Dutch radiologists showcase pioneering vendor-neutral AI setup

Researchers at Utrecht University Medical Center (UMC) built a flagship AI infrastructure that can run any algorithms in clinical practice, regardless of vendor equipment.
By Mélisande Rouger
12:00 AM

WHY IT MATTERS

The new platform enables easier deployment of AI tools and helps improve clinical workflow at the hospital.

The IMAGR infrastructure took three years to be build, as well as €150k in hospital funding and a €650k university grant. It is run through Dockers, a platform that takes the algorithm and deploys it through the infrastructure, without users noticing the source code.

The result is a tangibly enhanced workflow in daily routine practice with complete vendor independence. Most AI solutions available today are vendor dependent, which complicates their deployment in clinical practice.

THE LARGER TREND

IMAGR is able to monitor information passing through the HIS, RIS and PACS and take action whenever necessary. For example, it can recognise a brain MRI and choose appropriate algorithms that can help diagnose and quantify. The infrastructure will then recover the patient’s images from the PACS and activate the pipelines it deems appropriate, and either send resulting images back to the PACS or present them in a dedicated AI viewer on the PACS station - or both. Physicians can also help enhance the system, by pointing out when an algorithm fails or succeeds to perform a task.

As a result, many tasks that were time consuming can now be automatically done in daily practice thanks to the algorithms deployed in the infrastructure. For example, it can trigger white matter hyperintensities segmentation and run brain segmentation at the same time, meaning that radiologists no longer have to quantify these lesions manually, a task that could take up to 20 minutes per slide.

The Utrecht team has also helped improve an existing algorithm that can quantify fat and muscle in the body, an increasingly useful piece of information in oncology and cardiovascular disease, within seconds.

ON THE RECORD

“We created an in-house infrastructure to upscale the AI algorithms to the eyeballs of the radiologist. Not on some laptop away from the reading room, but right inside the room, on every work station,” said Dr Tim Leiner, professor of radiology and chair of cardiovascular imaging at the department of radiology and nuclear medicine at Utrecht UMC.

“Our infrastructure is completely vendor agnostic. We don’t want any single company’s platform, because then you are stuck with their product and can’t use anything else. We wanted our own tool, so that we could very rapidly incorporate every new algorithm out there,” he said.

“If we have a patient with a request for dementia and that patient is sent to the MRI, we can already trigger this pipeline upfront, when the images are sent from the MRI scanner to the PACS, so that when the radiologist opens and reads the case, the data is already there,” he explained.

Healthcare IT News is a publication of HIMSS Media.