An international competition is showcasing the power of crowdsourcing by helping to launch a new method to detect, predict and prevent epileptic seizures.

Epilepsy is a brain disorder that affects roughly 1 percent of the population. It is characterized by seizures or episodes of disturbed brain activity that have been likened to electrical storms in the brain. The difficulty for sufferers is knowing when these seizures are going to hit and taking appropriate measures to prevent them.

Currently, the most popular method for treating epilepsy is medication, but the side effects of these drugs can leave people feeling tired or dizzy. Surgery is another option, as is a new type of implanted device that uses electrical pulses to prevent seizures. But until now, doctors were at a loss when it came to fine tuning these treatments so that they could be effective when a seizure strikes.

Thanks to a contest hosted by the online platform Kaggle, researchers have been able to use data on electrical activity in the brain to develop an algorithm that predicts seizures 82 percent of the time.

The contest by Kaggle — which calls itself The Home of Data Science — received entries from participants from all over the world with backgrounds in everything from mathematics to computer science to engineering. The winning team of five participants from the United States and Australia featured a software engineer and mathematician — but oddly enough, no doctor. More than 500 teams competed in two challenges, one for seizure detection and another for seizure prediction. 

The two challenges were sponsored by the American Epilepsy Society, the National Institutes of Health's National Institute of Neurological Disorders and Stroke and the Epilepsy Foundation.

The contest has helped advance the treatment of patients with epilepsy by giving health care providers the tools they need to better predict seizures. And it has proven the value of crowdsourcing collaboration with experts from a variety of fields when it comes to solving tough problems. 

Related on MNN: