R W W 4 W W K bio R W W bio (28) Step 5 (29) if K (reader) � K (tag), obtain ( N , ID ); bio bio W W (30) Calculate S2 � HMAC(K , N | N |ID ); bio R W W (31) if S2 � S1 then (32) Calculate (K, K′) K � H (K | N ), K’ � H (K | N ); bio W W (33) Send < Seq , HMAC(K‘, N | Seq )> to WISP 1 W 1 (34) Step 6 (35) WISP verifies the session keys’ equality calculated by both sides (WISP, reader) (36) If the session keys calculated on both sides are equal (37) WISP records (K′, Seq1) to awaken the IMD antenna (38) When IMD detects the request, begins to collect (K′, Seq1), and employs them to exchange data securely with the programmer (39) }; ALGORITHM 1: ,e suggested mutual authentication protocol in the emergency mode. new nonce N′ R in the second round, the attacker cannot 4. Security and Performance Analysis of counterfeit the original reader. Protocol 3 and Protocol 4 ,e protocol 3 and protocol 4 are more suitable to store 4.1.2. Key Leak Attack Resistance. In order to resist the key physiological signals in medical applications. leak attacks, WISP calculates B � N ⊕ ID ; the reader W W calculates K � PRNG (K ||N ); and K′ � PRNG (K||N ). bio W W 4.1. Security Analysis for Protocol 3. Scheme 3 overcomes the weaknesses of protocol 1, and the protocol 4 overcomes the 4.1.3. Provision of Data Integrity Verification. In order to weaknesses of protocol 2. meet data integrity, the protocol 3 has used HMAC hash calculation to protect the integrity of messages (K1, Seq ). 4.1.1. "e Reader Impersonation Attacks Resistance. In order to resist the reader impersonation attacks, the reader 4.1.4. Provision of Scalability and Efficiency. In order to calculates K � PRNG (Q||N ) using N . Even if the at- satisfy the scalability, each tag identifier does not match the R R bio tacker monitors the output information ( K � PRNG (Q|| corresponding key in DB. ,erefore, the identifications of bio 2 2 N′ ), M �(ID , I, HMAC ( K , I||Q||ID ))) using the tag keys do not match one by one in DB of the improved R 3 R bio R 10 Journal of Healthcare Engineering IMD WISP RFID reader Programmer Step 1.1 M : <N , A, flag > Step 1. Record ECG 1 R A = N + ID R R Step 2. Record ECG create vault Step 2.1 M : <Vault V> Step 3. Identify common features Q compute key: K = PRNG (Q || N ) bio R Step 3.1 M : <I, HMAC (K , I | Q | ID )> 3 bio R Step 4. Compute key K = PRNG (Q || N ) bio R Select B = ID N W W Step 4.1 M : <{N , B}K , HMAC (K , N | N | ID )> 4 W bio bio R W W Step 5. Compute keys Compute keys K = PRNG (K | N ) bio W K = PRNG (K | N ) bio W K′ = PRNG (K | N ) K′ = PRNG (K | N ) K1 = K′ N Step 5.1 M : <Seq , HMAC (K′, N | Seq )> 5 1 W 1 Step 6. Verify K′ K2 = K′ + N Step 6.2 Step 6.1 <K2, Seq > <K1, Seq > Secure communication using K′ Figure 7: ,e proposed mutual authentication protocol in the emergency mode (protocol 3). protocol, which guarantees the efficiency of tag authenti- 4.2. Security Analysis for Protocol 4 cation and satisfies the scalability property. 4.2.1. Secret Key Disclosure Attacks Resistance. In order to achieve anonymous and privacy requirements in improved protocol 4, the protocol uses the XOR function to encrypt 4.1.5. Replay Attacks Resistance. ,e attacker replays the the transmitted keys as follows: messages to authenticate by monitoring the previous in- formation. In order to resist replay attacks, all messages are B � ID ⊕ N , K1 � K′⊕ N , K2 � K′⊕ N . W W R W encrypted by using the random numbers (Nbr, N , and N ) W R and combined with PRNG function. 4.2.2. Tracing Attacks Resistance. ,e key updating mecha- th nism K′ � PRNG (K|N ) involves the i keys and the nonces th th (N , K). ,e i key K cannot be cracked by the (i+1) keys 4.1.6. Provision of Data Integrity Verification. In order to W i th achieve the property of data integrity, we have used K and the i sessions. ,e reasons are that PRNG functions i+1 protect the parameters by the encrypted messages. ,erefore, PRNG calculation K′ � PRNG (K|N ) to protect the in- tegrity of K′. the enhanced protocols resist the tracing attacks. Journal of Healthcare Engineering 11 RFID IMD WISP Programmer reader Key K Step 1.1 Step 1. A = ID N M′ : <{N , A, flag, HMAC (K, N | ID )> R R 1 R R R Request Step 1.2 Check remaining Nbr key lifetime [Invalid key] <Deny, flag> [valid key] Step 2. Select random N compute B = N + ID W W Step 2.1 M′ : <{Nbr, N , B} , HMAC (K, N | N | ID )> 2 W K R W W Step 3. Compute session key K′ = PRNG (K | N ) K1 = K′ + N Step 3.1 M′ : <Seq , HMAC (K′, N | Seq )> 3 1 W 1 K′ = PRNG (K | N ) K2 = K′ N Step 4.2 Step 4.1 <K2, Seq , Nbr> <K1, Seq > Secure communication using K′ Figure 8: ,e proposed mutual authentication protocol in the regular mode (protocol 4) (Algorithm 2). 4.2.3. Availability and Desynchronization Attacks Resistance. 5. Blockchain Framework for Security and In order to provide anonymity, the communication com- Privacy Storage and Sharing ponents (tag and DB) update the shared messages after completing the conversation. If the opponents destroy the A framework is developed to share physiological signals’ updating process, the authentication scheme is subjected to cross domain and build the radiological studies’ ledger and desynchronization attacks. In order to guarantee the con- patient-defined access permissions by applying the block- fidentiality and anonymity of K, the messages synchronously chain as the distributed data store. Relative disadvantages of should be updated. In addition, the attacker knows the the framework include the privacy’s complexity and security shared key K′ during the updating processes, which is models. Ultimately, the large-scale feasibility of the approach protected by the random numbers (N , N ). ,e improved remains to be demonstrated. W R protocol is desynchronization resistance. ,e peculiar health-care technologies are required, such as parallel processing, distributed data network, scalable storage, frameworks, and infrastructures. ,e fog 4.3. "e Comparisons of Security and Performance Analysis. computing is economical and customizable, since fog Table 3 lists the computational cost for five protocols. ,e computing handles these complex problems in the virtual computational costs of tags in protocol 3 are 3PRNG + Xor, environment and only needs to pay for the used services and the computational costs of tags in protocol 4 are and resources. 2PRNG + Xor. ,e safety performances of the enhanced ,e sharing physiological signals systems are important protocols are superior to other schemes. Compared with the in different medical institutions, but the current infra- original protocol 1 and protocol 2, the improved protocols structure for transmitting physiological signals relies on support the security enhancements and ensure the function the trust third-party intermediaries. We propose the such as integrity, efficiency, and user privacy. framework of cross-domain sharing image where the 12 Journal of Healthcare Engineering ,e proposed mutual authentication protocol in the regular mode is in Figure 8 as follows: (1) Step 1 (2) ,e reader generates (N , ID , flag � 0); R R (3) Calculate A � ID ⊕ N and K � H (N |ID ); R R R R (4) Transmit M � < N , ID , flag, HMAC(K, N | ID )> 1 R R R R (5) When WISP receives the request, it confirms that the primary key is expired and verifies that how many session keys which originated from the primary key exceeds the predetermined threshold (6) if t< T then If the primary key has not expired, WISP receives the messages (7) else the key expired, access denied; (8) Step 2 (9) After WISP successful authentication, the random number N is generated; (10) Calculate K’ � H (K | N ), B� ID ⊕ N ; W W W (11) Transmit M � < {Nbr, Nw, I Dw} , HMAC(K, N |N |ID )> 2 k R W w Calculate S1 � HMAC(K, N |N |ID ); R W w (12) Step 3 (13) After receiving the messages, the reader starts to parse the first part of the message through the key K to obtain (Nbr, NW, IDW); (14) Calculate S2 � HMAC(K, N |N ID ); R W w If S2 � S1 then ,e message is true; Calculate K’ � PRNG (K |N ), K1 � K’⊕ NR; (15) Transmit M � Seq , HMAC(K , N |Seq ) 3 1 w 1 (16) Step 4 (17) Based on the received HMAC, WISP can confirm the timeliness of the message and the equality of the session keys calculated on both sides (18) After verifying successfully, Nbr++, K2 � � K’⊕ NW; (19) WISP records (K2, Seq1, Nbr) to awaken the IMD antenna (20) When IMD detects the request, collects (K1, Seq1), and employs them to exchange data securely with the programmer. ALGORITHM 2: ,e proposed mutual authentication protocol in the regular mode. blockchain is used as the distributed data storage to es- 5.1. Physiological Signals Data Sharing Model Based on tablish patient-defined access rights. ,e blockchain Blockchain [22]. Intelligent contract based on blockchain is framework is verified to eliminate the access permission of used to promote the security analysis and management of the third-party to protected physiological signal infor- medical sensors. Intelligent device invokes intelligent con- mation, meets many standards of the interoperable medical tract and writes records of all events on blockchain. ,e system, and easily generalizes to fields beyond physio- intelligent contract systems support real-time patient logical signal. We summarize the framework based on monitoring and medical intervention by sending notifica- blockchain to allow patients to securely grant electronic tions to patients and medical professionals. ,e provider of medical records can modify the physiological signals, but it access permission to their physiological signal data and describe the advantages and disadvantages of the approach. needs patient’s consent, and the patient can assign access ,e actual transmission of physiological signals re- authority to medical records. quires the physiological signals receiver who transmits the When applying blockchain