Escrever um comentário

Summary:

This research mainly explores the advantages and disadvantages of using Artificial Intelligence (AI) in healthcare and examines how cybersecurity tools can help reduce the risks associated with AI-based healthcare systems. The study is important because, as AI becomes increasingly integrated into healthcare applications for disease diagnosis and patient management, protecting patient data and ensuring cybersecurity remain major challenges. The work provides insights into how integrated cybersecurity techniques can be applied to improve the security of healthcare applications that use AI.

The authors conducted a literature review using recent studies from sources such as MDPI, Institute of Electrical and Electronics Engineers (IEEE), and Springer. They also used comparative synthesis methods in Table 1, to evaluate different cybersecurity approaches.

The study found that although AI in healthcare has both benefits and risks, many of the data security challenges can be reduced through the combined use of cybersecurity tools. The authors highlighted cryptographic and defense techniques such as Blockchain, Zero-Knowledge Proofs (ZKP), and Honeypots, which achieved hybrid system security scores of 0.85, 0.75, and 0.65 respectively. According to the study, these techniques may play an important role in the future protection of AI-based healthcare applications.

One interesting aspect of the study is the discussion of Blockchain as a strong tool for improving data security and transparency in healthcare AI systems. The paper also highlights the potential of mathematical hybrid security models and explains how AI tools can speed up medical tests, scans, and diagnoses compared to manual methods.

A major strength of the study is its comprehensive and up-to-date literature review, as well as its clear integrative framework. However, an important weakness is the lack of empirical validation or real-world data implementation to support the proposed solutions. Another limitation is that the three major cryptographic techniques discussed still have practical limitations that may affect their ability to fully address healthcare data security challenges.

List of major concerns and feedback:

• The preprint does not appear to follow PRISMA guidelines for the selection, screening, and reporting of literature included in the review. This affects the transparency and reproducibility of the literature review process. The authors should revise the methodology section in line with PRISMA standards and include a PRISMA flow diagram showing the number of studies identified, screened, excluded, and included.

• The preprint does not discuss risk of bias assessment using established tools such as the Cochrane Risk of Bias framework or other quality assessment methods. The authors should evaluate the quality and potential bias of the 10 included studies and their findings summarized in table 1. This would improve the reliability and credibility of the review.

• The study is entirely theoretical and does not include real-world data testing, simulations, or experimental validation of the proposed cybersecurity methods. This limitation should be acknowledged in the manuscript. If possible, future work should include empirical testing, or case studies to demonstrate how the proposed security techniques perform in real healthcare AI systems.

• Table 1 is informative but too wide. It also lacks a suitable title, making interpretation difficult. No quantitative data to support how the authors assessed the 10 studies included in the table.

• In the review section on Page 3, the statement “Papers we analyze focused on hybrid security frameworks that integrate multiple advanced security techniques as we can see on Figure 1” is not fully supported by Figure 1. The figure does not clearly present the advanced security techniques mentioned in the text. The authors should revise Figure 1 to clearly identify and label the advanced security techniques discussed or modify the text so that it accurately reflects the information presented in the figure.

List of minor concerns and feedback:

• In Figure 3, the phrase “The overall security score of your hybrid system—combining Blockchain, Zero-Knowledge Proofs (ZKP), and Honeypots” uses the word “your,” which appears inappropriate in academic writing. Replace “your” with “our” to improve the readability of the preprint.

• The sentence on Page 3, “Were inspected mainly newer works (to get the latest information) and using different keywords relevant for attacks”, contains grammatical and typographical issues that reduce readability. The sentence should be rewritten for clarity.

• In the Discussion section, the statement “The discussion surrounding modern healthcare security in the 4.0 era” is unclear and may confuse readers unfamiliar with the term. Replace “4.0 era” with “Industry 4.0 era” or provide a brief explanation of the term within the text.

• Under the future work section, the word “privacy” appears to have been used incorrectly. The correct word would be “private”.

Limitations and ethics discussed:

• Limitations of the proposed cyber security techniques (ZKP, Blockchain, Honeypots) were discussed. However, concerns about their experimental validation and real-world application were not adequately addressed.

• The study is a review article and does not appear to violate ethical standards. However, the preprint does not explicitly include an ethical guidelines statement.

List of additional comments and feedback:

• A few typographical and grammatical errors should be corrected to improve the readability and overall quality of the manuscript.

• Some paragraphs are too long and may be difficult for readers to follow. The authors should consider breaking them into shorter paragraphs and simplifying some of the technical explanations to improve clarity and make the manuscript more accessible to a broader audience.

• One of the most interesting aspects of the study is how AI tools can improve healthcare services by speeding up the processing of medical tests, scans, and other clinical procedures compared to traditional manual methods.

• The use of Blockchain technology is also very interesting. It appears to be a robust tool for improving data security and transparency in AI-based healthcare systems. Due to its decentralized and immutable nature, it may help reduce the risk of adversarial attacks and unauthorized data manipulation.

• Another notable aspect of the research is the proposed mathematical hybrid security model, which provides an integrated approach to strengthening cybersecurity in healthcare AI applications.

• The preprint relates well to existing literature on AI security and healthcare data protection. It confirms known challenges such as the vulnerability of AI systems and the need for stronger cybersecurity measures. It also highlights future research directions, especially the need for real-world testing of hybrid security frameworks and the development of more energy-efficient Blockchain systems for healthcare applications.

• The conclusions are supported by the reviewed literature, especially regarding the potential effectiveness of hybrid cybersecurity systems. However, because the study lacks empirical testing or real-world validation, the conclusions are more suggestive than definitive. The preprint mainly reveals that these approaches could be effective rather than demonstrating that they have been conclusively proven to work in practice.

• The preprint may be suitable especially for researchers in AI healthcare security and cybersecurity. However, it may be less suitable for those seeking strong experimental or empirical evidence, as the study is only theoretical. Publication is recommended provided that the concerns raised are adequately addressed.

Competing interests

The authors declare that they have no competing interests.

Use of Artificial Intelligence (AI)

The authors declare that they did not use generative AI to come up with new ideas for their review.