Doctors may soon use technology similar to that in airport security and passport control to help better predict if stroke patients will have a negative reaction to the popular intravenous thrombolysis treatment. Although the treatment is meant to save the lives of those affected with blood clots, in about six percent of cases it worsens the patient's condition by causing bleeding in the skull and even death. Doctors also hope this software will help identify stroke patients that may benefit from the treatment but are not currently offered it.
During intravenous thrombolysis treatment, a chemical is injected into the blood vessels to break up or “bust” the clots. This allows the blood to flow again. According to the press release, the treatment also thins the blood, which is what causes the negative reaction of bleeding in the brain in a small percentage of patients. Doctors try to predict patients most at risk of bleeding by using brain scans. The risk signs are often very small and hard to be picked up by human judgment. This severely affects the test's accuracy and reliability.
In a new study, researchers were able to successfully train a computer program to recognize patterns in the brain scans that represented high-risk patients for thrombolysis treatment. These signs included brain-thinning or small-vessel narrowing. The computer program was able to successfully predict brain bleeding caused by the intravenous thrombolysis treatment with 74 percent accuracy, in comparison to 63 percent for the standard prognostic approach.
This is great news for patients who suffer from strokes and could potentially make the difference between living or dying from their treatment. “For each patient that doctors see, they have to weight up whether the benefits of a treatment will outweigh the risks of side effects. Intravenous thrombolysis carries the risk of very severe side effects for a small proportion of patients, so having the best possible information on which to base our decisions is vital,” Dr. Paul Bentley, lead author of the study, explained in the press release.
Strokes affect over 15 million people every year worldwide, with the majority of these patients suffering from ischemic strokes. These occur when a small clot interrupts the blood supply to the brain. Researchers are also interested in discovering if this same technique could be used to identify stoke patients who might be helped by intravenous thrombolysis but are not currently offered the treatment because of pre-existing conditions.
Source: Bentley P, Ganesalingam J, Jones ALC, et al. Prediction of stroke thrombolysis outcome using CT brain machine learning. NeuroImage:Clinical. 2014