Frimley Park Hospital in Surrey has become the first in the UK to implement a deep learning algorithm designed to improve the quality of CT scan reconstructions.

The algorithm, which is integrated with three new Canon CT scanners installed at Frimley Health NHS Foundation Trust, has been trained to differentiate ‘noise’ from true signal, reducing distortions and maintaining details in image outputs.

As such, the deep learning system can deliver quicker, more reliable and higher-quality image reconstructions from patient CT scans.

This provides reporting clinicians with more accurate image information to support diagnosis and treatment.

Carmina Esperanza, CT lead radiographer at Frimley Health NHS Foundation Trust, said: “From a CT scanning operative point of view, we don’t even know that we’ve entered the world of AI as we don’t need to do anything different when setting protocols before patient examinations, it’s all part of the CT system.

“The advantages to our radiologists…is much sharper images to report on. The benefits to our patient population is confidence that they are gaining the highest standard of care via investment in the latest medical imaging equipment.”

The implementation of the artificial intelligence (AI) algorithm forms part of the installation of three new CT scanners provided by Canon Medical Systems at the trust, which replace incumbent systems.

Mark Hitchman, managing director of Canon Medical Systems UK, said: “Advanced intelligent Clear-IQ Engine (AiCE) is a deep learning medical imaging innovation borne from the momentum of big data and health collaboration projects where high quality information from real patient datasets is drawn into AI systems to automate routine or repetitive tasks.”

“AI is an exciting future for healthcare, where patients will be able to spend more time with their consulting clinicians who are freed from repetitive or routine image review tasks to focus back onto patient care and specialist cases.”