Artificial Intelligence-Based Decision Support Systems for Predictive Maintenance and Diagnostics of Medical Devices and Equipment

The article is devoted to the development and scientific justification of an integrated conceptual model of artificial intelligence (AI)-driven decision support systems for predictive maintenance and diagnostics of medical devices. It critically analyzes the underlying causes of the low efficiency of existing quality management systems (QMS) in the healthcare sector, which hinder maintaining the proper technical condition of equipment, ensuring patient safety, and optimizing costs. Enhanced metrics and criteria are proposed for proactive assessment of QMS reliability and performance, integrating classical indicators (reliability, operational availability, MTBF, MTTR) with modern condition indicators (Health Index, HI, Remaining Useful Life, RUL). Particular attention is paid to the potential of Predictive Maintenance (PdM), the Internet of Medical Things (IoMT), and big data for creating continuous telemetry streams that enable accurate diagnostics and early fault detection. The role of Digital Twins (DT) is substantiated as an operational digital layer for real-time decision support, improved medical equipment readiness, and maintenance scenario modeling. It is demonstrated how the integration of AI/Machine Learning (ML) with DT enables a closed-loop “data → model → decision” cycle, contributing to increased patient safety, asset management transparency, and product lifecycle management (PLM) optimization. Ethical, regulatory, and organizational challenges of implementation are addressed, including data protection, KPI standardization, algorithm transparency (DECIDE-AI), and workforce training. The obtained results form the foundation of a roadmap for the digital transformation of maintenance in healthcare and create prerequisites for managed quality of medical services