to the construction of the signed request to the URL endpoint. ,e individual ser- credit system, we promote the collection and supervision of vice is the requesting entity that the access permission of credit information in the medical field and build the new the physiological signals study is authorized to by the relationship platform. It is significant to the improvement of owner (patient). ,e studies of all patients’ physiological the credit system construction. According to the unified signals result in the huge blockchain, far too large to evaluation criteria, the credit rating is evaluated, the result of download, store, and validate for nodes running on the rating level is publicized on the platform of block chain, mobile devices. ,e size of the blockchain has been proven the credit rating is rewarded, and the violation of credit is punished, so as to strengthen the construction of the credit to be the limiting element for chains storing the trans- actional data. system in the medical field in the real sense. Considering all of these factors, sharing the physiological ,e asymmetric information encryption methods need signals by using blockchain helps the interoperable health two keys: public key and private key. After the physiological system and has greater ability to access patients’ physio- signals are encrypted with public key, only the corre- logical signals electronically. sponding private key can be used for decryption. On the Journal of Healthcare Engineering 13 IMD WISP Programmer RFID reader M′ : ⟨N , C, flag, HMAC (K, N | ID )⟩ 1 R R R Request Key K C = ID ID Check remaining R R key lifetime Nb r + ID = CN R R [Invalid key] <Deny, flag> [Valid key] Select random M′ : ⟨{Nbr, N , ID } , HMAC (K, N | N | ID )⟩ 2 W W k R W W K′ = PRNG (K | N ) K′ = PRNG (K | N ) W + Nbr K1 = K′ M′ : ⟨Seq , HMAC (K′, N | Seq )⟩ 3 1 W Verify K′ ⟨K1, Seq ⟩ update Nbr K1 = K′ Nbr ⟨K1, Seq , Nbr⟩ Secure communication using K1 Figure 9: Secure communication protocol between the IMD and the programmer (protocol 5) (Algorithm 3). contrary, if the private key is used to encrypt data, only the medical diagnosis in the hospital, which confirms the corresponding public key can be used for decryption. If the statement and promised to provide the study in the blockchain can be grafted, scientific research institutions previous block. ,e patient’s signature declaration is understand the probability of disease occurrence, the oc- obtained through the mobile application, which shares currence of accidents, the level of hospital management, and and stores the values required allowing access to the claims cases and other actual situations. transaction in the future. ,en, the hospital signs the follow-up information of the patients and broadcasts (1) Use the fog-based blockchain and fog warehouse to the transaction to the blockchain. store medical data, as shown in Figure 10. (3) Allow access: the transaction allows the owner of the (1) List of medical research and patients in each medical information research to authorize the other institute. party to retrieve its medical data. Patient K signs a transaction to grant the function to doctor K . ,e (2) Patients are authorized to access the entity set of signed verification blocks are embedded in block- each study. ,e entities are represented by the chains. As shown in Figure 11, patients publish the common part of the asymmetric key pair on the transaction after verifying the key with the doctor blockchain. through the APP platform. ,e patient can be au- (2) Definition study: the transaction builds the patient as thorized to the legitimate doctor or institution, and the the master of a UID which is the specific unique doctor can associate any medical information received identifier and the source as the creator. Tuples stored with the correct local medical record number. in block chains are transactions with double signa- ,e middle column (Block Chain Medical Data Sharing tures, similar to documents with signatures from Sequence) describes the interaction between entities and patients and hospital representatives. ,e patients judgments in each stage and reflects the sharing medical claim that the definition study has received the 14 Journal of Healthcare Engineering ,e programmer can use the session key calculated by the protocol to establish the secure communication after IMD authenticates the programmer in Figure 9. (1) Step 1: (2) ,e reader initially generates (K, C � ID ⊕ N ) and transmits the values M � (N , C, flag, HMAC (K, N | ID )) to the WISP. R R R R R (3) Step 2: (4) ,e IMD returns Nbr and updates ID � C⊕ N . R R (5) Step 3: (6) If the key is valid then (7) ,e WISP selects NW and transmits the values M �((Nbr, N , ID ), HMAC (K, N | N |ID )) to the reader. W W R W W (8) else (9) ,e WISP transmits the sequences (Deny, flag) to the reader. (10) Step 4: (11) ,e WISP updates K′ � PRNG (K | NW), and the reader updates K′ � PRNG (K | N ) and K1 � K’⊕ Nbr. ,e reader sends the value M3 � (Seq1, HMAC (K′, N , Seq1)) to the WISP and sends the messages (K1, Seq1) to the programmer. (12) Step 5: (13) ,e WISP identifies K′ by comparing the value K′ of the WISP with the K′ value of the reader. ,e WISP updates Nbr, K � K’⊕ Nbr, and sends (K , Seq1, Nbr) to the IMD. 1 1 ALGORITHM 3: Secure communication protocol between the IMD and the programmer. Table 3: ,e comparisons of the performance analysis and safety performance. Performance Protocol 1 Protocol 3 Protocol 2 Protocol 4 Protocol 5 F0 No Yes No Yes No F1 3H + Xor 3PRNG + Xor 2H + Xor 2PRNG + Xor 1PRNG+2Xor F2 No Yes No Yes Yes F3 No Yes No Yes Yes F4 No Yes No Yes Yes Attack types Protocol 1 Protocol 3 Protocol 2 Protocol 4 Protocol 5 R1 No Yes No Yes No R2 Yes Yes Yes Yes Yes R3 Yes Yes Yes Yes Yes R4 Yes Yes Yes Yes Yes R5 No Yes No Yes Yes R6 No Yes Yes Yes Yes F0: provision of scalability and efficiency; F1: storage cost (tag); F2: blockchain-enabled; F3: cloud computing-enabled; F4: fog computing-enabled. R1: key leak attacks resistance; R2: replay attacks resistance; R3: desynchronization attacks resistance; R4: reader impersonation attacks resistance; R5: tracking attacks resistance; R6: tag impersonation attacks resistance. information by supporting distributed block chains and out- which is used to construct the blockchain for medical data of-block transactions. sharing. ,e actual medical data transmission requires the We have showed the technology fundamentals of medical data receiver to deliver the signature request to the blockchain and provided a summarization of the blockchain medical source’s URL endpoint which creates the research. application that can be used as a tool to allow the patient- Both requests and responses are transmitted through the controlled, physiological signal’s cross-domain sharing secure link of the transport layer to prevent eavesdropping. without the central authority. In particular, we highlighted ,e effective blocks are generated in the timely manner by the way blockchain satisfies many requirements of the in- generating the distributed database with access permissions teroperable health system. However, these technologies also and stimulating the block generator in some way. Only those have several important limitations, and the relative merits of nodes with security deposits can participate in the expansion existing alternatives must be considered before any large- of the chain, and any node with misconduct will be forced to scale and blockchain-based application for sharing physi- ological signals. abandon its investment. ,e nature of blockchain provides the direct audit of the activity of each node such as the When receiving query request, the physiological signal number of blocks generated and the failure status of the data source verifies the correctness of the signature, ensures blocks generated. ,e node operator can prove the node that the hashed data matches the previously published data ownership by using the private key which is corresponded to for K -owner via Block B, and confirms that the K -owner P P the identity public key of the node to sign the message. ,e has allowed the requestor access to these physiological signal enhanced model adds the fog computing in the original data via Block C. If meeting all the conditions, the response blockchain medical data sharing the sequence model [22], containing the physiological signal study is returned from Journal of Healthcare Engineering 15 Blockchain and fog warehouse Blockchain Public and Hash history technology private keys Structured data types of physiological signal The serial ID of the patient in hospital number Fraction MRI, ECG, Heart, Rate, Signal, Image Websites My webpage com User authentication layer Figure 10: Blockchain based on fog warehouse. Block/time A CD kD kD kp kH kH kp kH Fog query computing Actors Physiological signals Physiological signals Step4: Physician: I am K-owner Step7: Step2: and need to review imaging from On-blockchain Physician uses the information in Patient: e physiological transaction your hospitalization. blocks (A,C) and to submit an query signals were acquired for Kp- Step5: request for physiological owner by K-owner Off-blockchain Assertions Patient: Agreed, I am K-owner signals, signed by K. Step3: commuication and will allow access Step8: Hospital: K-owner’s Hospital valid at the Physician’s Step1: assertion above is accrate and Step6: signature,uses the information in Hospital: kH-owner will K-owner will share these data Patient: As the K-owner blocks (B,C) and to confirm service physiological signals at its establised end point I permit the K-owner to access the authorization, and transmits the retrieval request at https. physiological signals that were physiological signals study in an query acquired for me by K-owner. response. Figure 11: Blockchain medical data sharing sequence diagram based on fog computing. the source. In order to prevent eavesdropping, the requests Step 1: for hospital (K -owner), K -owner will service H H and responses are sent to prevent eavesdropping. ,e spe- physiological signals retrieval requests at https by using cific steps of blockchain medical data sharing sequence on-blockchain transaction and off-blockchain diagram are as follows: communication. 