Brainomix has announced the publication of a prospective real-world study demonstrating that its artificial intelligence (AI) imaging platform, Brainomix 360 Stroke, significantly increased rates of endovascular therapy (EVT) procedures in large vessel occlusion (LVO) stroke patients, and reduced delays in patient triage and transfer. The data have been published in a paper in The Lancet Digital Health.
“This landmark study confirms what stroke physicians have been witnessing on their wards every day—that this AI technology is changing lives,” said David Hargroves, UK National Health Service (NHS) national clinical director for stroke and co-author of the study. “It is slashing waits for a stroke diagnosis, hugely increasing the chance of a stroke patient to receive life- and disability-saving treatments, improving their chances of returning to independent living and a better quality of life. This publication is the result of a strategic investment by the UK government into AI research and its integration into the NHS, providing robust evidence that supports national clinical guidelines.”
The present study evaluated the impact of AI imaging software on stroke treatment, encompassing data collected by the national stroke audit registry—the Sentinel Stroke National Audit Programme (SSNAP)—from 107 hospitals in England over a five-year period, representing more than 450,000 patients. Brainomix 360 Stroke was deployed across four regional stroke networks consisting of 26 hospitals in England, including both primary and comprehensive stroke centres, where data from more than 71,000 individuals with ischaemic stroke were collected, and a pre- and post-implementation analysis was conducted.
Evaluation sites using Brainomix 360 Stroke saw a 100% increase in EVT compared with a 63% increase at non-evaluation sites. According to Brainomix, at the level of the individual patient, the use of AI stroke software was also associated with an increased likelihood of EVT compared with patients for whom AI software was not used (odds ratio [OR], 1.57; p<0.0001). Furthermore, the increase in treatment rates as a result of the use of Brainomix 360 Stroke was associated with improved patient outcomes—meaning less severe disability at the time of hospital discharge.
The increase in patient treatment with EVT was most marked in primary stroke centres, suggesting that the greatest benefit of stroke AI software lies at hospitals with fewer specialist doctors who routinely use accessible imaging like non-contrast computed tomography (CT) and CT angiography, Brainomix states. The company adds that primary stroke centres using its 360 Stroke platform also saw a notable reduction of 64 minutes in door-in-door-out (DIDO) times between admitting a patient and transferring them to a specialised centre for EVT.
A final finding highlighted by Brainomix is a higher rate of intravenous thrombolysis (IVT) for stroke being observed at evaluation sites, with patients reviewed via AI software found to be more likely to receive IVT (OR, 1.99; p<0.0001).
The company notes in a recent press release that these study results support current clinical guidelines from both the UK NHS and the American Heart Association (AHA)/American Stroke Association (ASA), which recommend the routine use of AI imaging decision-support software in acute stroke care.
“This is an AI solution that is meeting a clear need, and delivering impact and benefit for patients across the NHS,” commented Gary Ford, chief executive of Health Innovation Oxford and Thames Valley (Oxford, UK), which carried out the evaluation. “Our evaluation is the largest into the impact of using stroke AI on increasing access and timeliness of the delivery of thrombectomy to people with stroke. It has highlighted the large impact on improving access to this vital treatment, particularly for patients who initially present to hospitals without the capability to deliver thrombectomy.”
“As a company that prides itself on a foundation of evidence-based medicine, we are delighted with the publication of this landmark study of Brainomix 360 Stroke in The Lancet Digital Health. The significant increase in mechanical thrombectomy and IVT with an associated improvement in clinical outcomes validates what we have seen in other studies, but at an unprecedented scale,” added George Harston, chief medical and innovation officer at Brainomix. “It is particularly exciting to see that access to care improves in all hospitals and particularly in non-specialist general hospitals, where there is the greatest unmet need, and to which most stroke patients will initially present. Through this unique collaboration with clinical teams across NHS England and the Health Innovation Network, we have demonstrated the generalisable impact that stroke AI technology can have for improving patient care.”
