A4 Refereed article in a conference publication
Enhancing Privacy Transparency in Remote Patient Monitoring with Explainable AI
Authors: Trivedi, Jolly; Isoaho, Jouni; Mohammad, Tahir
Editors: Shakshuki, Elhadi
Conference name: International Conference on Mobile Systems and Pervasive Computing
Publisher: Elsevier BV
Publication year: 2025
Journal: Procedia Computer Science
Book title : 20th International Conference on Future Networks and Communications/ 22nd International Conference on Mobile Systems and Pervasive Computing/15th International Conference on Sustainable Energy Information Technology (FNC/MobiSPC/SEIT 2025)
Volume: 265
First page : 149
Last page: 156
eISSN: 1877-0509
DOI: https://doi.org/10.1016/j.procs.2025.07.167
Publication's open availability at the time of reporting: Open Access
Publication channel's open availability : Open Access publication channel
Web address : https://doi.org/10.1016/j.procs.2025.07.167
Self-archived copy’s web address: https://research.utu.fi/converis/portal/detail/Publication/500454550
This paper presents an Explainable Privacy-Preserving Intelligent System for Monitoring (X-PRISM) framework designed to enhance transparency and privacy in AI-based remote patient monitoring (RPM) systems. The framework addresses the critical demand for explainable AI (XAI) in healthcare by integrating explainability techniques, such as SHapley Additive exPlanations (SHAP), to provide clarifications and reasoning behind AI-driven decisions based on key patient metrics such as heart rate and body temperature. X-PRISM implements pseudonymization and encryption to improve privacy and secure sensitive healthcare data while ensuring compliance with global regulations such as the General Data Protection Regulation (GDPR). Federated learning plays a vital role in the framework by enabling decentralized training of AI models across multiple healthcare nodes without directly sharing patient data. The layered architecture of X-PRISM enables seamless data collection from wearables and IoT devices, preprocessing in Azure Cloud, and AI model development using TensorFlow. X-PRISM is designed to offer transparent decision-making, protect patient data through decentralized training, and enable real-time interpretable feedback in RPM applications. Although the framework is not empirically tested in this study, it is presented as a foundation for future research and development in ethical AI deployment in healthcare. Existing gaps in RPM systems and synthesizing best practices in AI explainability and data privacy are reviewed in this study, which lays the foundation for developing secure, interpretable, and regulatory-compliant RPM systems. X-PRISM establishes the groundwork for advanced ethical, scalable, and trustworthy AI applications in RPM by addressing the dual goals of privacy and explainability, thereby enhancing patient trust and healthcare outcomes.
Downloadable publication This is an electronic reprint of the original article. |  |