16 Journal of Healthcare Engineering Step 2: for the patient (K -owner), the physiological transform healthcare management. In the future, we will signals are acquired for KP-owner and KH-owner. focus on heterogeneous physiological signal data issues through fog computing, blockchain, and AI technology in Step 3: for hospital, K -owner’s assertion is accrate and the realistic medical environment. KH-owner shares the physiological signals at the established endpoint. Data Availability Step 4: for physician (K -owner), K -owner reviews D D the physiological signals from the hospitalization. ,e paper gives an outline about the framework, and in- Step 5: for patient, if the patients agree, they are K - ternal working and protocols for handling heterogeneous owner and will allow access. physiological signal data. Once the hybrid technologies are Step 6: for patient, the patient permit K -owner to integrated, big data systems and AI technology have the access the physiological signals that were acquired by potential to offer privacy protection and data sharing, KH-owner. transform healthcare management. Step 7: physician uses the information in blocks (A, C) to submit the query request for physiological signals, Conflicts of Interest signed by K . ,e authors declare no conflicts of interest. Step 8: hospital valid at the physician’s signature, uses the data in blocks (B, C) to confirm authorization and Acknowledgments transmits the physiological signals study in the query response. ,e requests are sent by the K -owner at ,is work was supported in part by Jiangsu Postdoctoral timepoint D. Science Foundation (Grant nos. 1701061B and 2017107007); ,e ecosystem is consisted of the blockchain nodes and Xuzhou Medical University Affiliated Hospital Postdoctoral fog storage. For example, one of the main reasons for in- Science Foundation (Grant nos. 2016107011, 183822, corporating fog storage technology into the ecosystem is to 53120225, and 53120226); Xuzhou Medical University Ex- supply the offline storage solution, especially for large cellent Persons Scientific Research Foundation (Grant nos. physiological signals. For security and privacy, the client side D2016006, D2016007, and 53591506); the Practice Inovation would encrypt the physiological signals uploaded to the fog Trainng Program Projects for the Jiangsu College Students storage. With the maturity of the fog storage, personal (Grant nos. 20161031308H and 201610313043Y); the Natural storage may be replaced by it. Science Foundation of the Jiangsu Higher Education In- Most significantly, blockchain technology can create the stitutions of China (Grant no. 16KJB180028); and 333 physiological signal-driven marketplace, where patients can Project of Jiangsu Province (no. BRA2017278). get real return by offering their data to research institutions, pharmaceutical and consumer companies, the application References development community, and producing new physiological signal data. [1] J. P. Rajan and S. E. Rajan, “An internet of things based physiological signal monitoring and receiving system for virtual enhanced health care network,” Technology and Health 6. Conclusions Care, vol. 26, no. 2, pp. 379–385, 2018. [2] O. Faust, Y. Hagiwara, T. J. Hong, O. S. Lih, and We extend the architecture of the IMD with blockchain, U. R. Acharya, “Deep learning for healthcare applications RFID, and WISP, which increases the physiological signal based on physiological signals: a review,” Computer Methods data’s confidentiality and authenticity. ,e enhanced RFID and Programs in Biomedicine, vol. 161, no. 1, pp. 1–13, 2018. protocols provide protection against tracking attacks, [3] R. Shanthapriya and V. Vaithianathan, “ECG-based secure readers’ impersonation attacks, and secret disclose attacks. healthcare monitoring system in body area networks,” in ,e physiological signal records have proved the im- Proceedings of the 2018 Fourth International Conference on portance for the patients, and sharing and acquiring Biosignals, Images and Instrumentation (ICBSII), pp. 206–212, physiological signals is essential for intelligent and advanced IEEE, Montreal, Canada, October 2018. medical services. ,e blockchain application of e-commerce [4] C. Orphanidou, “A review of big data applications of phys- iological signal data,” Biophysical Reviews, vol. 11, no. 1, has proven that trusted and auditable transaction in peer-to- pp. 83–87, 2019. peer networking is possible. In the paper, we have intro- [5] E. O. Tartan and C. Ciflikli, “An android application for duced a blockchain-based architecture model for physio- geolocation based health monitoring, consultancy and alarm logical signal data on fog computing environment. Our system,” in Proceedings of the IEEE 2018 IEEE 42nd Annual contributions are mainly consisted of the proposed solution Computer Software and Applications Conference (COMPSAC), and introduction to future medical data directions in pp. 341–344, IEEE Computer Society, Tokyo, Japan, July 2018. blockchain. ,e paper proposes the outline to show the [6] L. Yuri alvarez, ´ F. Jacqueline, N. Guillermo alvarez ´ et al., framework and schemas for dealing with heterogeneous “RFID technology for management and tracking: e-health physiological signals. Once the hybrid technologies are in- applications,” Sensors, vol. 18, no. 8, pp. 2663–2678, 2018. tegrated, big data systems and AI technology have the po- [7] M. Mart´ınez Perez, ´ C. Dafonte, and A. Gomez, ´ “Traceability tential to offer privacy protection and data sharing and in patient healthcare through the integration of RFID Journal of Healthcare Engineering 17 technology in an ICU in a hospital,” Sensors, vol. 18, no. 5, [23] C. Tang, C. Li, X. Yu, Z. Zheng, and Z. Chen, “Cooperative pp. 1627–1641, 2018. mining in blockchain networks with zero-determinant [8] T. Adame, A. Bel, A. Carreras, J. Melia-Segu ` ´ı, M. Oliver, and strategies,” IEEE Transactions on Cybernetics, vol. 2, no. 3, pp. 1–6, 2019. R. Pous, “CUIDATS: An RFID–WSN hybrid monitoring system for smart health care environments,” Future Gener- ation Computer Systems, vol. 78, no. 2, pp. 602–615, 2016. [9] H. Q. Omar, A. Khoshnaw, and W. Monnet, “Smart patient management, monitoring and tracking system using radio- frequency identification (RFID) technology,” in Proceedings of the Biomedical Engineering and Sciences, vol. 1, no. 2, pp. 1–12, IEEE, Kuala Lumpur, Malaysia, December 2016. [10] N. Ellouze, S. Rekhis, N. Boudriga, and M. Allouche, “Powerless security for cardiac implantable medical devices: use of wireless identification and sensing platform,” Journal of Network and Computer Applications, vol. 107, no. 1, pp. 1–21, [11] C. Xu, K. Wang, and M. Guo, “Intelligent resource man- agement in blockchain-based cloud datacenters,” IEEE Cloud Computing, vol. 4, no. 6, pp. 50–59, 2018. [12] Z. Aiqing and L. Xiaodong, “Towards secure and privacy- preserving data sharing in e-health systems via consortium blockchain,” Journal of Medical Systems, vol. 42, no. 8, pp. 140–154, 2018. [13] A. Dubovitskaya, Z. Xu, S. Ryu et al., “Secure and trustable electronic medical records sharing using blockchain,” in Proceedings of the AMIA. Annual Symposium proceedings/ AMIA symposium, vol. 1, no. 1, pp. 1–13, AMIA Symposium, Washington, DC, USA, January 2017. [14] C. S. Lebech, S. M. Nibe, K. ,omas, P. Vestergaard, and O. Hejlesen, “How to use blockchain for diabetes health care data and access management: an operational concept,” Journal of Diabetes Science and Technology, vol. 13, no. 2, pp. 1–14, 2018. [15] X. Yue, H. Wang, D. Jin, M. Li, and W. Jiang, “Healthcare data gateways: found healthcare intelligence on blockchain with novel privacy risk control,” Journal of Medical Systems, vol. 40, no. 10, pp. 218–242, 2016. [16] B. M. N. Kamel, J. T. Wilson, and K. A. Clauson, “Geospatial blockchain: promises, challenges, and scenarios in health and healthcare,” International Journal of Health Geographics, vol. 17, no. 1, pp. 1–19, 2018. [17] T. Jen-Hung, L. Yen-Chih, C. Bin, and L. Shih-wei, “Gov- ernance on the drug supply chain via gcoin blockchain,” International Journal of Environmental Research and Public Health, vol. 15, no. 6, pp. 1055–1067, 2018. [18] K. Harleen, A. M. Afshar, J. Roshan, A. K. Mourya, and C. Victor, “A proposed solution and future direction for blockchain-based heterogeneous medicare data in cloud en- vironment,” Journal of Medical Systems, vol. 42, no. 8, pp. 156–167, 2018. [19] J. J. Gu, R. C. Huang, L. Jiang, G. Qiao, X. Du, and M. Guizani, “A fog computing solution for context-based privacy leakage detection for android healthcare devices,” Sensors, vol. 19, no. 5, pp. 1–16, 2019. [20] C. A. Silva, G. S. Aquino, S. R. M. Melo, and D. J. B. Eg´ıdio, “A fog computing-based architecture for medical records man- agement,” Wireless Communications and Mobile Computing, vol. 2019, no. 1, pp. 1–16, 2019. [21] Y. Guan, J. Shao, G. Wei, and X. Mande, “Data security and privacy in fog computing,” IEEE Network, vol. 32, no. 5, pp. 1–6, 2018. [22] V. Patel, “A framework for secure and decentralized sharing of medical imaging data via blockchain consensus,” Health Informatics Journal, vol. 1, no. 1, pp. 1–14, 2018. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Healthcare Engineering Hindawi Publishing Corporation

Merging RFID and Blockchain Technologies to Accelerate Big Data Medical Research Based on Physiological Signals

Loading next page...
