Fast brain waves: A better biomarker for epilepsy

EEG and MEG detection of HFOs, fast brain waves associated with epilepsy
Localization of fast brain waves, called HFOs, with scalp EEG (left) and MEG (right). HFOs present a new biomarker for areas of the brain responsible for epileptic seizures.

In the U.S., about one in 100 people have some form of epilepsy. A third of those people have seizures that cannot be controlled with drugs, eventually requiring surgery to remove the area of their brain tissue that is triggering seizure activity.

“If you can identify and surgically remove the entire epileptogenic zone, you will have a patient who is seizure-free,” says Christos Papadelis, PhD, who leads the Boston Children’s Brain Dynamics Laboratory in the Division of Newborn Medicine and is an assistant professor in pediatrics at Harvard Medical School.

Even experts in this field were skeptical for years about the non-invasive detection of HFOs. But now, thanks to our study and other researchers’ work, these people are changing their minds. At present, however, these surgeries are not always successful. Current diagnostics lack the ability to determine precisely which parts of an individual’s brain are inducing his or her seizures, called the epileptogenic zone. In addition, robust biomarkers for the epileptogenic zone have been poorly established.

But now, a team at Boston Children’s Hospital is doing research to improve pre-surgical pinpointing of the brain’s epileptogenic zone. They are using a newly-established biomarker for epilepsy — fast brain waves called high-frequency oscillations (HFOs) — that can be detected non-invasively using scalp electroencephalography (EEG) and magnetoencephalography (MEG).

An overriding need for better diagnostics

Related story: What if seizure origins could be mapped without having to actually observe a seizure? An economic forecasting algorithm could make this possible.

Conventionally, the epileptogenic zone has been approximated based on a variety of non-invasive diagnostic tests, which can sometimes be inconclusive.

In these cases, surgeons must open the skull to insert intracranial electrodes. But intracranial recordings are costly and carry some risk of infection, bleeding and, rarely, neurological damage. What’s more, intracranial recordings require access to the brain’s surface, therefore it would not be feasible to access all surface areas of someone’s brain so invasively. Thus, it’s possible to miss the area that generates the seizures, leading to a poor roadmap for brain surgery.

More recently, studies have shown better surgical outcomes through removal of brain tissue that specifically generates HFOs. Yet for a long time, HFOs have been hard to detect using non-invasive methods. Now, that’s finally changing.

Getting more precise, less invasive

MEG sensors showing HFOs, responsible for epilepsy
Schematic showing placement of MEG sensors and the HFO area localized in the cortex of the patient.

“We now know that we can detect and localize HFOs using non-invasive techniques, such as scalp EEG and MEG,” says Papadelis.”This has great potential to improve the outcome for epilepsy patients undergoing surgery.”

“Using scalp EEG and MEG data, our team has developed a protocol and an automated algorithm that allows the reliable non-invasive detection and precise localization of HFOs, which we’ve been able to confirm by comparing with data from invasive intracranial recordings,” says Eleonora Tamilia, PhD, a research fellow in pediatrics in the Brain Dynamics Lab.

The team recently published a paper in Frontiers in Neurology about their findings. The project is funded by the American Epilepsy Society and the Office of Faculty Development.