How Do We Hear While We Sleep?

Using electrodes implanted directly on the human cortex, a Johns HopkinsUniversity undergraduate has located the part of the brain thatappears to process sounds while people sleep. This site, in thefrontal lobe, may be part of a vigilance system that, for instance, rouses amother when her baby cries but lets the woman sleep when a truckrumbles by.

Serena J. Gondek, a 21-year-old junior majoring in biomedicalengineering, is slated to present her findings on Tuesday, April28, at the annual meeting of the American Academy of Neurology inMinneapolis, Minn. About 8,000 people are expected to attend theworld's largest gathering of neurologists and neuroscienceprofessionals. Administrators at the academy said it is unusualfor an undergraduate to be chosen to make an oral researchpresentation at the event.

Previous studies on hearing during sleep have relied onelectrodes attached to a subject's shaved scalp. Gondek'sexperiment is believed to be the first of its type to useelectrodes implanted directly on the brain, a technique thatyields far more precise information about which parts areactivated by sounds during sleep.

Gondek is from the Chicago suburb of Oak Brook, Ill. Sheconducted her experiment under the supervision of Gregory L.Krauss, an assistant professor of neurology at the Johns HopkinsSchool of Medicine. Krauss is co-author of the paper, titled, "Howdo we hear while we sleep?"

"It is controversial how we monitor our environment while wesleep," says Krauss. "It's a pretty big part of our lives, butsleep is poorly understood. The main thing that Serena did was toshow where on the cortex we hear while we sleep. She did aterrific job, particularly in the brain mapping."

Gondek is the second Hopkins undergraduate in recent months topresent research findings at a major medical conference incollaboration with Krauss. "Undergraduates are great for thesekind of projects because they're very enthusiastic," says Krauss,"and they work very hard."

Hopkins professors often encourage undergraduates to take part inhigh-level scientific studies. "That's why I really lookedforward to coming to Hopkins," says Gondek, "It's been as easy asknocking on a few doors to become involved in research projects,even during my freshman year."

For her experiment, Gondek obtained the cooperation of fivepatients who were about to undergo brain surgery to curtailepileptic seizures. To find the focal points of these seizures, asurgeon had cut open the patients' skulls and implanted electrodegrids directly on their brains.

Prior to their surgery, Gondek placed special plugs in thepatients' ears. The plugs blocked out room noise but allowed thepatients to hear sequences of two tones emitted by Gondek'sequipment, one tone pitched at 500 hertz, the other at 1,000.

She played various tone patterns while the patients were awake,during light sleep and during deep sleep. The electrodes detectedwhich portions of the brain were activated. Gondek analyzed theresults and mapped them onto MRI and CT scans made of thepatients' brains.

"We found that during waking, only areas around primary auditorycortex are activated by the tones," she says. "Then, during lightand deep sleep, you find not only primary auditory activation,but the frontal lobe also responds."

The frontal lobe is believed to play a key role in vigilancefunctions, such as screening new stimuli and preparing the bodyto react. During sleep, Gondek speculates, this part of the brainmay analyze sounds to decide whether the person needs to beawakened to respond. This mechanism would allow a camper in thewoods to sleep through non-threatening cricket chirps. But itmight awaken the camper quickly to the growl of a bear.

Gondek and Krauss are planning follow-up experiments to learnmore about how the brain processes specific environmental soundsduring sleep. Their research may also shed light on connectionsbetween sleep disorders and mental illnesses such as depressionand dementia.

Finding the place where the brain processes sounds during sleephas been a critical first step to cracking a tough physiologicalpuzzle. "At this point," Gondek explains, "we can say that thisis where it's happening. But it's still very unclear what's goingon there, how this processing works."

Her project was supported by a General Electric FoundationUndergraduate Engineering Research Stipend.

###

Reporters, please note: Color slides are available. Contact Phil Sneiderman atthe phone number or e-mail address at the top of this release.