 
/lp/hindawi-publishing-corporation/merging-rfid-and-blockchain-technologies-to-accelerate-big-data-h2WjvSuMeL
Publisher
Hindawi Publishing Corporation
Copyright
Copyright © 2020 Xiuqing Chen et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
ISSN
2040-2295
eISSN
2040-2309
DOI
10.1155/2020/2452683
Publisher site
See Article on Publisher Site

Abstract

Hindawi Journal of Healthcare Engineering Volume 2020, Article ID 2452683, 17 pages https://doi.org/10.1155/2020/2452683 Research Article Merging RFID and Blockchain Technologies to Accelerate Big Data Medical Research Based on Physiological Signals Xiuqing Chen , Hong Zhu, Deqin Geng, Wei Liu, Rui Yang, and Shoudao Li School of Medicine Information, Xuzhou Medical University, Xu Zhou 221000, China Correspondence should be addressed to Xiuqing Chen; xiuqingchen@126.com Received 4 October 2019; Revised 20 December 2019; Accepted 16 January 2020; Published 14 April 2020 Guest Editor: Liang Zou Copyright © 2020 Xiuqing Chen et al. ,is is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. ,e proliferation of physiological signals acquisition and monitoring system, has led to an explosion in physiological signals data. Additionally, RFID systems, blockchain technologies, and the fog computing mechanisms have significantly increased the availability of physiological signal information through big data research. ,e driver for the development of hybrid systems is the continuing effort in making health-care services more efficient and sustainable. Implantable medical devices (IMD) are ther- apeutic devices that are surgically implanted into patients’ body to continuously monitor their physiological parameters. Patients treat cardiac arrhythmia due to IMD therapeutic and life-saving benefits. We focus on hybrid systems developed for patient physiological signals for collection, storage protection, and monitoring in critical care and clinical practice. In order to provide medical data privacy protection and medical decision support, the hybrid systems are presented, and RFID, blockchain, and big data technologies are used to analyse physiological signals. computing, and blockchain in the medical applications 1. Introduction provide security and privacy protection for storing and ,e medical applications are continually increasing. For sharing physiological signal records. It can provide doctors handling physiological signals efficiently, specific tech- with collaboration ways through IMD [10] and RFID to help nologies, such as data gathering using RFID protocols, patients in case of emergencies mode. ,e new model based infrastructures, and distributed information storage based on blockchain can support medical background rural on blockchain frameworks, are required. ,e hospitals healthcare and analyse data for medicines and medical re- applications are adopting physiological signals to realize a search [11–15]. quicker way to visit these records. ,e physiological signals It is urgent for different research institutions to share the are responsible to offer patient care, enhance the clinical encrypted physiological signals. ,erefore, privacy and se- performances, and promote the clinical data research curity problems of physiological signals are the data owners [1–5]. and research institutions’ primary focus, when the physio- logical signals include a lot of sensitive information and the Since the fog computing solves the secure storage issues of big data in the clinical data research with minimal cost, the attackers are continually trying novel approaches to steal the fog computing technology is customizable and economical physiological signals. In order to handle these problems, the and offers infrastructure, platform, and software. Physio- medical databases adapted blockchain, and fog computing logical signals’ analysis and migration have been proposed are proposed [16, 17]. ,e medical application ecosystems for accessing and sharing physiological signal data by dif- allow the regulators to share and exchange physiological ferent research labs and health-care experts, which can signal data in Figure 1. ,e introduction of the blockchain- enable exchange of physiological signals more rapid and fog-RFID based on data ecosystems ensures that the indi- suitable by using RFID technologies and smart phone app viduals take control over physiological signal information. platforms. ,e advantages of RFID protocols [6–9], the fog ,e proposed sharing data-driven economy shares the 2 Journal of Healthcare Engineering FDA Private National health companies organization Contract Physiological research AI signal date set organisation Universities Blockchain lifedate Machine learning Insirance Hospitals companies Doctors Patients Blockchain Date Figure 1: ,e flow of data from the individuals to the companies and research institutions. physiological signals for research and commercial purposes patient management in data collection based on RFID; data in Figure 1. storage based on fog computing; and dealing with data In the paper, we protect cardiac IMD against security breaches by using blockchian. In the future, we will discuss threats by presenting a security scheme. First, we verify and the method’s applications in physiological signals research: classify the IMD’s major security attacks. Second, we in- basic research; disease management; aetiology; detection troduce blockchain and the RFID systems to extend the IMD and diagnosis; health services research; treatment devel- architecture [10] and discuss the structures of the interop- opment; and treatment evaluation. ,e possibilities of the erability in the medical environment, as shown in Figure 2. blockchian-fog-RFID method for accelerating big data ,e motivation of the blockchian-fog-RFID method for medical research in physiological signals are enormous. accelerating big data medical research based on physio- ,e paper contribution consist of four parts as follows: logical signal is as follows: the method is becoming more (1) ,e security scheme is a low energy cost RFID common due to the application of powerful computers and system in IMD. ,e applied authentication protocol the availability of physiological signals from various is implemented on the RFID circuit without energy. sources. However, although the complexity of physiolog- (2) ,e applied energy harvesting scheme uses the en- ical signals makes the complex methods particularly ap- hanced WISP, which performs computational plicable, their application of physiological signals is functions and uses the harvested energy to go beyond generally considered earlier than in other fields. Big data passive RFID tags. has become a buzzword in medical innovation. Rapid advances in artificial intelligence particularly promise to (3) ,e presented authentication protocol enables the reform medical practice from the resource allocation to the authorized health-care professionals to obtain the complex diseases’ diagnosis. However, big data brings huge access permission to cardiac IMD securely in the risks and challenges, including major questions about regular and emergency model which are determined patient privacy: the importance of fairness, consent, and according to the patient’s ability to supply valid Journal of Healthcare Engineering 3 Doctor 1 creates an order, which receives an unique ID e android medical called as hash which points to TAG device records the a record in the blockchain patient’s status Reader IMD Blockchain Lab assistant queries the blockchain to access Doctor 2 may replace the the order, does the work doctor 1 during his absence and report to the record and need to observe the Reader Reader patient’s record. Figure 2: Blockchain in the medical environment. credentials, thanks to a biometric key distribution suitable for collecting, storing, and handling heterogeneous scheme implemented. physiological signal. ,e proposed model can be used for physiological signals management. (4) ,e schemes generate and share a master key se- curely based on the physiological sets of the patient 2. Related Work collected by IMD. Monitoring and ensuring data integrity during clinical trials is not always feasible in ,e industry of healthcare has changed dramatically because current research systems. Blockchain makes the data of the boom in clinical research for physiological signal data collected immutable, traceable, and probably more sharing. We summarize the healthcare studies including trustworthy during clinical trials. We also improve physiological signal data, patient information obtained by the way we currently report adverse events. fog computing, and improvements to blockchain technol- In conclusion, we argue that the blockchain can improve ogy. ,e health-care applications of physiological signal data the management of clinical trial data, enhance trust in the adopt big data and deep learning technologies and provide clinical research process, and simplify regulatory oversight with data confidentiality and identity authentication, so as to of trials. Finally, we evaluate the security solution’s security maintain patients’ privacy. In order to more conveniently and performance. serve big data medical analysis, Rajan and Rajan [1] and ,e proposed model covers the many aspects of the Faust et al. [2] proposed the importance of medical big data health industry such as doctors, patients, and pharmacies to privacy and the impact of data analysis on medical care. insurance suppliers and government. ,e paper shows the Rajan and Rajan [1] proposed a physiological signal applications of using RFID, blockchain technologies, and fog monitoring scheme by using the Internet of ,ings (IoT). computing for storing and managing the physiological signal Our schemes use IoT to improve the access method of data. A blockchain model for sharing physiological signals is physiological signals and the real-time dynamic monitoring proposed. In the next section, the combination of block- method of the remote monitoring system, which enhances chain, RFID, and artificial intelligence (AI) technologies is the efficiency of the remote monitoring systems. Faust et al. 4 Journal of Healthcare Engineering attackers are continually trying novel approaches to steal [2] summarized the application of deep learning algorithms in physiological signals and pointed out that deep learning information. In order to meet the privacy needs and deal with the security problems, medical databases which use methods performed better than classical analysis and ma- chine classification methods for large and diverse datasets. blockchain and fog computing technology are proposed. Shanthapriya and Vaithianathan [3] proposed the health ,e enhanced trusted sharing physiological signals monitoring system for human regional network. ,e steg- model features highly secured data encryption and de- anography technologies monitor patients’ health safety and cryption schemes. ,e model requires permission from the provide patients with data confidentiality and identity au- blockchain network to share patient information among thentication. Orphanidou [4] reviewed big data applications medical staff. ,e proposed model encrypts and analyzes the physiological signals through the blockchain network, big of physiological signals, pointed out how the applications use physiological signals to provide real-time support for data analysis technology, and AI technologies. Kamel et al. [16] pointed out that blockchain technology is becoming medical decision making in both clinical and family settings, and need to be overcome in clinical practice. Tartan et al. [5] more and more important in the research of medicine and medical care, proposed eight solutions of blockchain ap- proposed a heart rate monitoring system based on mobile devices and geographical location, which