AI helps radiology team spot concerns sooner on chest X-rays

Patients with potentially serious findings on a chest X-ray can now be identified sooner thanks to an artificial intelligence (AI) tool supporting radiology teams across Kettering and Northampton general hospitals. 

The software reviews chest X-rays for patients over 16-years-old and flags areas that may need closer attention. It also helps the team put scans with potentially critical findings, including suspected cancer, nearer the top of the reporting list. 

A clinician still examines each X-ray independently, checks the AI’s findings and makes the final judgement. 

Cornilius Mushambi, Consultant Radiographer, described the tool as a “second pair of eyes”. He explained the process can draw attention to subtle abnormalities on an image that might otherwise be difficult to see. 

The radiology team can then assess whether the finding needs further investigation. For patients, that could mean a possible lung cancer is investigated earlier, while those whose scans appear most urgent can be prioritised for review. 

Cornilius shared an example of how a quicker pathway can benefit a patient. He said: “AI acts as a second pair of eyes, highlighting subtle abnormalities that might otherwise be missed, this increases the confidence of reporters, who retain the final judgement.  

“It means potentially a patient can leave our hospital with both their X-ray and CT completed and reported on by a radiographer. Without input from AI, these steps that might otherwise have taken place days apart.” 

The technology was introduced in November 2024 with NHS England funding, and the team began using it to prioritise reporting in April 2025.  

Since then, AI has been helping reporters to look at 62,661 chest X-rays during that time. That figure is a count of scans over the period, not a measure of cancers detected or patient outcomes. 

For now, the AI findings are visible to radiology staff. The team is considering whether marked-up images could in future be made available to other clinical teams, such as emergency department staff. This could help clinicians see a flagged finding sooner, particularly out of hours. 

By helping radiology teams focus on findings that may need urgent attention, AI offers another layer of support for patient care. It does not replace clinical expertise: every X-ray still requires a clinician’s assessment and final judgement. The aim is to help patients who need further investigation move to the next stage of their care sooner.