Trustworthy Intelligence: Split Learning–Embedded Large Language Models for Smart IoT Healthcare Systems

Mahbuba Ferdowsi; Nour Moustafa; Marwa Keshk; Benjamin Turnbull

doi:10.3390/electronics15071519

journal article Open Access Apr 04, 2026

Trustworthy Intelligence: Split Learning–Embedded Large Language Models for Smart IoT Healthcare Systems

Mahbuba Ferdowsi

Nour Moustafa

Marwa Keshk

Benjamin Turnbull

Electronics Vol. 15 No. 7 pp. 1519 · MDPI AG

View at Publisher Save 10.3390/electronics15071519

Abstract

The Internet of Things (IoT) plays an increasingly central role in healthcare by enabling continuous patient monitoring, remote diagnosis, and data-driven clinical decision-making through interconnected medical devices and sensing infrastructures. Despite these advances, IoT healthcare systems remain constrained by persistent challenges related to data privacy, computational efficiency, scalability, and regulatory compliance. Federated learning (FL) reduces reliance on centralised data aggregation but remains vulnerable to inference-based privacy risks, while edge-oriented approaches are limited by device heterogeneity and restricted computational and energy resources; the deployment of large language models (LLMs) further exacerbates concerns surrounding privacy exposure, communication overhead, and practical feasibility. This study introduces Trustworthy Intelligence (TI) as a guiding framework for privacy-preserving distributed intelligence in IoT healthcare, explicitly integrating predictive performance, privacy protection, and deployment-oriented system design. Within this framework, split learning (SL) is examined as a core architectural mechanism and extended to support split-aware LLM integration across heterogeneous devices, supported by a structured taxonomy spanning architectural configurations, system adaptation strategies, and evaluation considerations. The study establishes a systematic mapping between SL design choices and representative healthcare scenarios, including wearable monitoring, multi-modal data fusion, clinical text analytics, and cross-institutional collaboration, and analyses key technical challenges such as activation-level privacy leakage, early-round vulnerability, reconstruction risks, and communication–computation trade-offs. An energy- and resource-aware adaptive cut layer selection strategy is outlined to support efficient deployment across devices with varying capabilities. A proof-of-concept experimental evaluation compares the proposed SL–LLM framework with centralised learning (CL), federated learning (FL), and conventional SL in terms of training latency, communication overhead, model accuracy, and privacy exposure under realistic IoT constraints, providing system-level evidence for the applicability of the TI framework in distributed healthcare environments and outlining directions for clinically viable and regulation-aligned IoT healthcare intelligence.

Topics

No keywords indexed for this article. Browse by subject →

References

144

[1]

Shafi, I., Din, S., Farooq, S., Díez, I.T., Breñosa, J., Espinosa, J.C.M., and Ashraf, I. (2024). Design and development of patient health tracking, monitoring and big data storage using Internet of Things and real time cloud computing. PLoS ONE, 19. 10.1371/journal.pone.0298582

[2]

Alsabah "A comprehensive review on key technologies toward smart healthcare systems based on IoT: Technical aspects, challenges and future directions" Artif. Intell. Rev. (2025) 10.1007/s10462-025-11342-3

[3]

World Health Organization (2021). Global Strategy on Digital Health 2020–2025, World Health Organization. Available online: https://www.who.int/publications/i/item/9789240020924.

[4]

Internet of Things in Healthcare Research: Trends, Innovations, Security Considerations, Challenges and Future Strategy

Attiq Ur Rehman, Songfeng Lu, Md Belal Bin Heyat et al.

International Journal of Intelligent Systems 2025 10.1155/int/8546245

[5]

(2013). Information Technology—Security Techniques—Information Security Management Systems—Requirements. Standard No. ISO/IEC 27001:2013. Available online: https://www.iso.org/standard/54534.html.

[6]

(2025). Health Informatics—Information Security Controls in Health Based on ISO/IEC 27002. Standard No. ISO 27799:2025. Available online: https://www.iso.org/standard/84647.html.

[7]

Xu "Federated Learning for Healthcare Informatics" J. Healthc. Inform. Res. (2021) 10.1007/s41666-020-00082-4

[8]

The future of digital health with federated learning

Nicola Rieke, Jonny Hancox, Wenqi Li et al.

npj Digital Medicine 2020 10.1038/s41746-020-00323-1

[9]

Advances and Open Problems in Federated Learning

Peter Kairouz, H. Brendan McMahan

Foundations and Trends® in Machine Learning 2021 10.1561/2200000083

[10]

Zhu, L., Liu, Z., and Han, S. (2019, January 8–14). Deep Leakage from Gradients. Proceedings of the Advances in Neural Information Processing Systems (NeurIPS), Vancouver, BC, Canada.

[11]

Edge Computing: Vision and Challenges

Weisong Shi, Jie Cao, Quan Zhang et al.