can monitor plication in medical care, and predicted that blockchain and physiological signals and send alarm information when AI solve various medical problems in the future. Jen Hung abnormal heart rate changes. et al. [17] used blockchain in the drug supply chain to create ,e health-care systems [6–9] are data-distribution transparent drug transaction data, prevent counterfeit drugs, domains where many physiological signals are generated, and protect public health. stored, scattered, and accessed daily by using RFID. Yuri ,e abovementioned research findings do not apply blockchain to RFID systems. However, the protocol [18] alvarez ´ et al. [6] described that the contribution of RFID technology can improve medical services, can offer hospital proposed the RFID system based on blockchain and did not apply fog computing to medical fields. It is our innovative tracking of patients, drugs, and medical assets, and can improve the efficiency and safety of electronic medical work to propose RFID protocol based on fog computing and ´ block chain technology in medical systems. applications. Martinez Perez et al. [7] used RFID technology in the ICU (information management system) to track ICU RFID protocol framework based on fog computing and patients’ admission, nursing plan, life monitoring, pre- blockchain is used for medical big data collection and data scription, and drug management process, improving the privacy protection [19–21]. Gu et al. [19] proposed a security quality of patients’ care during hospitalization. Adame et al. and privacy protection solution for fog computing, which [8] proposed the monitoring systems for intelligent designs a framework for security and privacy protection healthcare which provides location status and tracks patients using fog computing and a privacy leakage based on context- based dynamic and static information to improve health and and health-care assets. Omar et al. [9] proposed the reliable, secure, and privacy-based medical automation and orga- medicine infrastructure. Silva et al. [20] proposed a medical records management architecture based on fog computing. nizational information management system that can provide real-time monitoring of vital signs of patients during hos- ,e architecture used blockchain technology to provide pitalization for intelligent patient management. necessary privacy protection and to allow fog nodes to ,e literatures [11–15] have been tremendous concen- execute authorization processes in a distributed manner. tration in blockchain applications. Xu et al. [11] provided a Guan et al. [21] discussed data security and privacy issues in decentralized resource management framework based on fog computing. ,ey pointed out that the data security and blockchain by studying resource management issues. Aiqing privacy challenges posed by fog layers and data protection and Xiaodong [12] proposed a blockchain-based security technologies in cloud computing cannot be directly applied and privacy protection sharing protocol to improve the to fog computing. Patel added the fog computing in the original blockchain medical data sharing sequence model diagnosis of electronic health systems. ,e private block- chain is responsible for storing personal medical informa- [22]. Tang et al. [23] proposed a new game theory framework to improve the mining efficiency of blockchain network and tion (PHI), while alliance blockchain keeps the secure index record of PHI. Dubovitskaya et al. [13] proposed a frame- maximize the total benefits of blockchain network. In order work for sharing EMR data for cancer patients based on the to improve the diagnosis of an electronic medical system, blockchain and implemented. Lebech et al. [14] used mul- Zhang and Lin [12] proposed a security and privacy pro- tisignature blockchain protocol for diabetes data manage- tection based on the blockchain PHI sharing (BSPP) scheme. ment and access control, as well as sharing and encryption. ,e consensus mechanism (private blockchain and joint ,e new approach helps to share diabetes data more ef- blockchain) is constructed by designing a blockchain data structure. fectively in different institutions. Yue et al. [15] proposed the medical data gateway (HGD) architecture based on block- chain, which enabled patients to safely own, control, and 3. Mutual Authentication Protocol Using IMDs share the data without infringing privacy. When different research institutions share the physio- ,e presented mutual authentication protocols for the WISP logical signals, the issues of privacy and security are the have two modes: the regular mode shares the IMD and the primary focus of research institutions because the physio- same credentials; the emergency mode is initiated when one logical signals include the sensitive information, and the of the following status appear. ,e IMD credentials are not Journal of Healthcare Engineering 5 shared by the programmer; the patients cannot communi- K′ � H (Q), and verifies K′ ? � K . If the key is bio bio bio cate with the shared credentials; and the credentials con- successfully confirmed, WISP generates N and com- figured are expired. putes K � H (K | N ) and K′ � H (K | N ). WISP ad- bio W W mits the reader by transmitting M � ((N , ID ) , 4 W W Kbio HMAC (K , N |N |ID )). bio R W W 3.1. "e "reats and Its Influence on the Medical Record. Step5: in order to determine (N , ID ), the reader W W ,e threats and its influence on physiological signals are as decodes the message’s first part using K . After that, it bio follows: privacy, equity, consent, and patient governance in verifies the authenticity of (N , ID ) by employing W W health information collection; discrimination in information HMAC function and comparing the result to the re- applications; and handling data breaches. ceived message’s second section. If they are equal, the Because of newly developing data collection and storage reader calculates K � H (K |N ) and K′ � H (K | N ) bio W W technologies to collect and analyse vast amounts of data, the and then sends M � (Seq , HMAC (K′, N |Seq )). ,e 5 1 W 1 technologies (RFID, blockchain, and artificial intelligence) reader sends messages (K′, Seq ) to the programmer. enable more human experience. While strict clinical testing Step 6: WISP verifies the session keys’ equality. IMD is still required for handling data breaches, the technologies collects the key of session and the relevant sequence will fuel a new age of precision medicine in various methods, number. as shown in Table 1. Two modes (emergency mode and regular mode) have the same shortcomings. First, neither model talks about how 3.2. Physiological Signals Data Privacy Rules. While physi- to store large amounts of data on the database. Second, both ological signals are the lifeblood of today’s digital society, models have secret key leakage attacks and tracking attacks. numerous people are not fully aware of appropriate data ,ird, neither model uses cloud storage technology or collection and processing. ,e privacy issues are the con- blockchain technology. cerns in the process of generating data. It is more significant to be considered privacy protection in healthcare, where personal physiological signals consist of a large percentage of 3.5. Attacks for Mutual Authentication Protocol in the the data. ,e rules and regulations guide the process of data Emergency Mode generation, transmission, access, and exchange. ,e privacy storage rules are as follows: entitles patients more control 3.5.1. "e Reader Impersonation Attacks. ,e reader com- over physiological signals; establishes boundaries of physi- putes K � H (Q) and then sends M � (ID , I, HMAC bio 3 R ological signals’ use and release; protects the privacy of (K , I|Q|ID )) to WISP. bio R physiological signal; enables patients to make choices wisely; In order to simplify the analysis steps, the steps 3–6 in and enables patients to be aware of methods for preventing Figure 4 are omitted here. ,e tracing attacks in the data leakage. It is completely important to maintain the emergency mode have three phases. security and privacy of physiological signals by using RFID, (1) ,e testing phase: the attacker chooses the target tag fog computing, and blockchain. 1 1 1 ∗ ∗ R , monitors the first round ( M , M , M ) to R , 1 2 3 and obtains the outputs keys K � H (Q), and the bio reader applies M � (ID , I, HMAC (K , I|Q|ID )) 3.3. Security Attacks and Requirements for IMDs. ,is part 3 R bio R to WISP. shows IMDs’ main security attacks [10] and discusses the security requirements in Figure 3. Table 2 explains the (2) ,e reader impersonation attacks phase: the attacker symbols and definitions of all the authentication protocols. (the counterfeit reader R′) chooses the monitored information M . ,e attacker monitors the output 2 2 information ( K � H (Q), M � (ID , I, HMAC bio 3 R 3.4. Mutual Authentication Scheme in the Emergency Mode. (K , I|Q|ID ))) in the second round. bio R ,e IMD and programmer can securely produce and offer (3) ,e decision phase: the adversary obtained the values the major key which is extracted from the patient’s data by 1 1 2 2 1 1 2 ( K , M ) and ( K , M ). If ( K , M )≠ ( K , bio 3 bio 3 bio 3 bio executing the presented mutual authentication protocol’s M ), and the attacker confirms that R is not R′ with emergency mode in Figure 4. 1 1 2 2 the probability 1; if ( K , M ) � ( K , M ), the bio 3 bio 3 Step1: the reader initiates the presented mutual au- attacker makes sure that R is the counterfeit R′. thentication protocol’s emergency mode by transmit- ,erefore, the protocol does not meet the weak ting the synchronization request M � (ID , N , and 1 R R indistinguishability property and suffers from the flag) to the IMD. reader impersonation attacks. Step2: WISP computes features V � RandPermute (F ∪ F′ W) and sends V to the reader. 3.5.2. Reducing the Calculation Cost of Reader and WISP. Step3: the reader computes K � H (Q) and sends bio In order to reduce the computation of the whole systems, the M � (ID , I, HMAC (K , I|Q|ID )) to WISP. 3 R bio R HASH computational expense of the reader and WISP are Step4: if the number of matching characters is greater high, the proposed protocol uses the PRNG function to than the predefined threshold, the WISP calculates replace HASH function. 