NHS Trust deploys AI algorithms for radiology reporting


Dartford and Gravesham NHS Trust has deployed artificial intelligence (AI) algorithms in a bid to support quicker radiology reporting.

The Trust has partnered with Behold.ai, a radiology AI specialist to implement its red dot prioritisation platform. The red dot AI algorithm separates radiology studies into normal and abnormal categories alerting radiographers to urgent findings at the point of acquisition and helping clinicians prioritise workloads and reduce reporting backlogs.

Neil Perry, associate director Digital Transformation for Dartford and Gravesham NHS Trust said: “Radiology departments play a crucial role in our healthcare network and the ability to report and diagnose as quickly and accurately as possible has a major impact on patient outcomes.

“Behold.ai’s AI technology is specifically designed to support our skilled radiology workforce by understanding and supporting their existing workflows and enabling them to manage reporting more efficiently. It also allows us to work with our existing radiology technology which further demonstrates its ability to support our existing working processes. We are extremely excited to work with Behold.ai and take the technology to the next level with automated reporting of key diagnoses and expanding to other investigation groups.”

The Behold.ai algorithms have been developed using more than 30,000 example images, all of which have been reviewed and reported by highly experienced consultant radiology clinicians in order to shape accurate decision making. The red dot prioritisation platform employed at Dartford is capable of sorting images into normal and abnormal categories in less than 30 seconds post image acquisition.

This means that the radiology team know, before the patient leaves the radiology department, whether their examination is normal, or whether there is an abnormality present – an instant triage.  This is valuable information for the radiologists, clinicians around the Trust and ultimately the patient. The algorithm into existing workflows for report prioritisation and distribution, as well as critical findings alerts.

Radiology departments play an essential role in the diagnostic process, however a consequence of fewer radiologists and growing demand for images has left services stretched beyond capacity across many trusts, resulting in reporting delays – in some cases impacting cancer diagnosis. These service issues have been highlighted by the Care Quality Commission and the Royal College of Radiologists.

Simon Rasalingham, CEO and chairman of Behold.ai said: “Our algorithms have been developed by highly-trained and experienced consultant radiologists who have continually reviewed, annotated and reported on patient images to help train the algorithm correctly.

“This means we are able to replicate the real-life work of a radiologist via our red dot algorithm, with the aim of ensuring exceptional real-world clinical accuracy. This is the key difference, we believe, in the performance of AI. Our solution seamlessly integrates into local trust workflows augmenting clinical practice and delivering state-of-the-art, safe Artificial Intelligence.

“The implementation at Dartford is a great opportunity to showcase the practical application of AI technology in the healthcare sector and demonstrates how it can be used to deliver improved patient care. We are in the process of rolling out the red dot solution across the NHS, delivering an essential, cost effective and safe alternative to the way radiology departments can manage their workload.”

NHS Trust deploys AI algorithms for radiology reporting