An AI-driven innovation hailing from Oxford University promises to avert thousands of heart attack fatalities by unearthing latent data within CT scans, potentially revolutionizing cardiovascular care across the UK. Currently under review by the National Institute for Health and Care Excellence (Nice), the technology could see nationwide integration within the National Health Service (NHS) by year-end, pending a favorable decision.
Professor Charalambos Antoniades, spearheading the ORFAN (Oxford Risk Factors And Non-Invasive Imaging) study, expressed optimism about the technology's trials, stating, "The technology's performance across multiple UK hospitals has been exceedingly encouraging, indicating its potential to save thousands from premature heart attacks or cardiovascular-related deaths."
Each year, over 300,000 Britons experience acute chest pain, leading to CT scans for cardiac anomaly detection. Despite this, fewer than 20% of scans reveal arterial obstructions warranting immediate concern. Prof. Antoniades, also the Chair of Cardiovascular Medicine at Oxford, highlighted a critical oversight: "The vast majority, 80% or more, receive a clean bill of health and are discharged, often without further intervention. Our findings suggest these reassurances may be misguided."
Alarmingly, two-thirds of this seemingly low-risk group later encounter significant, and occasionally fatal, cardiac episodes. "Our scans have overlooked vital indicators of danger," Prof. Antoniades admitted. "Addressing this glaring healthcare gap necessitates a sophisticated solution, and we contend that AI is optimally poised to do so."
Published recently in The Lancet, the research by Oxford's Radcliffe Department of Medicine aims to illuminate unnoticed irregularities in routine CT scans. By leveraging AI to enhance image resolution, hitherto unseen inflammation-induced arterial damage becomes visible, enabling preemptive interventions such as anti-inflammatory medications.
The methodology harnesses plaque characteristics and periarterial fat changes, signaling inflammation-linked artery health. "This analysis provides a definitive gauge of a patient's likelihood of experiencing a fatal cardiac event in the coming decade," Prof. Antoniades elaborated.
Initially calibrated with US datasets, the algorithm has since been validated through examination of 40,000 UK patient cases, confirming its efficacy. Prof. Antoniades shared, "There's a clear correlation: high coronary inflammation, as identified by our AI-augmented scans, correlates strongly with a heightened risk of severe cardiac incidents like heart attacks. We're effectively exposing the underlying precursors to these attacks."
Supported by the British Heart Foundation, the study found AI-assisted analysis prompted treatment alterations in 45% of cases, introducing intensified statin regimens or cardio-protective medications like colchicine.
Looking ahead, Prof. Antoniades disclosed plans to export this domestically-developed technology to the US, where it awaits Food and Drug Administration (FDA) approval, and Europe, where it has already gained clinical clearance, underscoring a global ambition to redefine cardiac risk assessment and prevention.
