How Shields Health is using AI to revolutionize MRI wait times, patient experience

Few patients enjoy getting an MRI. Often, there’s a long wait to schedule the scan, and then the time required in the machine — which can be close to an hour — can be scary, loud and claustrophobic.
Shields Health, a large Quincy-based provider of medical imaging at over 50 locations across New England, says it is using artificial intelligence to cut that time in half.


“This is about as exciting a technology that we’ve seen come into MRIs in decades,” said Peter Ferrari, Shields Health president.


Shields is working with GE Healthcare to implement AI software in the medical diagnostics process, not only shortening the time a patient spends in a machine by more than half, but also improving image quality and diagnostic accuracy, Shields said.


The AI software, used by radiologists, delivers a clearer image and acts almost like a co-pilot in pinpointing diagnoses, executives said.


Shields first started using the technology in 2020. Now approximately one-third of its fleet of MRI machines are equipped with the software. The buildout of AI comes at acritical time, Ferrari said, because
the bottlenecks that are plaguing other parts of the healthcare system are impacting patients’ access to MRIs too.


The shorter run time of the MRI machines, down from an hour to about 15 to 20 minutes with the new technology, means more patients can be seen. Ferrari said patient throughput has improved by 20% to as much as 50% across locations — and patients are no longer having to wait weeks to be scheduled for the scan.


“Our ability to shorten those backlogs is incredibly important,” Ferrari said.


Shields receives patient referrals primarily from community physicians, and being able to access an MRI outside of a major Boston hospital can be both more convenient and less expensive for patients, Ferrari said.


Shields is “pecking away” at updating the remainder of its locations with the machine-learning and AI reading capabilities, but to do so requires significant capital expenditure for each setup. A typical machine update is about $2 million, Ferrari said, but that does not include updates to other parts of the exam room, which can also be in the million-dollar range.


Ferrari said he expects that the near future will see nearly all MRI machines equipped with a similar technology.


“It’s been quite honestly revolutionary in the world of MRI,” he said.