Publication Spotlight

How do stroke AI algorithms help with patient treatment?

Generally used to confirm a clot blocking blood flow to the brain tissue, in this field neuroimaging can give much more thanks to stroke AI algorithms.

hands holding mobile with StrokeViewer app

The importance of detecting the stroke’s cause

When a patient comes into the emergency department with a suspected stroke, the first thing to determine is whether the stroke is caused by a clot blocking blood flow to the brain tissue.

If this is the case, the next thing to figure out is which treatment is most likely to benefit the patient:

  • a procedure using a catheter to mechanically extract the clot;
  • a combination of the previous treatments – often the most suitable to restore perfusion;
  • doing nothing at all – sometimes the only course of action.

Nowadays, in stroke patients, neuroimaging can be used to confirm the presence of the clot, but it can give so much more1.

Related Article

With neuroimaging, it’s possible to analyze specific characteristics of the clot, such as location, size and composition, which have the potential to predict the treatment that has the highest chance of success.

While location and size can be intuitively evaluated from visual assessment of the scans, the composition is more difficult to assess, driving many researchers to investigate surrogate measurements of clot composition that can be derived from neuroimaging.

Clot perviousness: a new metric in stroke treatment

Some simple metrics compare the clot’s density to a healthy vessel on the opposite side of the brain2. Others use complex features to relate treatment success to the clot’s perceived texture3,4,5.

A metric that recently has gained much traction is perviousness.

Stroke AI algorithms will help physicians perform density measurements

Perviousness provides a good balance between being easy to measure and harnessing the information provided by images while remaining intuitive. It measures the increase of density (HU) of the clot on a contrast-enhanced scan with respect to a non-contrast scan.

Clots that are more permeable will appear brighter in a contrast-enhanced scan. Patients with pervious clots are more likely to respond better to treatment6.

Easier measurements with stroke AI algorithms

However, by requiring the measurements of the density values at anatomically correspondent locations on both scans, computing perviousness is a cumbersome and time-consuming process that does not fit the acute stroke environment. But here is where stroke AI algorithms play their role in supporting neuroimaging.

The stroke AI algorithms automatically align the images and delineate the clot, allowing physicians to perform the density measurements with ease and confidence7 in the acute stroke environment.

Using stroke AI algorithms to analyze clot characteristics will soon become an integral part of the triage8 of stroke patients.

Discover how you can use stroke AI algorithms in your job and what StrokeViewer can do for you and your stroke workflow.

Related articles

International Stroke Conference 2022 logo

News from ISC22

StrokeViewer's Collateral algorithm made its appearance at the International Stroke Conference 2022 (ISC22), held in New Orleans.