To err is human, but can AI tip the scales in our favour?

Medica Reporting Limited uses the qER tool from Qure.ai for detection and prioritisation of acute intracranial haemorrhage. We have previously demonstrated a faster turnaround (reduction of 80 seconds cf baseline) for these scans with life-threatening bleeds that will dictate modification to potential stroke thrombolysis pathways. Now a longitudinal audit of the qER tool in an acute setting has demonstrated outcomes similar to a radiologist in confirming and excluding acute haemorrhage (positive predictive accuracy 85.7 per cent, negative predictive accuracy 94.3 per cent). In addition qER identified several subtle findings that may have been missed by the radiologist alone.

This does not suggest that AI tools should be considered an autonomous replacement for radiologists. Moreover, augmented intelligence tools should assist the radiologist in the scan reporting cockpit. Don’t fly without it!

Join us for an in depth look at the data and some fascinating talks from other experts who have gone through the AI tool deployment challenges. Bring your thoughts on the governance aspects of this development in radiology.

Sign up for a view into our radiology AI present and future at the Royal College of Surgeons on Wednesday, June 26, with an accompaniment of food and liquid refreshments. And best of all, it is CPD, but free!

This news story has been sponsored by the companies concerned and does not represent the views or opinions of RAD Magazine.