About the Session
AI algorithms are rapidly entering radiology workflows, but integrating them into production imaging environments introduces complex infrastructure challenges. PACS teams must manage study routing, DICOM variability, compute latency, result ingestion, and workflow orchestration across systems that were never designed for algorithm pipelines. These challenges multiply when organizations attempt to integrate multiple AI tools across fragmented imaging platforms and vendors.
This session examines the engineering realities of deploying AI in clinical imaging environments, drawing from real-world infrastructure experience. Damien Evans will walk through common failure points PACS teams encounter when implementing AI and explore architectural patterns that improve performance, reliability, and scalability. Particular attention will be given to platform-based approaches that consolidate routing, AI orchestration, and workflow integration within a unified imaging environment, reducing system complexity while enabling AI to operate seamlessly within clinical workflows.
Objectives
- Analyze the infrastructure challenges associated with integrating AI algorithms into PACS and enterprise imaging environments, including study routing, DICOM inconsistencies, compute latency, and results integration.
- Evaluate architectural patterns used to support AI pipelines in production imaging systems, including routing layers, orchestration frameworks, and workflow-aware result delivery.
- Assess how unified imaging platform architectures can reduce operational complexity by consolidating AI processing, workflow management, and imaging infrastructure into a single system rather than relying on fragmented multi-vendor integrations.
Sponsored By
