Lifecycle Framework for Image-Guided Therapy Systems

Background
The Philips Image-Guided Therapy Systems (IGTS) portfolio spans over three decades and includes complex imaging and therapy platforms such as Azurion, Allura, iApps, and EchoNavigator. These systems continue to evolve through hardware and software updates, while earlier generations remain in clinical use. Because each generation was developed under different international standards (e.g., IEC 60601-1, ISO 14971, IEC 62366), maintaining consistent risk management, usability validation, and regulatory traceability is challenging. To address this, Philips IGTS implemented a data-model-driven lifecycle framework linking requirements, risk assessments, usability evidence, and verification data across generations.

Approach
The framework integrates digital data modeling with human factors (HF) engineering and risk management. A relational lifecycle data model connects requirements, hazards, and risk controls to applicable standards. HF engineering is embedded through targeted post-market usability studies, including simulated use evaluations, interviews, and expert reviews, particularly for legacy subsystems. HF specialists re-analyze field events to update designs and risk files. A living FMEA database links hazards, mitigations, and usability findings across generations. Verification and usability testing focus on modified risk controls and worst-case use scenarios.

Results
Implementation improved efficiency, consistency, and risk visibility. Integrated traceability strengthened alignment between usability findings, risk assessments, and verification outcomes, enhanced audit readiness, and reduced redundant testing. Post-market usability data supported more precise residual risk evaluation and timely field issue resolution.

Conclusion
This framework demonstrates how data-driven risk management and HF engineering can sustain safety and regulatory compliance across long-lived medical systems, providing a scalable model for evolving and AI-enabled technologies.