EUVIMEDEUVIMED entdecken
Crossref

Medsecai: A Federated Multi-Layer Artificial Intelligence Framework For Cyber Security Threat Detection In Electronic Health Record Systems

Arunkumar Palanichamy, Sivakumar Dhandapani

International Journal of Drug Delivery Technology · 11.05.2026 · 16 (26s)

DOI: 10.25258/ijddt.16.26s.91 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Drug Delivery Technology, online erschienen am 11.05.2026. Verfasst von Arunkumar Palanichamy und 1 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Electronic health record (ehr) systems aggregate longitudinal patient data at a scale that renders them the primary target of healthcare-sector cyberattacks. Conventional signature-based intrusion detection systems are unable to identify novel, polymorphic threats that adapt faster than static rule-bases can be updated. This study proposes medsecai, a four-layer ai/ml security framework designed to address this limitation. The architecture integrates: (i) a bidirectional long short-term memory (bi-lstm) autoencoder with a dynamic sliding-percentile threshold for session-level anomaly detection; (ii) a gradient boosting (xgboost) multi-class threat classifier that ingests lstm reconstruction error as an explicit feature — a two-stage coupling not previously demonstrated in ehr security literature; (iii) a reinforcement learning (rl) policy engine based on proximal policy optimisation (ppo) for adaptive access control; and (iv) a federated learning aggregation layer supporting privacy-preserving collaborative training across multiple hospital clients under differential privacy (ε = 0.5). Evaluated on the cicids-2017 benchmark comprising 2.8 million network flow records, the proposed framework achieves 98.7% intrusion detection accuracy, a false positive rate of 0.9%, and a 428× reduction in inter-institutional data transmission compared with a centralised equivalent. Ablation against four baselines confirms that each architectural component contributes measurably to overall performance. The compliance implications of the federated design are examined under hipaa, gdpr, and india's digital personal data protection (dpdp) act 2023.

Zitieren

APA

Palanichamy, A., & Dhandapani, S. (2026). Medsecai: A Federated Multi-Layer Artificial Intelligence Framework For Cyber Security Threat Detection In Electronic Health Record Systems. International Journal of Drug Delivery Technology, 16(26s). https://doi.org/10.25258/ijddt.16.26s.91

Vancouver

Palanichamy A, Dhandapani S. Medsecai: A Federated Multi-Layer Artificial Intelligence Framework For Cyber Security Threat Detection In Electronic Health Record Systems. International Journal of Drug Delivery Technology. 2026;16(26s). doi:10.25258/ijddt.16.26s.91

BibTeX
@article{Palanichamy2026,
  author = {Palanichamy, Arunkumar and Dhandapani, Sivakumar},
  title = {Medsecai: A Federated Multi-Layer Artificial Intelligence Framework For Cyber Security Threat Detection In Electronic Health Record Systems},
  journal = {International Journal of Drug Delivery Technology},
  year = {2026},
  volume = {16},
  number = {26s},
  doi = {10.25258/ijddt.16.26s.91}
}
Kostenlos starten →