Current Issue
Download Full Article
Cite this paper as:
Lidia BĂJENARU, Cristian PETRACHE, Dragoș NICOLAU, "Model-Centric Governance for AI Modules in Multimodal Telemonitoring Platforms: A Conceptual Framework for the NeuroPredict Ecosystem", Romanian Cyber Security Journal, ISSN 2668-6430, vol. 8(2), pp. 45-61, 2026. https://doi.org/10.54851/v8i2y202604
Romanian Cyber Security Journal / Fall 2026, No. 2, Vol. 8
Model-Centric Governance for AI Modules in Multimodal Telemonitoring Platforms: A Conceptual Framework for the NeuroPredict Ecosystem
Abstract
Digital health platforms that combine Internet of Medical Things (IoMT) devices, cloud–edge infrastructures, and artificial intelligence (AI) analytics are increasingly used to support remote monitoring of patients with chronic and neurodegenerative conditions. Within this context, NeuroPredict integrates heterogeneous physiological, cognitive, behavioural, and environmental data streams with embedded AI modules designed to generate indicators for clinical monitoring and decision support. Although prior work has addressed platform architecture, cybersecurity, role-based access control, and multimodal data integration, the governance of AI models deployed in such ecosystems remains insufficiently explored. This paper proposes a model-centric governance framework for AI-enabled telemonitoring platforms, organized around six complementary dimensions: model integrity and provenance, controlled deployment and update governance, threshold and calibration governance, inference traceability through prediction records, multimodal consistency validation, and subgroup/ fairness monitoring under human oversight. The framework consolidates existing modelmanagement and traceability capabilities and extends them within a unified governance layer. By linking infrastructure-level security with model-level governance, this work contributes to trustworthy AI deployment in digital health ecosystems and provides a foundation for reliable, auditable, and transparent AI-driven telemonitoring and future real-world deployment.
Keywords
Cybersecurity, NeuroPredict platform, Digital Health, AI Governance, Internet of Medical Things