Medica and Qure.ai white paper reveals: AI matches radiologists in 93.5% of CT scans in acute imaging workflow

Medica, a leading provider of teleradiology services, proudly hosted a landmark event at the Royal College of Surgeons, unveiling a joint white paper with Qure.ai on the transformative role of AI in radiology. This event, chaired by Medica's Chief Medical Officer Dr. Robert Lavis, marked a significant milestone in Medica's 20th year, showcasing the role of AI in radiology's evolution.

The findings of the white paper were presented by Surabhi Srivastava, Commercial VP UK at Qure.ai. The data was taken from analysing 1,315 non-contrast CT head scans with acute presentation histories using the qER algorithm. The study demonstrated a 93.5% overall agreement between radiologists and the AI algorithm, highlighting its potential as a reliable tool for bleed detection and prioritisation. Specific weaknesses, such as subtle subarachnoid hemorrhage detection, were identified, underscoring the need for continuous monitoring and improvement.

Dr Neelan Das, Consultant Interventional Radiologist at East Kent Hospitals University NHS Foundation Trust, shared insights from his extensive journey in AI deployment, outlining the five essential steps to success: identifying a specific need, engaging stakeholders, setting clear goals, pre-deployment validation, and ongoing surveillance. His insights emphasised the importance of understanding algorithm performance within specific populations.

The research and development perspectives were covered by Dr. Sarim Ather, Radiology Consultant and AI Lead, Oxford University Hospitals NHS Foundation Trust. Dr Ather highlighted the critical role of post-deployment monitoring and the potential of AI tools in healthcare screening for conditions like osteoporosis and coronary artery calcification. He stressed the necessity of rigorous assessment in real-world environments to ensure reproducibility and reliability.

The final presentation was delivered by Dr. Amrita Kumar, chair of AI Working Group at Frimley Health and chair of AI & Innovation SIG at British Institute of Radiology. Dr Kumar provided an in-depth exploration of the comprehensive clinical governance framework required for AI deployment. She emphasised the importance of stakeholder engagement, oversight, ethical considerations, and staff training, advocating for robust AI tools capable of hanwithout service disruption.

Following the keynote presentations, Medica CEO, Andrew Cannon, moderated an engaging panel discussion with questions from attendees who were highly engaged. Many of the questions centred around the legal challenges and the future of autonomous AI tool reporting. The discussion revealed diverse perspectives on the timeline for autonomous AI adoption, with estimates ranging from immediate implementation to a three-to-five-year horizon. The panel also debated the validity of using humans as the gold standard for AI comparison.

Dr. Robert Lavis expressed his gratitude to the Medica team, Qure.ai, the speakers, and the audience for their contributions to a thought-provoking and insightful evening. "The event underscored the remarkable advancements in AI technology and its potential to revolutionise radiology," said Dr. Lavis. "As we continue to explore and integrate AI tools, Medica remains committed to advancing patient care through innovation and collaboration.

Below are some pictures from this successful event: