Stroke wards with access to the AI imaging platform Brainomix 360 were able to double their thrombectomy rate and reduce door-in-door-out (DIDO) time by 65 minutes, its creator says.
The results of a three-year study into stroke-focused AI options involving 71,000 patients in 26 hospitals will be shared at the International Stroke Conference (ISC) in LA today.
Within the data is further evidence for the platform developed by Oxford-based Brainomix, which is powered by what the firm calls “the most advanced and clinically validated AI algorithms built around basic imaging”.
The AI tool assesses non-contrast CT scans to automatically generate an ASPECTS [Alberta Stroke Program Early CT Score] value – which determines stoke severity using tomography data. It features a heatmap that enables a more nuanced assessment of each brain region.
The data shows that sites using Brainomix 360 doubled their thrombectomy rates over the evaluation period, increasing an additional 37 per cent compared with non-evaluation sites, while also achieving a faster door-in-door-out (DIDO) time by 65 minutes.
Dr Waleed Brinjikji, a professor of radiology and neurosurgery at US healthcare group Mayo Clinic, which tested the tech, said: “These latest studies have yielded fascinating insights around eloquence and brain frailty, revealing a more nuanced understanding of stroke, and informing potentially novel approaches to patient selection for stroke treatment.”
Dr George Harston, chief medical and innovation officer at Brainomix, said: “The results from this large-scale study are unmatched in the field, robustly validating the power of Brainomix 360 to improve stroke care, enabling more patients to receive life-changing treatment.
“We are now seeing similarly impactful results from US stroke networks.”
