The Risk of Surveillance and Privacy in Health Care using AI

The Risk of Surveillance and Privacy in Health Care Using AI

Artificial intelligence (AI) in healthcare undoubtedly has excellent potential in virtually all areas of healthcare, from research to direct patient care. However, the threat to patient privacy and the likelihood of patient data misuse have gained much attention and concern. Even though institutions have been formed to regulate research and ethical policies are set to prevent such occurrences, a radical shift towards more significant innovation and patient-centered solutions is needed to solve such problems. In this way, it is possible to design and develop AI systems used in healthcare with the patient’s privacy in mind and protect their data rights from the beginning instead of trying to do that after the AI system has already been created. This is why it is crucial to go beyond the mere reliance on the regulatory frameworks and ethical guidelines to address the privacy concerns of AI in the healthcare sector and instead focus on the patient-centered approach that would incorporate patients’ privacy preferences and data rights into AI systems by design.

Patient Participation as Co-Creators in AI-Enabled Health Systems

In most cases, technology firms, scientists, and healthcare professionals initiate the creation of AI technologies rather than the users, that is, the patients. This has led to the development of systems where the targeted consumer’s privacy considerations and data preferences may not effectively be met. Thus, moving to a patient-centric design, it becomes possible to leverage patients’ knowledge and experience to guide the fundamental aspects of AI-based health solutions, namely, their functions and structures.

The studies also indicate that when the patient has an active role in designing the technologies, they are likely to have confidence in the technologies and feel like the personal data belongs to them. The Journal of the American Medical Informatics Association also revealed that patients’ engagement in the creation of an AI-based clinical decision support system impacted the system’s usability, data credibility, and compliance with privacy and security standards (Choudhury, 2022). Therefore, Peters et al. (2020) reported the necessity of “co-design” AI solutions in which patients and the general public are involved in defining such tools’ values, goals, and applications. This approach not only improves the protection of privacy but also guarantees the end product to be more relevant to the needs of society.

An example of such a patient-led approach is the creation of the Apple Heart Study, a massive research endeavor that utilized the Apple Watch to identify atrial fibrillation. In this project, Apple worked with the healthcare providers and researchers not only to devise the technical details of the study but also to engage patients in the decisions about data privacy and consent (Skaria et al., 2020). Thus, by engaging the patients in the process, the study was able to establish rapport, solicit the community’s cooperation, and respect the patients’ privacy.

Incorporating Privacy-Preserving AI Architectures

Apart from just involving patients in the healthcare AI systems’ design process, the systems’ developers can also incorporate privacy-preserving architectures and techniques into the systems’ architecture. These are federated learning, differential privacy, and homomorphic encryption, novel approaches that can be adopted to train and deploy AI models without necessarily having to transfer and share patients’ information (Banabilah et al., 2022). For instance, federated learning allows AI models to be trained on distributed datasets, such as individual patient devices, without the data transfer out of the local network (Cresswell et al.,  2023. This decentralized nature of the model training process also minimizes the vulnerability of data leaks and preserves the patient’s confidentiality. Likewise, the differential privacy methods overlay the data with noise, considerably hampering the re-identification process. At the same time, the information remains valid for the AI model training.

With the above privacy principles incorporated as the basic framework of healthcare AI systems, developers can develop technologies that are inherently privacy compliant to the patient’s choice without the need for future legislation or ethical standards to control the use of personal information. This preventive measure reduces the risks associated with surveillance and data abuse and helps create patients’ trust and independence (Rafiq et al., 2022).

One use case of privacy-preserving AI in healthcare that has been partially implemented is the creation of the NVIDIA Clara platform that employs federated learning and other similar techniques to train AI models on distributed medical data sources without violating the patient’s rights to privacy (NVIDIA, 2024). NVIDIA has integrated these privacy-preserving architectures into the very fabric of the platform so that HL7’s members can use them to build AI apps that are inherently compliant with patient data rights and privacy preferences.

Promoting a Culture of Accountability and Transparency

Such approaches can only be a part of a more prominent solution, with the patients being empowered as co-designers of the new healthcare AI systems, as well as ensuring that privacy-preserving architectures are used; in addition, a drastic change of the culture of transparency and accountability needs to be promoted as well. Those who practice medicine, design AI systems, and formulate policies must join forces and help patients comprehend how their data is being utilized, which AI tools are being implemented, and the steps taken to safeguard their rights. Some of the ways that this transparency can be affected include creating awareness, educating people, and developing effective data policies that are easy to understand. Also, robust processes to ensure the patients’ data rights like the right of access, the right to rectification or erasure of personal data, or the right to complain when the rights are infringed or when there is data bias must be in place (Torkzadehmahani et al., 2022).

An example of this approach is the creation of the European Union’s General Data Protection Regulation or GDPR, which not only sets high privacy standards but also requires organizations to give individuals understandable information on collecting and processing their data (Ryngaert & Taylor, 2020). When patients see that their concerns are being answered and their data is protected, healthcare stakeholders can involve them in the AI-driven change of the industry instead of turning them into mere objects affected by the technologies.

