Home / Video case study: Stroke and AI

Video case study: Stroke and AI

March 10, 2023

Stroke & AI: How the e-Stroke software is reducing door in door out time for stroke patients.

The number of strokes, especially in younger people, is increasing. And also over the years, we’re seeing an increasing number of people having strokes. 

There are two kinds of stroke. One is called ischaemic stroke and the other one is primary intracerebral haemorrhage. The ischaemic stroke is caused by a blockage of a blood vessel. The standard treatment up until 2015 is to give intravenous thrombolysis, which is injecting a medicine called Alteplase, and that disperses the clot.

But when a large blood vessel is occluded, the intravenous thrombolysis alone is not enough to disperse the clot. That’s when the mechanical thrombectomy, (mechanical removal of the clot) is beneficial. For every three people treated with this procedure, one person will be independent at three months. We believe 10% of stroke patients should receive the treatment in the UK.

But if you look at the numbers, we’re only providing this treatment to 1 to 2% of the patients actually.

Daily most stroke happens in the community, so patients get admitted to community hospitals and every 30 minutes a stroke patient dies or remains permanently disabled, not because of the stroke, but because they are admitted to a hospital that doesn’t have the right expertise in a timely manner to diagnose and treat this patient. 

So what we use with AI in our technology is essentially a tool to process the scans of stroke patients in an acute setting where every second counts.

In order to provide frontline clinicians with more information from the scans that they would anyway capture and have to interpret and ultimately help these frontline physicians to make more timely and better diagnosis and treatment decisions.

In stroke, we always say time is brain. When we lose time, we lose brain. So when a patient has a large blood vessel occlusion stroke, every minute we delay, patients lose 2 million nerve cells. Since we introduced artificial intelligence, the door in door out time has come down by more than 60 minutes. And when we looked at the outcome for the patients since we introduced artificial intelligence, the percentage of patients who achieved independence as a result of this treatment has tripled.

The second advantage is that we can have access to these images remotely. So when a clinician is on call, they can have access to the vital imaging data within two or three minutes of the scan being undertaken.

Currently, there are more than 70 stroke units in the UK that are routinely using e-Stroke to manage patients. And we expect that AI and stroke in particular could become the standard of care so that all hospitals have a solution to optimally diagnose and treat their patients.

This is a very good example of how NHS hospitals, providers, and commissioners can work together to improve care and in our case, artificial intelligence technology facilitated it.