6 Journal of Healthcare Engineering Table 1: Audiences and influence functions of medical record. Audiences Influence functions Patients Promote diagnoses and identification of physiological signals, facilitate preventive care, and reduce costs Doctors ,e rigorous diagnosis, treatment choices, monitoring disease progression, therapy response, and patient susceptibility Researchers Perform large-scale disease modelling and efficacious therapies Risk estimation, forecasting relapse possibility, designing criteria for discharge/readmission, predicting mortality, and Clinics conveying potential crisis episodes Battery Data reliability depletion attack Unsecure access Relay Renewable Energy during attack credentials emergency preservation situations Attacks Perfect forward secrecy Secure access in Attacks Traffic (PFS) property emergency situations targeting capture and authentication analysis protocl De Protection against battery synchronization depletion attacks attacks Figure 3: Security attacks and requirements for secure IMDs. Table 2: Symbols and definitions of the enhanced RFID system privacy protection authentication protocol. Symbols Definitions C ; TID ; COUNT Challenge from the DB to reader; temporary identity; count R ; R ; N Response for the reader; R ⊕ N ; random number generated i s i s th th Res ; CRP (C R ); K h(COUNT + 1‖Ri‖R )); i challenge-response; i key s i i i i PUF ;h (.); ⊕; ‖(|) PUF for the tag T; one-way hash function; XOR; concatenation K ; K ; K Hospital; patient; doctor H P D 3.6. Mutual Authentication in the Regular Mode. ,e regular Step6: WISP can confirm the message’s freshness and mode ensures the secure data exchange, as shown in the keys’ equality computed on both sides. WISP in- Figure 5. crements the Nbr parameter which represents the total number of session keys which originated from the Step1: the reader sends M � (N , ID , flag, HMAC (K, R R primary key. N |ID )) in the regular mode. R R Step7: WISP delivers the messages (K′, Seq , Nbr) to Step2: WISP can confirm the received request’s awaken IMD antenna. freshness and the reader’s authenticity. If the organized primary key has not run out, the received request from ,e attacks for mutual authentication protocol in the the keys is authenticated by the WISP. By contrary, the regular mode. WISP rejects access by sending the denial message. Step3: WISP computes K′ � H (K | N ), and sends 3.6.1. Secret Key Disclosure Attacks. ,e attackers monitor M � ((Nbr, N , ID ) , HMAC (K, N | N |ID )) to W W K R W W the delivery messages and reveal the secret keys as follows: reader. In Step1, M � (N , ID , flag, HMAC (K, N |ID )), the Step4: when receiving the messages, the reader decodes the 1 R R R R attacker discloses ID first part of the messages to obtain (Nbr, N , and ID ). R W W In Step3, M � ((Nbr, NW, IDW) K, HMAC (K, NR| Step5: after verifying successfully, the reader calculates 2 NW|IDW)), the attacker discloses IDW the key value K′ using N and sends the messages M � (Seq , HMAC (K′, N | Seq )). 1 W 1 In Step7, (K′, Seq1, Nbr), the attacker discloses K′ 3 Journal of Healthcare Engineering 7 IMD WISP RFID reader Programmer M : ⟨N , ID , flag⟩ 1 R R Record ECG Record ECG Create vault M : ⟨Vault V⟩ Identify common features Q compute key: K = H (Q) bio M : ⟨ID , I, HAMC(K , |Q| ID )⟩ 3 R bio R Compute key: K = H (Q) bio Select N M : ⟨N , ID  , HAMC (K ,N | N | ID )⟩ 4 W W K bio R W W bio Compute key: Compute key: K = H (K | N ) K = H (K | N ) bio W bio W K′ = H (K | N ) K′ = H (K | N ) bio W bio W M : ⟨Seq , HMAC, (K′, N | Seq )⟩ 5 1 W 1 ⟨K′, Seq ⟩ Verify K′ 1 ⟨K′, Seq ⟩ Secure communication using K′ Figure 4: Mutual authentication protocol in the emergency mode (protocol 1). 3.6.2. "e Tracing Attacks. In order to simplify the analysis 3.6.3. Medical Framework Based on RFID, Blockchain, and process, the steps 3–6 in Figure 5 are omitted here. ,e Artificial Intelligence. At present, amounts of patients have tracing attacks have three phases. the comprehensive datasets which consist of clinical history (the genetic, lifestyle data, drug, and blood biochemistry). In (1) ,e testing phase: the attacker chooses the target tag T . addition, the consumer companies and the pharmaceutical 1 1 1 ,en, she/he monitors the first round ( M , M , M , 1 2 3 are willing to pay much money for the vast personal 1 1 1 M ) to T and obtains the outputs keys ( ID , ID ). 4 R W physiological signal data applied to train their AI model via (2) ,e tracing attacks phase: we assume that the tag set using the machine learning. We proposed the medical 0 4 i ∗ (T , T ,. . . T ) includes T and the counterfeit tag T′. framework based on RFID, blockchain, and artificial in- 2 2 ,e attacker monitors the keys ( ID , ID ) in the R W telligence, as in Figure 6. second round. Previous researches based on RFID, blockchain, and artificial intelligence mainly focused on the medical ap- (3) ,e decision phase: the adversary obtained the values 1 1 2 2 1 ( ID , ID ) and ( ID , ID ). If ( ID , plication, respectively. ,e studies improve the time R W R W R 1 2 2 ID )≠ ( ID , ID ), the attacker confirms that T′ is proficiency of physiological signal data processing and W R W ∗ 1 1 2 not T with the probability 1; if ( ID , ID ) � ( ID , contribute to medical data management by combining R W R i ∗ ID ), the attacker makes sure that T is T (the three technologies. ,e effectiveness of the medical counterfeit tag T′). ,erefore, the original protocol in framework involves low resource usage, large computation the regular mode does not meet the weak indistin- time, more energy, less power, and low memory con- sumption (Algorithm 1). guishability property and suffers from the tracing attacks. 8 Journal of Healthcare Engineering IMD RFID reader Programmer WISP M′ : <N , ID , flag, HAMC (K, N | ID > Key K 1 R R R R Request Check remaining Nbr key lifetime [Invalid key] ⟨Deny, flag⟩ [Valid key] Select random M′ : <{Nbr, N , ID } , HAMC (K, N | N | ID )> 2 W W K R W W Compute session key Compute session key K′ = H (K | N ) K′ = H (K | N ) W W M′ : <Seq , HMAC, (K′, N | Seq )> 3 1 W 1 <K′, Seq > Verify K′ update Nbr <K′, Seq , Nbr> Secure communication using K′ Figure 5: Mutual authentication in the regular mode (protocol 2). Application Machine Feature reverse learning extraction engineering classification TAG Patient Data collection Reader Blockchain Figure 6: ,e medical framework based on RFID, blockchain, and artificial intelligence. Journal of Healthcare Engineering 9 ,e proposed protocol in the emergency mode (Figure 7) is as follows: (1) Step 1 (2) ,e reader initially generates the random numbers (N , ID , flag � 1); R R (3) Calculate A � N ⊕ ID ; R R (4) Broadcast M ; (5) Step 2 (6) Compare ID ; (7) if ID ≠ A⊕ N then R R (8) Process termination; (9) else (10) { (11) set up V�RandPermute (F ∪ F ); W W (12) Send M to reader (13) M in V; (14) Step 3 (15) for each f do (16) if f � F then r R (17) ,e reader and the tag match each other; (18) Calculate K � H (Q || N ); bio R (19) Send the message M � (I HMAC(K , I|Q|ID )); 3 bio R (20) Step 4 (21) If the number of matched characteristics is greater than the predetermined threshold in WISP (22) Calculate K � H (Q||N ); bio (23) if K � Kbio then bio (24) if HMAC(K , I|Q|ID ) � HMAC(K , I|Q|ID ); bio R bio R (25) Verify success, generate random number N , Calculate B � N ⊕ ID ; W W W (26) Calculate K � H (K | N ), and new key K’ � H (K | N ); bio W W (27) Send S1 � HMAC(K , N | N |ID ) M � <􏼈 N , ID 􏼉 , HMAC(K , N |N |ID )> bio R W W 4 W W K bio R W W bio (28) Step 5 (29) if K (reader) � K (tag), obtain ( N , ID ); bio bio W W (30) Calculate S2 � HMAC(K , N | N |ID ); bio R W W (31) if S2 � S1 then (32) Calculate (K, K′) K � H (K | N ), K’ � H (K | N ); bio W W (33) Send < Seq , HMAC(K‘, N | Seq )> to WISP 1 W 1 (34) Step 6 (35) WISP verifies the session keys’ equality calculated by both sides (WISP, reader) (36) If the session keys calculated on both sides are equal (37) WISP records (K′, Seq1) to awaken the IMD antenna (38) When IMD detects the request, begins to collect (K′, Seq1), and employs them to exchange data securely with the programmer (39) }; ALGORITHM 1: ,e suggested mutual authentication protocol in the emergency mode. new nonce N′ R in the second round, the attacker cannot 4. Security and Performance Analysis of counterfeit the original reader. Protocol 3 and Protocol 4 ,e protocol 3 and protocol 4 are more suitable to store 4.1.2. Key Leak Attack Resistance. In order to resist the key physiological signals in medical applications. leak attacks, WISP calculates B � N ⊕ ID ; the reader W W calculates K � PRNG (K ||N ); and K′ � PRNG (K||N ). bio W W 4.1. Security Analysis for Protocol 3. Scheme 3 overcomes the weaknesses of protocol 1, and the protocol 4 overcomes the 4.1.3. Provision of Data Integrity Verification. In order to weaknesses of protocol 2. meet data integrity, the protocol 3 has used HMAC hash calculation to protect the integrity of messages (K1, Seq ). 4.1.1. "e Reader Impersonation Attacks Resistance. In order to resist the reader impersonation attacks, the reader 4.1.4. Provision of Scalability and Efficiency. In order to calculates K � PRNG (Q||N ) using N . Even if the at- satisfy the scalability, each tag identifier does not match the R R bio tacker monitors the output information ( K � PRNG (Q|| corresponding key in DB. ,erefore, the identifications of bio 2 2 N′ ), M �(ID , I, HMAC ( K , I||Q||ID ))) using the tag keys do not match one by one in DB of the improved R 3 R bio R 10 Journal of Healthcare Engineering IMD WISP RFID reader Programmer Step 1.1 M : <N , A, flag > Step 1. Record ECG 1 R A = N + ID R R Step 2. Record ECG create vault Step 2.1 M : <Vault V> Step 3. Identify common features Q compute key: K = PRNG (Q || N ) bio R Step 3.1 M : <I, HMAC (K , I | Q | ID )> 3 bio R Step 4. Compute key K = PRNG (Q || N ) bio R Select B = ID N W W Step 4.1 M : <{N , B}K , HMAC (K , N | N | ID )> 4 W bio bio R W W Step 5. Compute keys Compute keys K = PRNG (K | N ) bio W K = PRNG (K | N ) bio W K′ = PRNG (K | N ) K′ = PRNG (K | N ) K1 = K′ N Step 5.1 M : <Seq , HMAC (K′, N | Seq )> 5 1 W 1 Step 6. Verify K′ K2 = K′ + N Step 6.2 Step 6.1 <K2, Seq > <K1, Seq > Secure communication using K′ Figure 7: ,e proposed mutual authentication protocol in the emergency mode (protocol 3). protocol, which guarantees the efficiency of tag authenti- 4.2. Security Analysis for Protocol 4 cation and satisfies the scalability property. 4.2.1. Secret Key Disclosure Attacks Resistance. In order to achieve anonymous and privacy requirements in improved protocol 4, the protocol uses the XOR function to encrypt 4.1.5. Replay Attacks Resistance. ,e attacker replays the the transmitted keys as follows: messages to authenticate by monitoring the previous in- formation. In order to resist replay attacks, all messages are B � ID ⊕ N , K1 � K′⊕ N , K2 � K′⊕ N . W W R W encrypted by using the random numbers (Nbr, N , and N ) W R and combined with PRNG function. 4.2.2. Tracing Attacks Resistance. ,e key updating mecha- th nism K′ � PRNG (K|N ) involves the i keys and the nonces th th (N , K). ,e i key K cannot be cracked by the (i+1) keys 4.1.6. Provision of Data Integrity Verification. In order to W i th achieve the property of data integrity, we have used K and the i sessions. ,e reasons are that PRNG functions i+1 protect the parameters by the encrypted messages. ,erefore, PRNG calculation K′ � PRNG (K|N ) to protect the in- tegrity of K′. the enhanced protocols resist the tracing attacks. Journal of Healthcare Engineering 11 RFID IMD WISP Programmer reader Key K Step 1.1 Step 1. A = ID N M′ : <{N , A, flag, HMAC (K, N | ID )> R R 1 R R R Request Step 1.2 Check remaining Nbr key lifetime [Invalid key] <Deny, flag> [valid key] Step 2. Select random N compute B = N + ID W W Step 2.1 M′ : <{Nbr, N , B} , HMAC (K, N | N | ID )> 2 W K R W W Step 3. Compute session key K′ = PRNG (K | N ) K1 = K′ + N Step 3.1 M′ : <Seq , HMAC (K′, N | Seq )> 3 1 W 1 K′ = PRNG (K | N ) K2 = K′ N Step 4.2 Step 4.1 <K2, Seq , Nbr> <K1, Seq > Secure communication using K′ Figure 8: ,e proposed mutual authentication protocol in the regular mode (protocol 4) (Algorithm 2). 4.2.3. Availability and Desynchronization Attacks Resistance. 5. Blockchain Framework for Security and In order to provide anonymity, the communication com- Privacy Storage and Sharing ponents (tag and DB) update the shared messages after completing the conversation. If the opponents destroy the A framework is developed to share physiological signals’ updating process, the authentication scheme is subjected to cross domain and build the radiological studies’ ledger and desynchronization attacks. In order to guarantee the con- patient-defined access permissions by applying the block- fidentiality and anonymity of K, the messages synchronously chain as the distributed data store. Relative disadvantages of should be updated. In addition, the attacker knows the the framework include the privacy’s complexity and security shared key K′ during the updating processes, which is models. Ultimately, the large-scale feasibility of the approach protected by the random numbers (N , N ). ,e improved remains to be demonstrated. W R protocol is desynchronization resistance. ,e peculiar health-care technologies are required, such as parallel processing, distributed data network, scalable storage, frameworks, and infrastructures. ,e fog 4.3. "e Comparisons of Security and Performance Analysis. computing is economical and customizable, since fog Table 3 lists the computational cost for five protocols. ,e computing handles these complex problems in the virtual computational costs of tags in protocol 3 are 3PRNG + Xor, environment and only needs to pay for the used services and the computational costs of tags in protocol 4 are and resources. 