Conclusion

In conclusion, there is a vast potential and, at the same time, an enormous danger to using artificial intelligence in healthcare and patient data. However, in addition to the old-fashioned legal and ethical framework, looking for a proactive and innovative solution is essential – the patient-centered development of these technologies. Thus, by involving the patients in designing the healthcare AI systems, integrating privacy-preserving designs into the systems, and fostering openness and responsibility, we can develop tools consistent with patients’ privacy preferences and rights. This approach not only addresses the dangers of surveillance and data abuse but also helps build trust, involvement, and cooperation between patients and all the related stakeholders in the healthcare field.

References

Banabilah, S., Aloqaily, M., Alsayed, E., Malik, N., & Jararweh, Y. (2022). Federated learning review: Fundamentals, enabling technologies, and future applications. Information processing & management, 59(6), 103061. https://doi.org/10.1016/j.ipm.2022.103061

Choudhury, A. (2022). Factors influencing clinicians’ willingness to use an AI-based clinical decision support system. Frontiers in digital health, 4, 920662. https://doi.org/10.3389/fdgth.2022.920662

Cresswell, K., Rigby, M., Magrabi, F., Scott, P., Brender, J., Craven, C. K., … & Williams, R. (2023). The need to strengthen the evaluation of the impact of Artificial Intelligence-based decision support systems on healthcare provision. Health policy, 136, 104889. https://doi.org/10.1016/j.healthpol.2023.104889

NVIDIA (2024). NVIDIA Clara. AI-Powered Solutions for Healthcare. https://www.nvidia.com/en-us/clara/

Peters, D., Vold, K., Robinson, D., & Calvo, R. A. (2020). Responsible AI—two frameworks for ethical design practice. IEEE Transactions on Technology and Society, 1(1), 34-47. https://doi.org/10.1109/TTS.2020.297499

Rafiq, F., Awan, M. J., Yasin, A., Nobanee, H., Zain, A. M., & Bahaj, S. A. (2022). Privacy prevention of big data applications: A systematic literature review. SAGE Open, 12(2), 21582440221096445. https://doi.org/10.1177/21582440221096445

Ryngaert, C., & Taylor, M. (2020). The GDPR as global data protection regulation?. https://doi.org/10.1017/aju.2019.80

Skaria, R., Parvaneh, S., Zhou, S., Kim, J., Wanjiru, S., Devers, G., … & Khalpey, Z. (2020). Path to precision: prevention of post-operative atrial fibrillation. Journal of thoracic disease, 12(5), 2735. https://doi.org/10.21037%2Fjtd-19-3875

Torkzadehmahani, R., Nasirigerdeh, R., Blumenthal, D. B., Kacprowski, T., List, M., Matschinske, J., … & Baumbach, J. (2022). Privacy-preserving artificial intelligence techniques in biomedicine. Methods of information in medicine, 61(S 01), e12-e27. https://doi.org/10.1055/s-0041-1740630

a:link {text-decoration: none;}a:visited {text-decoration: none;
}a:hover {text-decoration: underline;} a:active {text-decoration: underline;}

We’ll write everything from scratch

---

In ENGL148, you will write a full Argument Research Essay. You will work each week to build this essay using tools to help you develop your topics and ideas, create a full essay plan, and draft and revise an essay.  Briefly, the essay will require:

• 4-5 pages of writing plus a properly formatted APA cover sheet and reference page
• 4 scholarly resources cited and incorporated into the essay using proper APA
• additional APA formatting and writing requirements are detailed in the Week 6 assignment area

The Argument Research Essay Steps by Week

Starting in Week 2, you will begin preparing to write your 4-5-page Argument Research Essay.

• In Week 2, you will study the argument pattern of writing, how to use prewriting, and how to research scholarly sources and cite them. You will apply all of this knowledge toward completing a proposal worksheet for your Argument Research Essay. You may also use the Cybercafe discussion to work with your classmates on your essay.
• In Week 3, you will study how to outline an argument, how to use the Toulmin and Aristotle argument models, how to properly summarize, paraphrase, and quote, and how to create an annotated bibliography. You will be participating in your first Writing Workshop, receiving and giving peer reviews over the Argument Research Proposal Worksheet.
• In Week 4, you will study how the essay is structured, how to write strong and balanced arguments, and how to write strong introductions and conclusions and format a simple essay in APA. For your assignment, you will draft the introduction, one body, and a conclusion paragraph for your Argument Research Essay.  You will be able to use the Cybercafe discussion to work with your classmates on your essay.
• In Week 5, you will participate in a Writing Workshop with peers again, getting your introduction, one body, and conclusion paragraphs in the best shape possible to pick up revision tools you might use for the other paragraphs when later doing your full essay revisions.
• In Week 6, you will submit the first draft of your Argument Research Essay to your instructor for grading. The topics you study in Week 6 include a checklist for a full argument research essay and argument genres. You will again have the opportunity to use the Cybercafe discussion to work with your classmates on your essay.
• In Week 7, you will study peer reviewing and writing workshops. You will then be revising editing and reflecting on your writing process and activities in your final Writing Workshop.
• In Week 8, you will study revisions from global to granular and how argument and research can be used in everyday life. You will make final revisions and edits to your Argument Research Essay and turn in your final draft.

Last Completed Projects