Who are we?

We are an small research team evaluating whether we can use deep-learning to assist with the detection of intracranial haemorrhage in the emergency department. To test this we have designed an AI co-pilot that assists with the interpretation of axial non-contrast CT head scans. The model is a deep neural network, that is able to recognize and locate intracranial haemoaahge on CT scan. Our hope is that this will be able to assist emergency doctors with their workflow, by helping to triage patients with head injury.

Why bother?

Intracranial haemorrhage is a neurosurgical emergency caused by bleeding within the skull. Delayed identification is associated with worse neurological outcomes if required neurosurgery is delayed. Rapid diagnosis depends on the accurate interpretation of CT head scans. However, this can sometimes represent a diagnosis bottle neck, becuase specialist radiologist have to formally interpret scans. Therefore the use of an AI co-pilot that could help triage CT head scans would a valuable clinical tool.
We anticipate the greatest beneficiaries of this technology to be frontline emergency clinicians who are often the first to review the imaging. We have designed a co-pilot system whereby the model will review the scan alongside the emergency doctor, providing an extra set of eyes. We believe these models would be particularly valuable for departments without on-site radiologist support, where time taken to await formal reports from off-site specialists may delay neurosurgical referral and timely intervention.

How do I get involved?

Trial recruitment has now closed.