IEEE Internet of Things Journal 2016 10.1109/jiot.2016.2579198

[12]

Qiu "Applications of Auction and Mechanism Design in Edge Computing: A Survey" IEEE Trans. Cogn. Commun. Netw. (2022) 10.1109/tccn.2022.3147196

[13]

Toward expert-level medical question answering with large language models

Karan Singhal, Tao Tu, Juraj Gottweis et al.

Nature Medicine 2025 10.1038/s41591-024-03423-7

[14]

Large language models encode clinical knowledge

Karan Singhal, Shekoofeh Azizi, Tao Tu et al.

Nature 2023 10.1038/s41586-023-06291-2

[15]

The imperative for regulatory oversight of large language models (or generative AI) in healthcare

Bertalan Meskó, Eric J. Topol

npj Digital Medicine 2023 10.1038/s41746-023-00873-0

[16]

Organisation for Economic Co-operation and Development (2019). OECD Principles on Artificial Intelligence, Organisation for Economic Co-Operation and Development. Available online: https://oecd.ai/en/ai-principles.

[17]

(2023). Information Technology—Artificial Intelligence—Guidance on Risk Management. Standard No. ISO/IEC 23894:2023. Available online: https://www.iso.org/standard/77304.html.

[18]

Gupta "Distributed learning of deep neural network over multiple agents" J. Netw. Comput. Appl. (2018) 10.1016/j.jnca.2018.05.003

[19]

Zhang, M., Cao, J., Sahni, Y., Chen, X., and Jiang, S. (2024). Resource-efficient Parallel Split Learning in Heterogeneous Edge Computing. Proceedings of the 2024 International Conference on Computing, Networking and Communications (ICNC), Big Island, HI, USA, 19–22 February 2024, IEEE. 10.1109/icnc59896.2024.10556386

[20]

Duan, Q., Hu, S., Deng, R., and Lu, Z. (2022). Combined Federated and Split Learning in Edge Computing for Ubiquitous Intelligence in Internet of Things: State-of-the-Art and Future Directions. Sensors, 22. 10.3390/s22165983

[21]

Shiranthika "Decentralized Learning in Healthcare: A Review of Emerging Techniques" IEEE Access (2023) 10.1109/access.2023.3281832

[22]

Hu, Z., Zhou, T., Wu, B., Chen, C., and Wang, Y. (2025). A Review and Experimental Evaluation on Split Learning. Future Internet, 17. 10.3390/fi17020087

[23]

Hukkeri "A Comprehensive Survey on Split-Fed Learning: Methods, Innovations, and Future Directions" IEEE Access (2025) 10.1109/access.2025.3547641

[24]

Alhindi "Advancements and challenges in privacy-preserving split learning: Experimental findings and future directions" Int. J. Inf. Secur. (2025) 10.1007/s10207-025-01045-9

[25]

Pham, N.D., and Chilamkurti, N. (2024). Data Leakage Threats and Protection in Split Learning: A Survey. Proceedings of the 2023 International Conference on Intelligent Computing and Its Emerging Applications, Kaohsiung Taiwan, 14–16 December 2023, Association for Computing Machinery. 10.1145/3659154.3659189

[26]

Lekadir "FUTURE-AI: International consensus guideline for trustworthy and deployable artificial intelligence in healthcare" BMJ (2025) 10.1136/bmj-2024-081554

[27]

Bollineni "IoT for Next-Generation Smart Healthcare: A Comprehensive Survey" IEEE Internet Things J. (2025) 10.1109/jiot.2025.3570188

[28]

Shah "A systematic review of emerging trends of IoT in healthcare and IoMT frameworks" Discov. Public Health (2026) 10.1186/s12982-026-01347-w

[29]

Rancea, A., Anghel, I., and Cioara, T. (2024). Edge Computing in Healthcare: Innovations, Opportunities, and Challenges. Future Internet, 16. 10.3390/fi16090329

[30]

Baker "Artificial Intelligence of Things for Smarter Healthcare: A Survey of Advancements, Challenges, and Opportunities" IEEE Commun. Surv. Tutor. (2023) 10.1109/comst.2023.3256323

[31]

Adil "Healthcare Internet of Things: Security Threats, Challenges, and Future Research Directions" IEEE Internet Things J. (2024) 10.1109/jiot.2024.3360289

[32]

Abohany "A Comprehensive Survey on Impact of Applying Various Technologies on the Internet of Medical Things" Artif. Intell. Rev. (2025) 10.1007/s10462-024-11063-z

[33]

McMahan "Communication-Efficient Learning of Deep Networks from Decentralized Data" Proceedings of the 20th International Conference on Artificial Intelligence and Statistics (AISTATS), Ft. Lauderdale, FL, USA, 20–22 April 2017 (2017)

[34]

Hartmann "FedPref: Federated Learning Across Heterogeneous Multi-objective Preferences" ACM Trans. Model. Perform. Eval. Comput. Syst. (2025) 10.1145/3708984

[35]

