TEFCA Synthetic Patient Cyber-Testing
DOI:
https://doi.org/10.63282/3050-9246.IJETCSIT-V6I2P115Keywords:
TEFCA, Synthetic Patients, Diffusion Models, EMR Deep Fakes, Healthcare Cybersecurity, Patient Matching, AI-Driven Fraud, Cyber-Testing FrameworkAbstract
The Trusted Exchange Framework and Common Agreement (TEFCA) help healthcare networks in the United States interoperate with each other securely and on a large scale. It is essentially a crucial infrastructure for national health information exchange. Many people are looking at synthetic data generation as a way to train models in a privacy, preserving manner, but there is still no research on how to use synthetic data for offensive cybersecurity stress, testing of healthcare infrastructure. Here we propose a new synthetic patient cyber, testing framework based on generative diffusion models that can produce high, quality artificial patient identities and deepfake electronic medical records (EMRs) for adversarial testing of interoperability systems. The framework evaluates systematically the vulnerabilities in patient, matching algorithms, identity resolution systems, and fraud, detection pipelines by injecting diffusion, generated synthetic identities into a simulated TEFCA exchange environment. We provide a detailed cyber, injection model, an AI, driven identity collision engine, and a multi, metric cybersecurity evaluation framework through which the resilience of systems against AI, generated insurance fraud, record manipulation, and identity spoofing attacks can be quantified. Running experimental simulations has revealed that diffusion, based synthetic identities are capable of uncovering the hidden loopholes in probabilistic and AI, based matching systems, thus facilitating the development of preemptive defense models and resilience benchmarking. This study presents a paradigm shift whereby synthetic data is not just a privacy tool but a strategic cybersecurity defense weapon that ensures the national healthcare interoperability infrastructures' readiness for future AI, driven cyber threats.
Downloads
References
[1] Johnson, A. E. W., Pollard, T. J., Shen, L., Lehman, L. H., Feng, M., Ghassemi, M., ... & Mark, R. G. (2016). MIMIC-III, a freely accessible critical care database. Scientific Data, 3, 160035.
[2] Choi, E., Bahadori, M. T., Song, L., Stewart, W. F., & Sun, J. (2016). RETAIN: An interpretable predictive model for healthcare using reverse time attention mechanism. Advances in Neural Information Processing Systems, 29.
[3] Che, Z., Purushotham, S., Cho, K., Sontag, D., & Liu, Y. (2017). Recurrent neural networks for multivariate time series with missing values. Scientific Reports, 8, 6085.
[4] Rajkomar, A., Oren, E., Chen, K., Dai, A. M., Hajaj, N., Hardt, M., ... & Dean, J. (2017). Scalable and accurate deep learning with electronic health records. NPJ Digital Medicine, 1, 18.
[5] Xu, L., Skoularidou, M., Cuesta-Infante, A., & Veeramachaneni, K. (2018). Modeling tabular data using conditional GAN. Advances in Neural Information Processing Systems, 31.
[6] Beaulieu-Jones, B. K., & Greene, C. S. (2018). Privacy-preserving generative deep neural networks support clinical data sharing. Circulation: Cardiovascular Quality and Outcomes, 11(9).
[7] Frid-Adar, M., Klang, E., Amitai, M., Goldberger, J., & Greenspan, H. (2019). Synthetic data augmentation using GAN for improved liver lesion classification. IEEE Transactions on Medical Imaging, 38(3), 677–685.
[8] Park, N., Mohammadi, M., Gorde, K., Jajodia, S., Park, H., & Kim, Y. (2019). Data synthesis based on generative adversarial networks. Proceedings of the VLDB Endowment, 13(10), 1857–1871.
[9] Rasmy, L., Xiang, Y., Xie, Z., Tao, C., & Zhi, D. (2020). Med-BERT: Pretrained contextualized embeddings on large-scale structured electronic health records. NPJ Digital Medicine, 4, 86.
[10] Li, Y., Rao, S., Solares, J. R. A., Hassaine, A., Ramakrishnan, R., Canoy, D., ... & Salimi-Khorshidi, G. (2020). BEHRT: Transformer for electronic health records. Scientific Reports, 10, 7155.
[11] Baowaly, M. K., Lin, C. C., Liu, C. L., & Chen, K. T. (2021). Synthesizing electronic health records using improved GANs. Journal of the American Medical Informatics Association, 28(1), 92–102.
[12] Van der Schaar, M., Alaa, A., Floto, A., Gimson, A., Scholtes, S., Wood, A., ... & Ercole, A. (2021). How artificial intelligence and machine learning can help healthcare systems respond to COVID-19. Machine Learning, 110, 1–14.
[13] Kotelnikov, A., Baranchuk, D., Rubachev, I., & Babenko, A. (2022). TabDDPM: Modelling tabular data with diffusion models. International Conference on Machine Learning (ICML).
[14] Tomašev, N., et al. (2022). Use of deep learning to develop continuous-risk models for adverse events in hospital. Nature, 572, 116–119.
[15] Palanisamy, P., Urooj, S., Arunachalam, R., & Lay-Ekuakille, A. (2023). A novel prognostic model using chaotic CNN with hybridized spoofing for enhancing diagnostic accuracy in epileptic seizure prediction. Diagnostics, 13(21), 3382.
[16] Preethi, P., Vasudevan, I., Saravanan, S., Prakash, R. K., & Devendhiran, A. (2023, December). Leveraging network vulnerability detection using improved import vector machine and Cuckoo search based Grey Wolf Optimizer. In 2023 1st International Conference on Optimization Techniques for Learning (ICOTL) (pp. 1-7). IEEE.
[17] Chen, X., Wang, Y., Li, Q., & Sun, J. (2024). Privacy-aware diffusion models for synthetic healthcare data generation. IEEE Journal of Biomedical and Health Informatics.
[18] Patel, R., Singh, A., & Kumar, S. (2024). Federated generative models for privacy-preserving medical data synthesis. Artificial Intelligence in Medicine, 149, 102749.