2PRNG + Xor. ,e safety performances of the enhanced ,e sharing physiological signals systems are important protocols are superior to other schemes. Compared with the in different medical institutions, but the current infra- original protocol 1 and protocol 2, the improved protocols structure for transmitting physiological signals relies on support the security enhancements and ensure the function the trust third-party intermediaries. We propose the such as integrity, efficiency, and user privacy. framework of cross-domain sharing image where the 12 Journal of Healthcare Engineering ,e proposed mutual authentication protocol in the regular mode is in Figure 8 as follows: (1) Step 1 (2) ,e reader generates (N , ID , flag � 0); R R (3) Calculate A � ID ⊕ N and K � H (N |ID ); R R R R (4) Transmit M � < N , ID , flag, HMAC(K, N | ID )> 1 R R R R (5) When WISP receives the request, it confirms that the primary key is expired and verifies that how many session keys which originated from the primary key exceeds the predetermined threshold (6) if t< T then If the primary key has not expired, WISP receives the messages (7) else the key expired, access denied; (8) Step 2 (9) After WISP successful authentication, the random number N is generated; (10) Calculate K’ � H (K | N ), B� ID ⊕ N ; W W W (11) Transmit M � < {Nbr, Nw, I Dw} , HMAC(K, N |N |ID )> 2 k R W w Calculate S1 � HMAC(K, N |N |ID ); R W w (12) Step 3 (13) After receiving the messages, the reader starts to parse the first part of the message through the key K to obtain (Nbr, NW, IDW); (14) Calculate S2 � HMAC(K, N |N ID ); R W w If S2 � S1 then ,e message is true; Calculate K’ � PRNG (K |N ), K1 � K’⊕ NR; (15) Transmit M � Seq , HMAC(K , N |Seq ) 3 1 w 1 (16) Step 4 (17) Based on the received HMAC, WISP can confirm the timeliness of the message and the equality of the session keys calculated on both sides (18) After verifying successfully, Nbr++, K2 � � K’⊕ NW; (19) WISP records (K2, Seq1, Nbr) to awaken the IMD antenna (20) When IMD detects the request, collects (K1, Seq1), and employs them to exchange data securely with the programmer. ALGORITHM 2: ,e proposed mutual authentication protocol in the regular mode. blockchain is used as the distributed data storage to es- 5.1. Physiological Signals Data Sharing Model Based on tablish patient-defined access rights. ,e blockchain Blockchain [22]. Intelligent contract based on blockchain is framework is verified to eliminate the access permission of used to promote the security analysis and management of the third-party to protected physiological signal infor- medical sensors. Intelligent device invokes intelligent con- mation, meets many standards of the interoperable medical tract and writes records of all events on blockchain. ,e system, and easily generalizes to fields beyond physio- intelligent contract systems support real-time patient logical signal. We summarize the framework based on monitoring and medical intervention by sending notifica- blockchain to allow patients to securely grant electronic tions to patients and medical professionals. ,e provider of medical records can modify the physiological signals, but it access permission to their physiological signal data and describe the advantages and disadvantages of the approach. needs patient’s consent, and the patient can assign access ,e actual transmission of physiological signals re- authority to medical records. quires the physiological signals receiver who transmits the When applying blockchain to the construction of the signed request to the URL endpoint. ,e individual ser- credit system, we promote the collection and supervision of vice is the requesting entity that the access permission of credit information in the medical field and build the new the physiological signals study is authorized to by the relationship platform. It is significant to the improvement of owner (patient). ,e studies of all patients’ physiological the credit system construction. According to the unified signals result in the huge blockchain, far too large to evaluation criteria, the credit rating is evaluated, the result of download, store, and validate for nodes running on the rating level is publicized on the platform of block chain, mobile devices. ,e size of the blockchain has been proven the credit rating is rewarded, and the violation of credit is punished, so as to strengthen the construction of the credit to be the limiting element for chains storing the trans- actional data. system in the medical field in the real sense. Considering all of these factors, sharing the physiological ,e asymmetric information encryption methods need signals by using blockchain helps the interoperable health two keys: public key and private key. After the physiological system and has greater ability to access patients’ physio- signals are encrypted with public key, only the corre- logical signals electronically. sponding private key can be used for decryption. On the Journal of Healthcare Engineering 13 IMD WISP Programmer RFID reader M′ : ⟨N , C, flag, HMAC (K, N | ID )⟩ 1 R R R Request Key K C = ID ID Check remaining R R key lifetime Nb r + ID = CN R R [Invalid key] <Deny, flag> [Valid key] Select random M′ : ⟨{Nbr, N , ID } , HMAC (K, N | N | ID )⟩ 2 W W k R W W K′ = PRNG (K | N ) K′ = PRNG (K | N ) W + Nbr K1 = K′ M′ : ⟨Seq , HMAC (K′, N | Seq )⟩ 3 1 W Verify K′ ⟨K1, Seq ⟩ update Nbr K1 = K′ Nbr ⟨K1, Seq , Nbr⟩ Secure communication using K1 Figure 9: Secure communication protocol between the IMD and the programmer (protocol 5) (Algorithm 3). contrary, if the private key is used to encrypt data, only the medical diagnosis in the hospital, which confirms the corresponding public key can be used for decryption. If the statement and promised to provide the study in the blockchain can be grafted, scientific research institutions previous block. ,e patient’s signature declaration is understand the probability of disease occurrence, the oc- obtained through the mobile application, which shares currence of accidents, the level of hospital management, and and stores the values required allowing access to the claims cases and other actual situations. transaction in the future. ,en, the hospital signs the follow-up information of the patients and broadcasts (1) Use the fog-based blockchain and fog warehouse to the transaction to the blockchain. store medical data, as shown in Figure 10. (3) Allow access: the transaction allows the owner of the (1) List of medical research and patients in each medical information research to authorize the other institute. party to retrieve its medical data. Patient K signs a transaction to grant the function to doctor K . ,e (2) Patients are authorized to access the entity set of signed verification blocks are embedded in block- each study. ,e entities are represented by the chains. As shown in Figure 11, patients publish the common part of the asymmetric key pair on the transaction after verifying the key with the doctor blockchain. through the APP platform. ,e patient can be au- (2) Definition study: the transaction builds the patient as thorized to the legitimate doctor or institution, and the the master of a UID which is the specific unique doctor can associate any medical information received identifier and the source as the creator. Tuples stored with the correct local medical record number. in block chains are transactions with double signa- ,e middle column (Block Chain Medical Data Sharing tures, similar to documents with signatures from Sequence) describes the interaction between entities and patients and hospital representatives. ,e patients judgments in each stage and reflects the sharing medical claim that the definition study has received the 14 Journal of Healthcare Engineering ,e programmer can use the session key calculated by the protocol to establish the secure communication after IMD authenticates the programmer in Figure 9. (1) Step 1: (2) ,e reader initially generates (K, C � ID ⊕ N ) and transmits the values M � (N , C, flag, HMAC (K, N | ID )) to the WISP. R R R R R (3) Step 2: (4) ,e IMD returns Nbr and updates ID � C⊕ N . R R (5) Step 3: (6) If the key is valid then (7) ,e WISP selects NW and transmits the values M �((Nbr, N , ID ), HMAC (K, N | N |ID )) to the reader. W W R W W (8) else (9) ,e WISP transmits the sequences (Deny, flag) to the reader. (10) Step 4: (11) ,e WISP updates K′ � PRNG (K | NW), and the reader updates K′ � PRNG (K | N ) and K1 � K’⊕ Nbr. ,e reader sends the value M3 � (Seq1, HMAC (K′, N , Seq1)) to the WISP and sends the messages (K1, Seq1) to the programmer. (12) Step 5: (13) ,e WISP identifies K′ by comparing the value K′ of the WISP with the K′ value of the reader. ,e WISP updates Nbr, K � K’⊕ Nbr, and sends (K , Seq1, Nbr) to the IMD. 1 1 ALGORITHM 3: Secure communication protocol between the IMD and the programmer. Table 3: ,e comparisons of the performance analysis and safety performance. Performance Protocol 1 Protocol 3 Protocol 2 Protocol 4 Protocol 5 F0 No Yes No Yes No F1 3H + Xor 3PRNG + Xor 2H + Xor 2PRNG + Xor 1PRNG+2Xor F2 No Yes No Yes Yes F3 No Yes No Yes Yes F4 No Yes No Yes Yes Attack types Protocol 1 Protocol 3 Protocol 2 Protocol 4 Protocol 5 R1 No Yes No Yes No R2 Yes Yes Yes Yes Yes R3 Yes Yes Yes Yes Yes R4 Yes Yes Yes Yes Yes R5 No Yes No Yes Yes R6 No Yes Yes Yes Yes F0: provision of scalability and efficiency; F1: storage cost (tag); F2: blockchain-enabled; F3: cloud computing-enabled; F4: fog computing-enabled. R1: key leak attacks resistance; R2: replay attacks resistance; R3: desynchronization attacks resistance; R4: reader impersonation attacks resistance; R5: tracking attacks resistance; R6: tag impersonation attacks resistance. information by supporting distributed block chains and out- which is used to construct the blockchain for medical data of-block transactions. sharing. ,e actual medical data transmission requires the We have showed the technology fundamentals of medical data receiver to deliver the signature request to the blockchain and provided a summarization of the blockchain medical source’s URL endpoint which creates the research. application that can be used as a tool to allow the patient- Both requests and responses are transmitted through the controlled, physiological signal’s cross-domain sharing secure link of the transport layer to prevent eavesdropping. without the central authority. In particular, we highlighted ,e effective blocks are generated in the timely manner by the way blockchain satisfies many requirements of the in- generating the distributed database with access permissions teroperable health system. However, these technologies also and stimulating the block generator in some way. Only those have several important limitations, and the relative merits of nodes with security deposits can participate in the expansion existing alternatives must be considered before any large- of the chain, and any node with misconduct will be forced to scale and blockchain-based application for sharing physi- ological signals. abandon its investment. ,e nature of blockchain provides the direct audit of the activity of each node such as the When receiving query request, the physiological signal number of blocks generated and the failure status