AI-powered MRI previews aim to reduce errors and rescans
Autonomous MRI powered by AI could improve patient experience, access, and diagnostic consistency across hospitals and regions.
Philips is creating AI-driven predictive MRI previews to improve scan planning and reduce operator variability. Using NVIDIA accelerated computing and foundation models, the system creates a pre-scan image to validate protocols, optimise positioning, and spot potential issues.
The technology is based on a dedicated MR foundation model trained on diverse datasets covering anatomies, field strengths, protocols, and artefacts.
When combined with NVIDIA’s NV‑Generate, NV‑Segment, and NV‑Reason models, the platform integrates image generation, segmentation, and interpretation. It creates a single intelligent workflow that supports consistent and efficient MRI procedures.
Predictive previews reduce rescans, enhance image quality, and increase technologist confidence, especially in complex exams or areas with limited expertise. Early guidance helps confirm protocols, optimise positioning, and flag issues that could affect diagnostic outcomes.
Philips envisions autonomous MRI, with AI monitoring image quality, guiding positioning, and assisting radiologists with actionable insights. Predictive imaging boosts consistency, efficiency, and access, improving patient experience and expanding MRI availability.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!