neuromorphogenesis:

Brain stethoscope will use music to help identify epileptic seizures

An artistic project developed by two researchers at Stanford University has turned into a quest to build a biofeedback tool that will help identify epileptic seizures through music.

Inspired by a performance based on radio signals from outer space, Stanford neurologist Josef Parvizi embarked on a project to discover what the brain’s electrical activity would sound like set to music.

He enlisted the help of Chris Chafe, a professor of music research, and passed on to him the electroencephalogram (EEG) recording data of a consenting patient. From there Chafe, who has experience in converting natural signals into electronic music, set the electrical spikes of the rapidly firing neurons against music with human voice-like tones.

When the pair listened back to the music, they realised they had gone beyond creating something artistic and had by chance stumbled upon a way to clearly differentiate seizure activity in the brain from non-seizure activity. The two states were clearly discernible from the change in the music, says Chafe. “It was like turning a radio dial from a static-filled station to a clear one.”

The results don’t necessarily make for particularly easy listening (as you can hear in the embedded video), but it’s easy to distinguish the distinct phases of a seizure as the music changes throughout the piece.

The first, fairly rhythmic sounds represent the brain activity during the pre-ictal stage, which is before the seizure begins. Just before the seizure occurs, the sounds become increasingly louder, more frenzied and more unpredictable, peaking as the seizure takes place and the brain enters the ictal state. Suddenly the chaos subsides and the high frequency noises die down as the seizure tails off and the brain enters the fatigued, post-ictal state.

Caregivers to people with epilepsy often struggle to identify when a seizure is occurring or when one might be about to occur, but Chafe and Parvizi concluded that if they could work out how to achieve the same result using real-time brain activity data, it might be possible to develop a tool — a brain stethoscope — that would be able to tell.

The pair is currently developing the tool, which could be used for listening for seizures or for distinguishing when the brain is in a post-ictal state.

“Someone — perhaps a mother caring for a child — who hasn’t received training in interpreting visual EEGs can hear the seizure rhythms and easily appreciate that there is a pathological brain phenomenon taking place,” says Parvizi.

A prototype of the stethoscope, which will consist of a headset that will transmit an EEG of the wearer’s brain activity to a handheld device, is due to go on display at Stanford next year.