Bering achieves UKCA mark for AI-Supported Chest X-Ray Classification

 

iCAIRD consortium partner Bering Limited has achieved a Class I UKCA mark for BraveCX, its Artificial Intelligence (AI)-augmented software for Chest X-Ray (CXR) classification.

BraveCX was trained on over 1,000,000 CXRs acquired across diverse clinical settings using the Safe Haven AI Platform (SHAIP). SHAIP is a trusted research environment constructed specifically for machine learning within the health board network and deployed in NHS Greater Glasgow & Clyde through Industrial Centre for Artificial Intelligence Research in Digital Diagnostics (iCAIRD).

Working within iCAIRD enabled secure access to vast quantities of high-quality training images and the invaluable clinical resources required to co-develop a safe and accurate CXR triage product. In pre-clinical studies, BraveCX detected normal CXRs with over 95% accuracy and achieved 97% concordance in radiological sign identification with three Consultant Radiologists.

The software can be deployed either as a cloud-based service, directly on premises, or integrated with third-party systems. It interfaces seamlessly with Radiology Information Systems (RIS) and outputs the probability of CXRs being normal or abnormal, identifies the likelihood of abnormal radiological signs, and highlights them on an image. This information could allow health professionals to prioritise their CXR reviews.

In achieving UKCA Class I, Bering has aligned its software development processes and architecture to meet the safety and performance requirements for software as a medical device as part of the UK MDR and European Directive 93/42/EEC on Medical Devices.

Professor David Lowe, Joint Clinical Lead for the West of Scotland Innovation Hub and Emergency Medicine Consultant at Queen Elizabeth University Hospital said: “This is another welcome development demonstrating the potential for AI to support clinicians ensuring patients are getting the highest quality, timely and safe treatment. We continue to see the positive potential impact AI could have on radiology, to reduce waiting times to supporting staff with decision making when it comes to diagnosis and clinical care.

“Technology like this does not mean patients will have less face-to-face time with doctors and could help speed up processing high numbers of similar cases, while retaining accuracy, allowing clinicians to spend more time with more complex patient cases.”

Dr Mark Hall, Radiology Consultant at NHS Greater Glasgow & Clyde added, “Emerging into this post pandemic world we are seeing huge resource demands on radiology. We are just starting to utilise the positive potential AI can have on radiology. With this development, there are potential positive impacts from prioritisation of reporting to improving accuracy and reducing pressures on staff. The more consistent use of AI within imaging will help patients get results faster and improve radiology confidence when issuing a report.”

iCAIRD CTO James Blackwood added: “This is an indication of major success for the programme and its work with Bering Ltd. To have progressed from first principles research through to development of an at-market product in the three years of iCAIRD is testament to the incredible hard work and achievements of colleagues and collaborators from across healthcare, academia and industry on the programme and couldn’t have been achieved without the consortium and the core iCAIRD team. This is exactly why UKRI created the From Data to Early Diagnosis in Precision Medicine funding challenge.

“Scotland has a rich life sciences ecosystem which can shorten the distance between inventors and their customers. iCAIRD is a national programme of collaborators which focusses that ecosystem on delivering AI solutions for digital diagnostics.”

Dr. Ignat Drozdov, CEO and founder of Bering said: “After over three years of research and collaboration with clinical teams, it’s so exciting to see BraveCX emerge as a state-of-the-art tool that has actually ‘listened to the end user’. Aligning with UKCA means Brave CX prioritises patient safety, whilst still delivering the most advanced Risk Stratification algorithms where they are needed the most.”

 

Full article available via Bering Ltd at https://beringresearch.com/bering-achieves-ukca-mark-for-ai-supported-chest-x-ray-classification/