Rehman "TrustFed: A Framework for Fair and Trustworthy Cross-Device Federated Learning in IIoT" IEEE Trans. Ind. Inform. (2021) 10.1109/tii.2021.3075706

[36]

Punugoti, R., Dutt, V., Anand, A., and Bhati, N. (2023). Exploring the Impact of Edge Intelligence and IoT on Healthcare: A Comprehensive Survey. Proceedings of the 2023 International Conference on Sustainable Computing and Smart Systems (ICSCSS), Coimbatore, India, 14–16 June 2023, IEEE. 10.1109/icscss57650.2023.10169733

[37]

Matathammal, A., Gupta, K., Lavanya, L., Halgatti, A.V., Gupta, P., and Vaidhyanathan, K. (2025). EdgeMLBalancer: A Self-Adaptive Approach for Dynamic Model Switching on Resource-Constrained Edge Devices. Proceedings of the 2025 IEEE 22nd International Conference on Software Architecture Companion (ICSA-C), Odense, Denmark, 31 March–4 April 2025, IEEE. 10.1109/icsa-c65153.2025.00081

[38]

Wang, X., Xu, Z., and Sui, X. (2025). Intelligent data analysis in edge computing with large language models: Applications, challenges, and future directions. Front. Comput. Sci., 7. 10.3389/fcomp.2025.1538277

[39]

Roth, H.R., Chang, K., Singh, P., Neumark, N., Li, W., Gupta, V., Gupta, S., Qu, L., Ihsani, A., and Bizzo, B.C. (2022). Split-U-Net: Preventing Data Leakage in Split Learning for Collaborative Multi-Modal Brain Tumor Segmentation, Springer. Lecture Notes in Computer Science. 10.1007/978-3-031-18523-6_5

[40]

Saldanha "Swarm learning for decentralized artificial intelligence in cancer histopathology" Nat. Med. (2022) 10.1038/s41591-022-01768-5

[41]

Schultze "Swarm learning for decentralized and confidential clinical machine learning" Nature (2021) 10.1038/s41586-021-03583-3

[42]

Systematic analysis of ChatGPT, Google search and Llama 2 for clinical decision support tasks

Sarah Sandmann, Sarah Riepenhausen, Lucas Plagwitz et al.

Nature Communications 2024 10.1038/s41467-024-46411-8

[43]

Shi "Comparison of LLMs in extracting synthesis conditions and generating Q&A datasets for metal–organic frameworks" Digit. Discov. (2025) 10.1039/d5dd00081e

[44]

Awan "Meta-analysis of large language models: Benchmarking DeepSeek-R1 against ChatGPT, Gemini, Qwen, and LLaMA" J. Big Data (2025)

[45]

Sandmann "Benchmark evaluation of DeepSeek large language models in clinical decision-making" Nat. Med. (2025) 10.1038/s41591-025-03727-2

[46]

A systematic review of large language model (LLM) evaluations in clinical medicine

Sina Shool, Sara Adimi, Reza Saboori Amleshi et al.

BMC Medical Informatics and Decision Making 10.1186/s12911-025-02954-4

[47]

Wu "Towards evaluating and building versatile large language models for medicine" NPJ Digit. Med. (2025) 10.1038/s41746-024-01390-4

[48]

Munir "Evaluating AI in Legal Operations: A Comparative Analysis of Accuracy, Completeness, and Hallucinations in ChatGPT-4, Copilot, DeepSeek, Lexis+ AI, and Llama 3" Int. J. Leg. Inf. (2025) 10.1017/jli.2025.10052

[49]

Menz "Current safeguards, risk mitigation, and transparency measures of large language models against the generation of health disinformation: Repeated cross sectional analysis" BMJ (2024) 10.1136/bmj-2023-078538

[50]

McGrath, M.J., Cooper, P.S., and Duenser, A. (2024). Users do not trust recommendations from a large language model more than AI-sourced snippets. Front. Comput. Sci., 6. 10.3389/fcomp.2024.1456098

Showing 50 of 144 references

Metrics

0

Citations

144

References

Details

Published: Apr 04, 2026
Vol/Issue: 15(7)
Pages: 1519
License: View

Authors

M

Mahbuba Ferdowsi

School of Systems and Computing, University of New South Wales, Canberra, ACT 2612, Australia

N

Nour Moustafa

School of Systems and Computing, University of New South Wales, Canberra, ACT 2612, Australia

M

Marwa Keshk

School of Professional Studies, University of New South Wales, Canberra, ACT 2612, Australia

B

Benjamin Turnbull

School of Systems and Computing, University of New South Wales, Canberra, ACT 2612, Australia

Cite This Article

Mahbuba Ferdowsi, Nour Moustafa, Marwa Keshk, et al. (2026). Trustworthy Intelligence: Split Learning–Embedded Large Language Models for Smart IoT Healthcare Systems. Electronics, 15(7), 1519. https://doi.org/10.3390/electronics15071519

Trustworthy Intelligence: Split Learning–Embedded Large Language Models for Smart IoT Healthcare Systems

You May Also Like