AI and big data analytics in health care: Opportunity or threat?
Authors: Devarati Bagchi1, Vidya Prabhu1, Divya Sussana Patil2
1. Department of Global Public Health Policy and Governance, Prasanna School of Public Health, Manipal Academy of Higher Education, Manipal, India
2. Centre for Evidence-informed Decision-making, Prasanna School of Public Health, Manipal Academy of Higher Education, Manipal, India
Introduction
The health care sector is experiencing a profound transformation, as artificial intelligence (AI) and big data analytics redefine how we understand, diagnose, and treat illnesses. As global challenges—from pandemics to climate change—intensify, these technologies have the potential to support a more equitable, effective, and sustainable health care system. This blog delves into how new technologies, under Pillar 2 of communication innovation, are shaping the evidence ecosystem. We ask the question: Are AI and big data in health care an opportunity for progress, or are they a threat?
Background
AI is the modeling of human intellect in computers, which can learn from and adjust to new information. AI is utilized in health care for personalized medication, therapy recommendations, and diagnostic algorithms. Processing huge volumes of medical data to identify patterns, trends, and insights is known as big data analytics. Big data and AI together have the potential to transform health care by making sense of the enormous amount of statistics produced on a daily basis in clinical settings. Big data's application in predictive analytics and AI's ability to customize health care are essential for contemporary medical practice.
The Problem
Big data analytics and AI provide many benefits, but there are also drawbacks. Given how sensitive health care data is, privacy issues are quite real. Data bias is another problem that might result in inappropriate therapies or incorrect diagnosis Vulnerable communities can be disproportionately affected by skewed data, and AI algorithms are only as good as the data they are trained on. This problem is made worse by the absence of standardization in data collection procedures between health care systems. Furthermore, AI may not perform well in high-stakes settings such as emergency rooms, where human intuition and unforeseen circumstances still matter greatly. Getting these technologies integrated into current health care systems is one of the main problems. It is challenging to gather and thoroughly evaluate health care data since it is frequently dispersed throughout several platforms and organizations. In addition, the health care industry is subject to stringent privacy and security regulations. Ethical concerns arise when handling sensitive patient data, particularly in the case of AI algorithms that need to access enormous volumes of personal data. One major obstacle to the broad application of AI and big data analytics in health care is the possibility of data breaches and information misuse. Health care workers also lack sufficient understanding of the application of AI and big data techniques.
Search for Answers
Decentralized learning allows machine models to be trained on disparate sources of data without sacrificing privacy. For example, hospitals can work together to train models based on different datasets that forecast patients' risk of developing illnesses such as diabetes, as long as the data are kept local to each hospital. By using a variety of datasets without jeopardizing patient privacy, federated learning can improve the dependability of AI algorithms in health care organizations. Standardizing the use of AI in health care through ethical guidelines aims to reduce prejudice and guarantee fair treatment for all patient groups. Personalized medicine is where AI can use genetic data analysis to customize medications for specific people, increasing effectiveness and lowering side-effects. This strategy is especially helpful in oncology, where patient reactions to treatment might differ greatly from one another.
Potential Global Impact
Big data and AI have the potential to revolutionize health care globally. AI-driven diagnostic technologies could compensate for a shortage of trained health care workers in nations with inadequate health care infrastructure. The use of AI systems in mammography interpretation led to a decrease in false positives and negatives, respectively. These nations may benefit from big data analytics to better allocate resources and carry out public health programs. On the other hand, unequal access to these technologies carries some hazards that could exacerbate worldwide health care disparities. The effect might be more complex in industrialized countries. Although AI may enhance patient care, it may also reduce the need for some health care jobs, which could result in worker displacement. The concern of relying too much on AI could result in diagnostic error.
Challenges and Obstacles
There are various obstacles in the way of implementing AI and big data analytics in the health care industry. Regulatory systems are still catching up to technological breakthroughs. The lack of international regulatory standards results in a disjointed environment that makes cross-border cooperation in health care research difficult. In addition, health care organizations need to make infrastructure investments to support big data and AI activities, which might be unaffordable for some systems. Another big worry is security. Cyberattacks on health care systems could compromise patient information and interfere with vital functions. Fraudsters are targeting health care businesses more frequently, with big data platforms being particularly susceptible to breaches.
