• A total of 30 patient datasets (21 ischaemic and 9 haemorrhagic) were processed for this primary study analysis. The primary end point was met, with significant data collected to inform the value proposition and guide improvements in device hardware and software.
• It was observed that the EMVision device was able to classify stroke type (haemorrhagic or ischaemic) with an overall accuracy of between 93.3% and 96%.1
• It was observed that the EMVision device was able to localize targets in the correct quadrant (compared to ground truth CT/MRI) with an overall accuracy of between 86.7% and 96%.1
• Fusion methodology, which leverages data from multiple algorithms, produced particularly encouraging results, alongside select individual algorithms, which continue to be advanced.
• Positive feedback was received from both operators and patient participants on all scans.
• This is a data acquisition study and not intended to be an interventional study. Hence appropriate caution should be used in extrapolating these results to those of the general population at this stage of the development.
For further detail; ASX announcement link here