Philips and NVIDIA Partner to Revolutionize MRI with AI Foundation Model
Philips and NVIDIA have announced a strategic collaboration to develop a powerful AI foundation model aimed at transforming magnetic resonance imaging (MRI). The announcement was made at the 2025 International Society for Magnetic Resonance in Medicine (ISMRM) meeting, capturing widespread attention across the medical imaging community.
Transforming MRI Through AI Innovation
The new AI model is designed to enhance MR image quality, accelerate scan times, and improve diagnostic workflows. By leveraging deep learning, this foundation model will support a wide range of clinical applications, Philips stated.
Also read: Nvidia Plans Downgraded H20 Chip for China Amid US Export Restrictions
According to the company, the model will allow zero-click scan planning across various anatomies. Radiologists can preview images and adjust parameters like speed and resolution before scanning, streamlining the diagnostic process and reducing manual workload.
This efficiency could result in significantly higher patient throughput, allowing imaging centers to serve more patients with improved accuracy.
Advanced Features for Enhanced Diagnostic Confidence
The AI model won't stop at image planning. It will also facilitate automatic detection and interpretation of findings, helping radiologists spot abnormalities faster and with more confidence.
This foundation model builds upon NVIDIA’s Vista-3D, a powerful tool for segmentation analysis, and MAISI (Medical AI for Synthetic Imaging), a 3D model capable of generating synthetic medical images. Both technologies are critical to training AI that can adapt to real-world MR imaging scenarios.
Philips emphasized that the collaboration will produce a domain-specific solution designed to meet the unique challenges of MR diagnostics.
