Philips announces strategic partnership aimed at advancing AI in MRI
Health tech giant Philips will be forming a strategic partnership to advance the use of artificial intelligence in the MRI space.
Philips announced the collaboration with Nvidia Monday during the International Society for Magnetic Resonance in Medicine annual meeting, citing the AI solution leader's expertise as key to advancing capabilities within the modality. Together, the duo will build an MRI foundational model for improving every aspect of MR imaging, from acquisition to scan times and post-processing.
“Our AI-powered MRI solutions are already enabling healthcare providers to deliver better care to more people,” Dr. Ioannis Panagiotelis, Business Leader of MRI at Philips, said in the announcement. “By partnering with NVIDIA to build an MR foundational model, we’re pioneering a new frontier for medical imaging, one that has the potential to transform the role of MR in the diagnosis and treatment of a wide range of diseases. The benefits for patients and healthcare providers could be enormous.”
The deep learning neural network will be trained on massive sets of data, with the goal of enabling zero-click scan planning for multiple sequences across all anatomical regions of interest. This has the potential to drastically speed up workflows and increase patient throughput. There are plans to include multiple image enhancement techniques in the model’s development, including deep learning denoising and image sharpening, providing super-resolution images and more accurate diagnoses. The model also will allow radiologists to preview and adjust parameters prior to imaging, and it will use AI to automatically detect and flag certain suspicious findings.
It is the team’s hope that by integrating the foundational model into Philips’ existing MRI scanners, providers can seamlessly improve the efficiency, accuracy and quality of current workflows.
The ISMRM meeting is set to wrap up later this week in Honolulu, Hawaii.
Learn more about the new partnership here.