New Algorithm Helps Predict When Seizures Will Occur In Epileptic Patients

Statistics show that 2.3 million adults in the United States alone are affected by seizure disorder; that's 1 in 26 people who will be diagnosed with the health issue--otherwise known as epilepsy--at some point in their lives. Knowing when a seizure will occur is half the battle.

Now, a new algorithm has been created to help predict when the next episode could occur as a product of a crowdsourcing competition.

Researchers said they believe that it can help predict the onset of a seizure with an 82 percent success rate, according to the National Institutes of Health (NIH). Before this, the best algorithum at predicting a related health issue was simply flipping a coin.

Furthermore, there was no specific accepted treatment for an individual epilepsy center as many followed differing diagnostic and therapeutic guidelines.
Previous studies have examined two specific algorithms that guide decisions in the treatment of refractory partial epilepsy, describing stepwise diagnostic and therapeutic approaches for intractable media temporal and neocortical epilepsy.

Thank the American Epilepsy Society Seizure Detection Challenge for coming up with the new algorithm last August, via Kaggle.com--a well-known online platform for data prediction competition that's co-sponsored by the American Epilepsy Society, NIH's National Institute of Neurological Disorders and Stroke (NINDS), and the Epilepsy Foundation.

The challenge was based on two contests that were comprised of more than 500 teams from around the world, based on both detection and predictions of seizures.

Michael Hills, a computer engineer from Australia, won the prediction contest, and won the winning algorithm by classifying various aspects of localized electrical field potential in the brain.

Epilepsy can result from a host of medical conditions that affect an individual's brain, including stroke, brain tumor, traumatic brain injury (TBI) or central nervous system infection. Fortunately, regular check-ups, therapy and medication can help keep related health issues at pay, depending on the severity of the problem.

For more great nature science stories and general news, please visit our sister site, Headlines and Global News (HNGN).