Lessons Learned
Notwithstanding these difficulties, the application of big data and AI in health care has taught us many important lessons. The idea that technology should enhance human skill rather than replace it is one of the main lessons learned. Health care systems that have effectively incorporated AI have done so by utilizing it to support physicians rather than replace them. The significance of transparent algorithms is a further lesson. Research has brought attention to the need for lucid, comprehensible AI models to guarantee that physicians can rely on and validate AI-driven decisions.
Next Steps
A multifaceted strategy is required to fully harness the potential of AI and big data analytics in health care. First and foremost, funding for health care IT infrastructure needs to be a top priority. This will not only increase AI's efficacy but also guard against intrusions into private medical data. Second, to guarantee the ethical and secure application of AI in health care, regulatory organizations must establish uniform frameworks. Global efforts to unify legislation could make cross-border sharing of AI models and medical data easier. In addition to knowing how to use AI technologies, clinicians must be able to recognize when these tools can be incorrect. Continuous professional development in AI-related disciplines will guarantee that health care personnel are prepared to incorporate new technologies into patient care.
Key Messages
Health care delivery can be greatly improved by AI and big data analytics, but there are drawbacks as well, including concerns about bias, security, and privacy of data. To securely integrate these technologies into health care institutions, uniform legislation, enhanced IT infrastructure, and ethical standards are necessary. AI ought to be viewed as an enhancement to human knowledge, not as a substitute for it. To fully utilize AI and big data in health care and overcome the obstacles they present, ongoing research and worldwide cooperation are necessary.
References
Alowais, S.A., Alghamdi, S.S., Alsuhebany, N. et al. (2023). Revolutionizing health care: The role of artificial intelligence in clinical practice. BMC Medical Education, 23, 689. https://doi.org/10.1186/s12909-023-04698-z
Batko, K., & Ślęzak, A. (2022). The use of big data analytics in healthcare. Journal of Big Data, 9(1), 3. https://doi.org/10.1186/s40537-021-00553-4
Chin, M. H., Afsar-Manesh, N., Bierman, A. S., et al. (2023). Guiding principles to address the impact of algorithm bias on racial and ethnic disparities in health and health care. JAMA Network Open, 6(12), e2345050. doi:10.1001/jamanetworkopen.2023.45050
Lee, E., & Viswanath, K. (2020). Big data in context: Addressing the twin perils of data absenteeism and chauvinism in the context of health disparities research. Journal of Medical Internet Research, 22(1): e16377.
Panch, T., Mattie, H., & Atun, R. (2019). Artificial intelligence and algorithmic bias: Implications for health systems. Journal of Global Health, 9(2), 010318. https://doi.org/10.7189/jogh.09.020318
Rahman, A., Hossain, M. S., Muhammad, G., Kundu, D., Debnath, T., Rahman, M., Khan, M. S. I., Tiwari, P., & Band, S. S. (2022). Federated learning-based AI approaches in smart healthcare: Concepts, taxonomies, challenges and open issues. Cluster Computing, 1–41. https://doi.org/10.1007/s10586-022-03658-4
Rieke, N., Hancox, J., Li, W., Milletarì, F., Roth, H. R., Albarqouni, S., Bakas, S., Galtier, M. N., Landman, B. A., Maier-Hein, K., Ourselin, S., Sheller, M., Summers, R. M., Trask, A., Xu, D., Baust, M., & Cardoso, M. J. (2020). The future of digital health with federated learning. NPJ Digital Medicine, 3, 119. https://doi.org/10.1038/s41746-020-00323-1
Teo ZL, Jin L, Li S, Miao D, Zhang X, Ng WY, Tan TF, Lee DM, Chua KJ, Heng J, Liu Y, Goh RSM, Ting DSW. (2024). Federated machine learning in health care: A systematic review on clinical applications and technical architecture. Cell Reports Medicine, 5(2), 101419. doi: 10.1016/j.xcrm.2024.101419.
To link to this article - DOI: https://doi.org/10.70253/VNHR9587
Disclaimer
The views expressed in this World EBHC Day Blog, as well as any errors or omissions, are the sole responsibility of the author and do not represent the views of the World EBHC Day Steering Committee, Official Partners or Sponsors; nor does it imply endorsement by the aforementioned parties.
