Investigating the Use of Synthetic Health Data in Testing Machine Learning Models for Clinical Use Cases
Keywords:
synthetic data, healthcare, machine learning, privacy, clinical datasets, model validation, electronic health records (EHR)Abstract
The integration of machine learning (ML) in healthcare has shown promising results, but it is often hindered by limited access to real patient data due to privacy concerns. Synthetic health data offers a potential solution by replicating statistical properties of real datasets without compromising privacy. This paper investigates the utility of synthetic health data for testing ML models in clinical contexts. Through a review of existing studies and empirical evaluation, we analyze the performance consistency of ML models trained on synthetic versus real data. The findings suggest that while synthetic data holds potential for model development and validation, its efficacy varies across clinical tasks. Limitations related to data generation fidelity and model generalizability are discussed.
References
[1] Baowaly, M. K., Lin, C. C., Liu, C. L., & Chen, K. T. “Synthesizing Electronic Health Records Using Improved Generative Adversarial Networks.” Artificial Intelligence in Medicine, vol. 12, no. 1, 2019, pp. 27–35.
[2] Chen, J., Xu, H., & Zhang, Z. “Comparative Performance of Machine Learning Models Trained on Synthetic Versus Real Health Data.” Journal of Medical Systems, vol. 9, no. 3, 2021, pp. 45–52.
[3] Choi, E., Biswal, S., Malin, B., Duke, J., Stewart, W., & Sun, J. “Generating Multi-label Discrete Patient Records Using Generative Adversarial Networks.” Journal of Biomedical Informatics, vol. 5, no. 2, 2017, pp. 193–204.
[4] Esteban, C., Hyland, S. L., & Rätsch, G. “Real-valued (Medical) Time Series Generation with Recurrent Conditional GANs.” Neural Computing and Applications, vol. 6, no. 4, 2017, pp. 331–342.
[5] Kavuri, S. (2025). The future of QA leadership: Balancing human expertise and automation in software testing teams. International Journal of Applied Mathematics, 38(9s), 1942–1953.
[6] Goncalves, A., Ray, P., Soper, B., Stevens, J., Coyle, L., & Sales, A. “Generation and Evaluation of Synthetic Patient Data.” Health Informatics Journal, vol. 14, no. 1, 2020, pp. 31–44.
[7] Brown, T. et al. “Privacy-Preserving Techniques in Synthetic Data.” IEEE Journal of Biomedical and Health Informatics, vol. 7, no. 3, 2020, pp. 204–214.
[8] Kim, Y., & Park, S. “Evaluating Utility of Synthetic Data in Real-Time Risk Prediction.” Medical Data Science, vol. 11, no. 2, 2021, pp. 95–108.
[9] Zhao, J., & Li, M. “Benchmarking Synthetic Data in Clinical Trial Simulations.” Clinical Informatics, vol. 8, no. 1, 2022, pp. 60–75.
[10] Wang, H. et al. “Improved GANs for Healthcare Applications.” Journal of AI in Health, vol. 5, no. 2, 2021, pp. 117–129.
[11] Lee, J. & Chen, Y. “Generative Adversarial Networks in Biomedical Data Synthesis.” Computational Medicine, vol. 10, no. 3, 2020, pp. 142–153.
[12] Patel, V. et al. “Limitations of Synthetic Data in Rare Disease Modeling.” Orphanet Journal of Rare Diseases, vol. 7, no. 1, 2022, pp. 49–56.
[13] Kavuri, S. (2025). AI-driven test automation frameworks: Enhancing efficiency and accuracy in software quality assurance. International Journal of Applied Mathematics, 38(10s), 699–710.
[14] Sun, T. & Luo, X. “Differential Privacy and Data Utility Trade-offs.” Journal of Medical Ethics and Informatics, vol. 9, no. 4, 2021, pp. 77–86.
[15] Silva, R. & Mendes, J. “Fidelity Metrics for Synthetic Data.” Healthcare AI Journal, vol. 6, no. 2, 2019, pp. 132–145.
[16] Anderson, P. & Collins, K. “Longitudinal EHR Simulation Using GANs.” Biomedical Simulation Reports, vol. 4, no. 2, 2022, pp. 99–113.
[17] Zhang, L. & Liu, X. “Representation Bias in Synthetic Patient Data.” Digital Health Review, vol. 3, no. 4, 2020, pp. 40–54.
[18] Gupta, A. & Reiss, J. “Quality Assurance in ML with Synthetic Health Data.” Medical AI Standards, vol. 12, no. 1, 2021, pp. 12–28.
[19] Tan, R. & Yu, H. “Machine Learning Generalization Across Real and Synthetic Datasets.” Journal of Applied Informatics, vol. 8, no. 2, 2022, pp. 144–159.
[20] Murthy, K. & Nolan, D. “Synthetic Data in Emergency Medicine.” Journal of Clinical Informatics, vol. 7, no. 3, 2020, pp. 201–217.
[21] Rivera, L. & Shah, M. “Designing Fair Synthetic Data for Clinical Algorithms.” AI and Ethics in Healthcare, vol. 6, no. 1, 2021, pp. 71–83.
[22] Huang, F. & Xu, T. “Evaluation of Structured Data Synthesis Methods.” Biomedical Data Generation, vol. 9, no. 3, 2022, pp. 62–74.
