EMVision Medical Devices has reported positive stage II interim data from its ongoing pre-validation trial of its artificial intelligence (AI) algorithms in diagnosing suspected haemorrhagic or ischaemic strokes.
The multi-site clinical trial is aimed at assessing EMVision’s neurodiagnostic AI technology, which could revolutionise the diagnosis of strokes.
Related: UroMems champions positive results for stress urinary incontinence implant
In the trial’s Stage 2 portion, 180 patients with stroke-like symptoms were enrolled across three hospitals: Liverpool, Royal Melbourne, and Princess Alexandra Hospital Brisbane.
Of these patients, 75 had confirmed ischaemic strokes, while the remaining 105 were non-ischaemic, including 18 haemorrhagic strokes, 67 stroke mimics, and 20 transient ischaemic attacks.
The Alberta Stroke Program Early CT Score (ASPECTS) was utilised to quantify early ischaemic changes in brain CT scans, with an average score of 7.4 in the ischaemic stroke cohort.
The interim analysis of the Stage II data has shown the AI’s capacity to accurately interpret complex patterns in patient data sets, particularly in detecting early onset hyperacute ischaemic strokes, which are typically difficult to identify using non-contrast computed tomography.
This success follows previous positive interim results that confirmed the AI model’s strong performance in determining the presence of haemorrhage.
As the trial progresses, recruitment for the final Stage 3 is underway, with over half of the intended cohort already participating.
The trial’s completion is anticipated in the coming months, with expectations to recruit up to 30 patients with haemorrhages.
EMVision CEO Scott Kirkland said: “The ability of our technology to also detect hyperacute and acute ischaemic cases in this cross validation interim analysis is incredibly exciting for our team, our clinical collaborators, and most importantly, what this may mean for the improvement of care pathways and outcomes for future stroke patients.”
