Epilepsy affects about 50 million people worldwide, and while anticonvulsant medications can reduce the frequency of seizures, the drugs are ineffective for as many as one in three patients.

The new treatment builds on an existing treatment for epilepsy, the Cyberonics Inc. vagus nerve stimulator (VNS), which is often used in patients who do not respond to drugs. A defibrillator typically implanted under the patient's collar bone stimulates the left vagus nerve about every five minutes, which has been shown to help reduce the frequency and severity of seizures in many patients.

The MIT researchers and colleagues at Beth Israel Deaconess Medical Center (BIDMC) seek to improve the treatment by combining it with a detector that measures brain activity to predict when a seizure is about to occur. The new device would sense the oncoming seizure and then activate the VNS, in principle halting the seizure before it becomes manifest.

"Our contribution is the software that decides when to turn the stimulator on," said John Guttag, MIT's Dugald C. Jackson Professor in the Department of Electrical Engineering and Computer Science. Guttag developed the system along with Ali Shoeb, a graduate student in the Harvard-MIT Division of Health Sciences and Technology.

"Our colleague Dr. Steven Schachter, professor of neurology at Harvard Medical School and epileptologist at BIDMC, suggested hooking our detector up to the VNS," he said. MIT and BIDMC researchers plan to test the new device in epilepsy patients this fall. If it seems effective, more comprehensive trials will be launched.

A look at brain patterns

The detector works by measuring brain activity with electrodes placed on the patient's scalp. In its current form, the patient wears something resembling a bathing cap, in which electrodes are embedded. In order to adapt the detector to work with the VNS, researchers connected wires from the cap to a laptop computer or microprocessor that activates the implanted defibrillator.

Guttag said he believes the technology could be refined so the electrodes could be worn inside of a headband or baseball cap, making the device less obvious to observers.

Each epilepsy patient has different brain activity patterns, so the detector is programmed to measure an individual's patterns to determine what the precursors to a seizure look like for each patient.

"It's quite tricky to try to detect very early signs of seizures because there are abnormal electrical signals that don't evolve into seizures," Guttag said. "If we can learn what the right profile is for an individual, we can build a seizure onset detector that works really well for that person."

Ideally, when the device senses an impending seizure, it sends a magnetic signal to the implanted stimulator, which in turn activates the left vagus nerve. The vagus nerve sends electrical signals up to the brain as well as down toward the viscera, controlling heart rate, gastrointestinal peristalsis, sweating and keeping the larynx open for breathing. The mechanism by which VNS prevents seizures is not known, but the technique has been FDA approved to treat epilepsy for about 10 years.

About 20,000 epilepsy patients already have VNS implants, according to Guttag. Some of them are able to use a handheld magnet to activate the VNS on demand, but many cannot. If the new detection device is successful, it would allow many more patients to use the VNS on demand.

The device could also be adapted to provide warnings for patients who don't need or want VNS implants. Once the device alerts the patient that a seizure is imminent, that person could take steps to minimize injury, such as sitting down or moving away from potentially dangerous objects, such as a hot stove.

"If you could just give someone a little bit of warning they're about to have a seizure, it could be hugely valuable," Guttag said. "The seizures themselves aren't usually damaging to the brain in the long term. It's mostly about the collateral damage."

Although the seizure detector could have a huge impact on epilepsy patients, there are plenty of other potential applications for technology that analyzes electrical activity in individual brains, Guttag said. Depression, schizophrenia and attention deficit disorder are just a few of the conditions that could be studied.

"My hope is that we'll be able to use some of the technology to get insight into a lot of those mysterious neurological conditions," he said.

A paper describing the seizure-detection technology was published in the August 2004 issue of Epilepsy & Behavior.

Other researchers involved in the project are Blaise Bourgeois, a neurologist at Children's Hospital, and Ted Treves, chief of nuclear medicine at Children's Hospital and professor of radiology at Harvard Medical School.

Source: MIT