Unsupervised FHIR Stream Anomaly Detection
Keywords:
Adaptive Control ,Autonomous Vehicle , Dynamic Environment , Model-Free Reinforcement Learning , Safety , Traffic Efficiency.Abstract
Reports of serious security breaches involving medical data have attracted increasing public attention. These incidents emphasize the importance of protecting health information and re-establishing patient trust in information systems that support healthcare delivery. Guaranteeing the confidentiality and integrity of healthcare data is challenging, given the intrinsic requirements dictated by openness, interoperability, accessibility, and availability. In addition to the overall emergence of disorder, active, unattended FHIR data streams produced by thousands of devices, systems, and services are particularly prone to contain high-level attributes and momentary deviations that fall outside accepted monitoring thresholds or general patterns. Such anomalies may indicate dangerous problems, progress toward disastrous events, or even the introduction of malware. However, it is hardly possible to use supervised techniques to define the rich set of expected patterns that characterize or govern real-time healthcare data streams.
Unsupervised learning methods are presented with an objective, evidence-based analysis and formal structure. Emphasis is placed on the application of these approaches to real-time anomaly detection within unattended and open healthcare data streams that follow the Fast Healthcare Interoperability Resources (FHIR) standard. The analysis offers a heuristic indication of the methods that can best exploit the availability and characteristics of real-time healthcare data streams. The methods are systematically examined against qualifying criteria and success factors for their application in real-time healthcare environments
References
1. Bossenko, I., Randmaa, R., Piho, G., & Ross, P. (2024). Interoperability of health data using FHIR Mapping Language: Transforming HL7 CDA to FHIR with reusable visual components. Frontiers in Digital Health, 6, 1480600.
2. Tabari, P., Costagliola, G., De Rosa, M., & Boeker, M. (2024). State-of-the-art Fast Healthcare Interoperability Resources (FHIR)-based data model and structure implementations: Systematic scoping review. JMIR Medical Informatics, 12, e58445.
3. Namli, T., Sınacı, A. A., Gönül, S., Herguido, C. R., Garcia-Canadilla, P., Muñoz, A. M., Esteve, A. V., & Ertürkmen, G. B. L. (2024). A scalable and transparent data pipeline for AI-enabled health data ecosystems. Frontiers in Medicine, 11, 1393123.
4. Kirilov, N. (2024). Comparison of the performance of the Fast Healthcare Interoperability Resources (FHIR) subscription channels. Studies in Health Technology and Informatics, 316, 131–135.
5. Klopfenstein, S. A. I., Sass, J., Vorisek, C. N., Jorczik, F., Schmidt, C. O., Löbe, M., Golebiewski, M., Abaza, H., & Thun, S. (2024). Bringing communities together: Mapping the Investigation-Study-Assay-Model (ISA) to Fast Healthcare Interoperability Resources (FHIR). Studies in Health Technology and Informatics, 310, 856–860.
6. Khan, M. M., & Alkhathami, M. (2024). Anomaly detection in IoT-based healthcare: Machine learning for enhanced security. Scientific Reports, 14, 5872.
7. Muspratt, R., & Mammadov, M. (2024). Anomaly detection with sub-extreme values: Health provider billing. Data Science and Engineering, 9(1), 62–72.
8. El Jabri, C., Frappier, M., & Tardif, P.-M. (2024). ASTD patterns for integrated continuous anomaly detection in data logs. arXiv Preprint arXiv:2411.07272.
9. Maeda, S., Gu, C., Yu, J., Tokai, S., Gao, S., & Zhang, C. (2024). Frequency-guided multi-level human action anomaly detection with normalizing flows. arXiv Preprint arXiv:2404.17381.
10. Ahmed, M., Mahmood, A. N., & Hu, J. (2024). A survey of network anomaly detection techniques in machine learning. Journal of Network and Computer Applications, 236, 103868.
11. Pang, G., Shen, C., Cao, L., & Van Den Hengel, A. (2024). Deep learning for anomaly detection: A review. ACM Computing Surveys, 57(2), 1–38.
12. Ruff, L., Kauffmann, J. R., Vandermeulen, R. A., Montavon, G., Samek, W., Kloft, M., Dietterich, T. G., & Müller, K.-R. (2024). A unifying review of deep and shallow anomaly detection. Proceedings of the IEEE, 112(3), 354–387.
13. Chalapathy, R., & Chawla, S. (2024). Deep learning for anomaly detection in healthcare systems: Recent advances and future directions. Information Sciences, 669, 120538.
14. Li, Y., Mahapatra, D., & Ge, Z. (2024). Self-supervised anomaly detection for healthcare data streams. Pattern Recognition, 151, 110387.
15. Zhou, C., Paffenroth, R. C., & Aggarwal, C. C. (2024). Outlier detection in streaming data: A comprehensive review. Data Mining and Knowledge Discovery, 38(2), 625–668.
16. Yang, C., Chen, Y., & Wang, J. (2024). Real-time anomaly detection in healthcare event streams using unsupervised representation learning. Expert Systems with Applications, 250, 123912.
17. Zhang, H., Liu, X., & Sun, J. (2024). Transformer-based unsupervised anomaly detection for temporal healthcare records. Knowledge-Based Systems, 298, 111905.
18. Kumar, A., Singh, P., & Verma, S. (2024). Federated anomaly detection for privacy-preserving healthcare analytics. Future Generation Computer Systems, 157, 358–372.
19. Garcia, S., Ramirez, D., & Lopez, J. (2024). Stream analytics and anomaly detection for interoperable healthcare platforms. IEEE Access, 12, 94712–94729.
20. Chen, T., Wu, H., & Zhao, Y. (2024). Unsupervised learning approaches for detecting anomalies in electronic health record streams. Artificial Intelligence in Medicine, 148, 102746.
