EUVIMEDEUVIMED entdecken
Crossref

Federated and Privacy-Preserving Edge-TinyML Framework for Secure Cardiac Drug Response Monitoring Using Embedded Hardware– Software Co-Design

Anupama P. Patil

International Journal of Drug Delivery Technology · 08.06.2026 · 16 (50s)

DOI: 10.25258/ijddt.16.50s.52 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Drug Delivery Technology, online erschienen am 08.06.2026. Verfasst von Anupama P. Patil.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The rapid advancement of intelligent healthcare technologies and wearable biomedical devices has significantly increased the demand for secure and real-time cardiac drug response monitoring systems capable of operating efficiently in decentralized healthcare environments. Conventional cloud-centric healthcare architectures used for continuous electrocardiogram monitoring and physiological signal analysis often suffer from major limitations including high communication latency, excessive bandwidth utilization, privacy vulnerabilities, centralized data exposure, and increased energy consumption. Although edge-based TinyML frameworks have recently emerged as promising solutions for lowlatency physiological inference on resource-constrained embedded devices, existing approaches remain limited in their ability to support collaborative learning across distributed patients while preserving the confidentiality of sensitive medical information. Most current TinyML cardiac monitoring systems rely on isolated local inference models that lack adaptive global intelligence and fail to generalize effectively across heterogeneous patient conditions and drug response variations. Furthermore, existing edge-learning approaches provide insufficient protection against model inversion attacks, parameter leakage, and communication interception threats. In response to these challenges, this research proposes a federated and privacy-preserving Edge-TinyML framework for secure cardiac drug response monitoring using embedded hardware–software co-design methodologies. The proposed framework integrates electrocardiogram signal acquisition, heart rate variability analysis, lightweight machine learning inference, federated learning optimization, secure aggregation protocols, and differential privacy mechanisms within a decentralized edge intelligence architecture. Embedded TinyML models including lightweight convolutional neural networks, support vector machine variants, and decision-tree-based classifiers are deployed on microcontroller-enabled edge devices for localized physiological analysis and cardiac anomaly prediction. Federated learning enables multiple distributed devices to collaboratively train a shared global model without transmitting raw patient data to centralized servers, thereby significantly improving privacy preservation and reducing data exposure risks. Encrypted model synchronization and differential privacy noise injection techniques are incorporated to strengthen security against inference attacks and unauthorized access during federated communication processes. The framework additionally employs communication-efficient federated averaging, quantization, pruning, and heterogeneity-aware aggregation strategies to optimize computational efficiency, minimize bandwidth overhead, reduce energy consumption, and support adaptive learning under resource-constrained embedded environments. Standard biomedical datasets including the MIT-BIH Arrhythmia Database and PhysioNet ECG datasets are utilized to evaluate the effectiveness of the proposed framework under varying cardiac monitoring scenarios involving patient-specific drug response patterns associated with beta-blockers and anti-arrhythmic medications. Experimental evaluation demonstrates that the proposed system achieves high prediction accuracy, reduced inference latency, lower communication costs, enhanced energy efficiency, and improved privacy protection compared with conventional centralized healthcare monitoring architectures. The research further establishes that integrating federated learning with TinyML-enabled edge intelligence and embedded hardware–software co-design provides a scalable, secure, and privacy-aware solution for next-generation wearable and implantable cardiac monitoring systems. The proposed framework contributes significantly toward the advancement of decentralized intelligent healthcare infrastructures capable of supporting personalized medicine, secure biomedical analytics, and adaptive real-time physiological monitoring within modern digital healthcare ecosystems

Zitieren

APA

Patil, A. P. (2026). Federated and Privacy-Preserving Edge-TinyML Framework for Secure Cardiac Drug Response Monitoring Using Embedded Hardware– Software Co-Design. International Journal of Drug Delivery Technology, 16(50s). https://doi.org/10.25258/ijddt.16.50s.52

Vancouver

Patil AP. Federated and Privacy-Preserving Edge-TinyML Framework for Secure Cardiac Drug Response Monitoring Using Embedded Hardware– Software Co-Design. International Journal of Drug Delivery Technology. 2026;16(50s). doi:10.25258/ijddt.16.50s.52

BibTeX
@article{Patil2026,
  author = {Patil, Anupama P.},
  title = {Federated and Privacy-Preserving Edge-TinyML Framework for Secure Cardiac Drug Response Monitoring Using Embedded Hardware– Software Co-Design},
  journal = {International Journal of Drug Delivery Technology},
  year = {2026},
  volume = {16},
  number = {50s},
  doi = {10.25258/ijddt.16.50s.52}
}
Kostenlos starten →