Clinical AI Pilot Concluded - Plain Film Reporting 90% Automated

Certainly an eye-opening future ahead!

Greater Fenland Hospitals NHS Trust - First English NHS Trust to liberate over 85% Radiologist time in unique collaboration between the “3 Clouds”: Google MedAI, Amazon Fog and Microsoft HazeHealth

Trust Transformation Manager Matt Stone: "Like many Trusts we have been continually drilled there is a shortage of Radiologists due to the ongoing lack of funding and training posts. The time came to take action. Our progressive forward-thinking I.T. led PACS Team took the brave step to move almost all ambulatory and in-patient Radiology and Pathology reporting to Google’s MedAI programme. MedAI has successfully reduced headcount and associated human cost across many sites in the USA and South Africa and it was time to replicate this in the NHS. Trialling this cloud based VNA-driven AI has allowed us to redeploy our remaining Radiologists back into trauma and ‘hands-on’ positions, a role many of them now relish with renewed vigour and which reduces pressures on Resident Doctors.

MedAI ingests images directly from the acquisition station via the Amazon hosted broker, cutting out the need for Radiographers to review or approve imaging, providing a simple ‘pass’ or ‘retake again’ (red/green traffic ANDON light) on our specially adapted modalities - a clear unequivocal signal to time pressed Radiographers. The workflow also allows Radiographers to now concentrate on the roles left vacant by wider recruitment freezes, broadening and sharpening their arsenal of skills across the healthcare environment. This is helped by the new patient direct results communication system hosted in Microsoft Haze which provides patients their final reports by SMS before they leave the car park.

PACS teams are also aided by the automated, AI led, patient fix-ups, merging and demographic changes - with imaging and metadata sorted appropriately without the typical interventions being required.

The only minor drawback has been the possible dose changes needed for acquiring images for computer review rather than human review, something our Medical Physics outsourced team are monitoring for possible further improvements. Student placements have also been reduced as there is no real need for precision positioning when utilising the AI systems collaboratively.

Overall, our Radiologists are happier working with direct patient contact, taking bloods and assisting with medication administration away from the department. Our Radiographers also enjoy the myriad of opportunities across the wards and clinics that other duties take them. We look forward to working with Google and Amazon and Microsoft to offshore more of our countries patient data in an unstructured manner plus reskill our essential clinical staff in areas beyond the boundaries of Radiology in the future".

- released for publication on 1st April 2026.

4 Likes

What’s been really great, and nobody seems to talk about, is that it’s freed up time for colleagues to generate more standard operating procedures, local and regional guidelines, departmental policy documents and we’ve managed to use it all for Fenland’s ISAS application whilst the operational team were on an away day. FantAIstic!

2 Likes

Who’d have thought AI would make radiologists happier instead of jobless? Guess even April has its surprises!

2 Likes

I think April 1st may be a factor here

1 Like