of the data source verifies the correctness of the signature, ensures blocks generated. ,e node operator can prove the node that the hashed data matches the previously published data ownership by using the private key which is corresponded to for K -owner via Block B, and confirms that the K -owner P P the identity public key of the node to sign the message. ,e has allowed the requestor access to these physiological signal enhanced model adds the fog computing in the original data via Block C. If meeting all the conditions, the response blockchain medical data sharing the sequence model [22], containing the physiological signal study is returned from Journal of Healthcare Engineering 15 Blockchain and fog warehouse Blockchain Public and Hash history technology private keys Structured data types of physiological signal The serial ID of the patient in hospital number Fraction MRI, ECG, Heart, Rate, Signal, Image Websites My webpage com User authentication layer Figure 10: Blockchain based on fog warehouse. Block/time A CD kD kD kp kH kH kp kH Fog query computing Actors Physiological signals Physiological signals Step4: Physician: I am K-owner Step7: Step2: and need to review imaging from On-blockchain Physician uses the information in Patient: e physiological transaction your hospitalization. blocks (A,C) and to submit an query signals were acquired for Kp- Step5: request for physiological owner by K-owner Off-blockchain Assertions Patient: Agreed, I am K-owner signals, signed by K. Step3: commuication and will allow access Step8: Hospital: K-owner’s Hospital valid at the Physician’s Step1: assertion above is accrate and Step6: signature,uses the information in Hospital: kH-owner will K-owner will share these data Patient: As the K-owner blocks (B,C) and to confirm service physiological signals at its establised end point I permit the K-owner to access the authorization, and transmits the retrieval request at https. physiological signals that were physiological signals study in an query acquired for me by K-owner. response. Figure 11: Blockchain medical data sharing sequence diagram based on fog computing. the source. In order to prevent eavesdropping, the requests Step 1: for hospital (K -owner), K -owner will service H H and responses are sent to prevent eavesdropping. ,e spe- physiological signals retrieval requests at https by using cific steps of blockchain medical data sharing sequence on-blockchain transaction and off-blockchain diagram are as follows: communication. 16 Journal of Healthcare Engineering Step 2: for the patient (K -owner), the physiological transform healthcare management. In the future, we will signals are acquired for KP-owner and KH-owner. focus on heterogeneous physiological signal data issues through fog computing, blockchain, and AI technology in Step 3: for hospital, K -owner’s assertion is accrate and the realistic medical environment. KH-owner shares the physiological signals at the established endpoint. Data Availability Step 4: for physician (K -owner), K -owner reviews D D the physiological signals from the hospitalization. ,e paper gives an outline about the framework, and in- Step 5: for patient, if the patients agree, they are K - ternal working and protocols for handling heterogeneous owner and will allow access. physiological signal data. Once the hybrid technologies are Step 6: for patient, the patient permit K -owner to integrated, big data systems and AI technology have the access the physiological signals that were acquired by potential to offer privacy protection and data sharing, KH-owner. transform healthcare management. Step 7: physician uses the information in blocks (A, C) to submit the query request for physiological signals, Conflicts of Interest signed by K . ,e authors declare no conflicts of interest. Step 8: hospital valid at the physician’s signature, uses the data in blocks (B, C) to confirm authorization and Acknowledgments transmits the physiological signals study in the query response. ,e requests are sent by the K -owner at ,is work was supported in part by Jiangsu Postdoctoral timepoint D. Science Foundation (Grant nos. 1701061B and 2017107007); ,e ecosystem is consisted of the blockchain nodes and Xuzhou Medical University Affiliated Hospital Postdoctoral fog storage. For example, one of the main reasons for in- Science Foundation (Grant nos. 2016107011, 183822, corporating fog storage technology into the ecosystem is to 53120225, and 53120226); Xuzhou Medical University Ex- supply the offline storage solution, especially for large cellent Persons Scientific Research Foundation (Grant nos. physiological signals. For security and privacy, the client side D2016006, D2016007, and 53591506); the Practice Inovation would encrypt the physiological signals uploaded to the fog Trainng Program Projects for the Jiangsu College Students storage. With the maturity of the fog storage, personal (Grant nos. 20161031308H and 201610313043Y); the Natural storage may be replaced by it. Science Foundation of the Jiangsu Higher Education In- Most significantly, blockchain technology can create the stitutions of China (Grant no. 16KJB180028); and 333 physiological signal-driven marketplace, where patients can Project of Jiangsu Province (no. BRA2017278). get real return by offering their data to research institutions, pharmaceutical and consumer companies, the application References development community, and producing new physiological signal data. [1] J. P. Rajan and S. E. Rajan, “An internet of things based physiological signal monitoring and receiving system for virtual enhanced health care network,” Technology and Health 6. Conclusions Care, vol. 26, no. 2, pp. 379–385, 2018. [2] O. Faust, Y. Hagiwara, T. J. Hong, O. S. Lih, and We extend the architecture of the IMD with blockchain, U. R. Acharya, “Deep learning for healthcare applications RFID, and WISP, which increases the physiological signal based on physiological signals: a review,” Computer Methods data’s confidentiality and authenticity. ,e enhanced RFID and Programs in Biomedicine, vol. 161, no. 1, pp. 1–13, 2018. protocols provide protection against tracking attacks, [3] R. Shanthapriya and V. Vaithianathan, “ECG-based secure readers’ impersonation attacks, and secret disclose attacks. healthcare monitoring system in body area networks,” in ,e physiological signal records have proved the im- Proceedings of the 2018 Fourth International Conference on portance for the patients, and sharing and acquiring Biosignals, Images and Instrumentation (ICBSII), pp. 206–212, physiological signals is essential for intelligent and advanced IEEE, Montreal, Canada, October 2018. medical services. ,e blockchain application of e-commerce [4] C. Orphanidou, “A review of big data applications of phys- iological signal data,” Biophysical Reviews, vol. 11, no. 1, has proven that trusted and auditable transaction in peer-to- pp. 83–87, 2019. peer networking is possible. In the paper, we have intro- [5] E. O. Tartan and C. Ciflikli, “An android application for duced a blockchain-based architecture model for physio- geolocation based health monitoring, consultancy and alarm logical signal data on fog computing environment. Our system,” in Proceedings of the IEEE 2018 IEEE 42nd Annual contributions are mainly consisted of the proposed solution Computer Software and Applications Conference (COMPSAC), and introduction to future medical data directions in pp. 341–344, IEEE Computer Society, Tokyo, Japan, July 2018. blockchain. ,e paper proposes the outline to show the [6] L. Yuri alvarez, ´ F. Jacqueline, N. Guillermo alvarez ´ et al., framework and schemas for dealing with heterogeneous “RFID technology for management and tracking: e-health physiological signals. Once the hybrid technologies are in- applications,” Sensors, vol. 18, no. 8, pp. 2663–2678, 2018. tegrated, big data systems and AI technology have the po- [7] M. Mart´ınez Perez, ´ C. Dafonte, and A. Gomez, ´ “Traceability tential to offer privacy protection and data sharing and in patient healthcare through the integration of RFID Journal of Healthcare Engineering 17 technology in an ICU in a hospital,” Sensors, vol. 18, no. 5, [23] C. Tang, C. Li, X. Yu, Z. Zheng, and Z. Chen, “Cooperative pp. 1627–1641, 2018. mining in blockchain networks with zero-determinant [8] T. Adame, A. Bel, A. Carreras, J. Melia-Segu ` ´ı, M. Oliver, and strategies,” IEEE Transactions on Cybernetics, vol. 2, no. 3, pp. 1–6, 2019. R. Pous, “CUIDATS: An RFID–WSN hybrid monitoring system for smart health care environments,” Future Gener- ation Computer Systems, vol. 78, no. 2, pp. 602–615, 2016. [9] H. Q. Omar, A. Khoshnaw, and W. Monnet, “Smart patient management, monitoring and tracking system using radio- frequency identification (RFID) technology,” in Proceedings of the Biomedical Engineering and Sciences, vol. 1, no. 2, pp. 1–12, IEEE, Kuala Lumpur, Malaysia, December 2016. [10] N. Ellouze, S. Rekhis, N. Boudriga, and M. Allouche, “Powerless security for cardiac implantable medical devices: use of wireless identification and sensing platform,” Journal of Network and Computer Applications, vol. 107, no. 1, pp. 1–21, [11] C. Xu, K. Wang, and M. Guo, “Intelligent resource man- agement in blockchain-based cloud datacenters,” IEEE Cloud Computing, vol. 4, no. 6, pp. 50–59, 2018. [12] Z. Aiqing and L. Xiaodong, “Towards secure and privacy- preserving data sharing in e-health systems via consortium blockchain,” Journal of Medical Systems, vol. 42, no. 8, pp. 140–154, 2018. [13] A. Dubovitskaya, Z. Xu, S. Ryu et al., “Secure and trustable electronic medical records sharing using blockchain,” in Proceedings of the AMIA. Annual Symposium proceedings/ AMIA symposium, vol. 1, no. 1, pp. 1–13, AMIA Symposium, Washington, DC, USA, January 2017. [14] C. S. Lebech, S. M. Nibe, K. ,omas, P. Vestergaard, and O. Hejlesen, “How to use blockchain for diabetes health care data and access management: an operational concept,” Journal of Diabetes Science and Technology, vol. 13, no. 2, pp. 1–14, 2018. [15] X. Yue, H. Wang, D. Jin, M. Li, and W. Jiang, “Healthcare data gateways: found healthcare intelligence on blockchain with novel privacy risk control,” Journal of Medical Systems, vol. 40, no. 10, pp. 218–242, 2016. [16] B. M. N. Kamel, J. T. Wilson, and K. A. Clauson, “Geospatial blockchain: promises, challenges, and scenarios in health and healthcare,” International Journal of Health Geographics, vol. 17, no. 1, pp. 1–19, 2018. [17] T. Jen-Hung, L. Yen-Chih, C. Bin, and L. Shih-wei, “Gov- ernance on the drug supply chain via gcoin blockchain,” International Journal of Environmental Research and Public Health, vol. 15, no. 6, pp. 1055–1067, 2018. [18] K. Harleen, A. M. Afshar, J. Roshan, A. K. Mourya, and C. Victor, “A proposed solution and future direction for blockchain-based heterogeneous medicare data in cloud en- vironment,” Journal of Medical Systems, vol. 42, no. 8, pp. 156–167, 2018. [19] J. J. Gu, R. C. Huang, L. Jiang, G. Qiao, X. Du, and M. Guizani, “A fog computing solution for context-based privacy leakage detection for android healthcare devices,” Sensors, vol. 19, no. 5, pp. 1–16, 2019. [20] C. A. Silva, G. S. Aquino, S. R. M. Melo, and D. J. B. Eg´ıdio, “A fog computing-based architecture for medical records man- agement,” Wireless Communications and Mobile Computing, vol. 2019, no. 1, pp. 1–16, 2019. [21] Y. Guan, J. Shao, G. Wei, and X. Mande, “Data security and privacy in fog computing,” IEEE Network, vol. 32, no. 5, pp. 1–6, 2018. [22] V. Patel, “A framework for secure and decentralized sharing of medical imaging data via blockchain consensus,” Health Informatics Journal, vol. 1, no. 1, pp. 1–14, 2018.

Journal

Journal of Healthcare EngineeringHindawi Publishing Corporation

Published: Apr 14, 2020

References