Page 39 - AIH-1-2

Artificial Intelligence in Health Blockchain for health-care security

the blockchain supply. This protects patient data integrity 4.4. PBFT
and increases the resilience of PoS-based health-care PBFT can solve the byzantine problem, as presented
blockchains against threats. For health-care applications in a published paper. A byzantine fault is a defective
30
based on blockchain, switching from complicated algorithm. Byzantine fault tolerance can ensure the safety
computational problems to POS results in lower energy and efficiency of the system so that hardly [(n-1)/ duplicate
3
consumption and enables randomized validator selection, data are defective over the system in a lifetime. In medical
node participation incentives, performance-based rewards, science, PBFT algorithms create an efficient impact
and continuous work to resolve distribution issues. because several nodes are being shared and maintained by
36
The security, effectiveness, and dependability of health- several nodes. 36,40 The fact that they hinder medical data
care blockchain applications are all improved by these from being disclosed or accessed by attackers significantly
contributions taken together. enhances the trustworthiness of PBFT.
Moreover, using the POS consensus mechanism in 4.5. RAFT
the context of an EHR system indicates that health-care
applications built on blockchain technology are more secure RAFT has five server nodes with three states, namely
and efficient in terms of making smart decisions. Real-time leader, follower, and candidate. Modified RAFT nodes
modifications to patient records can be made easier with work in a category accepting the same transitions. For
POS, which has advantages including faster transactions and instance, if a person is selected from a category assigned
less energy usage. In addition, the tasks of verifying patient as a leader, he must accept clients’ requests. 37,41 The
records are carried out by trusted health care providers who leader must replicate the log to other servers and the
use health care networks to ensure reliability and efficiency. data flow from the leader to the server. The leader’s task
37
POS is a good option for applications where timely access is divided into three subtasks: leader election, leader
to patient data is crucial and environmental concerns are log replication, and safety. A new leader is elected when
present since it combines the benefits of decentralization the assigned leader fails to monitor the works. In log
with a fast and streamlined consensus process, which replication, the leader can guide and command the
enhances the system’s overall security. followers to execute changes made by the leader. 34,41
Finally, RAFT uses different commands for the same log
4.3. DPOS index when the server changes the state of machine for
safety concerns. Figure 1 shows the process of cluster
DPOS is a decentralized model with high efficiency but algorithm of RAFT.
low consumption. There is an option to vote for creating a
panel with restricted trusted parties known as witnesses. 33,38 By comparing these five consensus algorithms, as
Some users act in the reputation system. It can create blocks shown in Figure 2, the POW algorithm stands outs as
and add them to the blockchain. The DPOS census is cost- the most efficient for the health-care sector because it
efficient and time-saving. Since few nodes are eligible has a robust security system, which is the primary goal of
for DPOS to be centralized, the central node can easily initiating blockchain algorithms in health-care sectors. The
monitor the election process. DPOS cannot maintain all summary of compared algorithms is shown in Table 2.
the nodes effectively, undercutting the trustworthiness in A previous study provides a framework for
29
security. Nowadays, the health-care domain is undergoing implementing the algorithms discussed previously, where
advancements through the incorporation of blockchain. 34,39 a number of computers with the same specifications were
The implementation of the DPOS algorithm ensures the used to act as nodes for the blockchain. By considering
38
privacy of EHRs through secure transactions. With this the typical framework with a computers of core I7, with
technology, the patients maintain control of their EHRs. the specifications of 16 GB memory size, and Window 10
The patients may share their medical records with different operating system, the experimented POW, POS, DPOS, and
institutions. PBFT algorithms with data size of 100 M/times can deliver
Blockchain technology can ensure the privacy and performance depicted in Figure 3. An extended period of
security of shared data. Once a doctor updates the EHR, it time is required to implement the proposed model in the
is encrypted by the SHA256 hashing algorithm, and then, system, as per empirical experiences, to compensate for the
it is stored in a different block. 35,40 The doctor receives a random delay between nodes.
unique key from the patients through mail for accessing the The analysis also showed that the POW algorithm
medical data. The DPOS algorithm secures the patient data takes longer time compared to DPOS and POS algorithms.
with a trustworthy guarantee and lowers the computational Furthermore, the PBFT consensus algorithm requires
time and minimizes the entire cost of processing EHRs. shorter time compared to other algorithms. The performance

Volume 1 Issue 2 (2024) 33 doi: 10.36922/aih.2580