Print

Researching Brain Music At Trinity College

WILLIAM WEIR

August 10, 2009

Music's effect on the brain has been the subject of a good deal of neuroscience research in recent years. Trinity College's Dan Lloyd has gone further by actually making music out of brains.

It started as a hobby two summers ago for Lloyd, who teaches in the philosophy department at Trinity, when he wondered: What does consciousness sound like?

So he developed a software program that assigns pitches to different regions of the brain. When a region becomes active in functional MRI brain scans, a note is sounded. If a few brain regions activate at once, a chord is struck. How loud the tones are depends on the degree of brain activity.

More than just an intriguing idea, Lloyd's brain music could also serve a practical use by providing one more way to identify disorders. The brains of patients with schizophrenia, for instance, sound notably different from healthy brains.

"I'm interested in brains and consciousness and the lack of scientific knowledge to account for it," he said.

He has posted a few of the results on YouTube, where viewers can see the brain scans and the sounds they make. Lloyd is also working with a harpist in Finland who plans to transcribe some of the brain scans to her own instrument, and a video artist plans to use Lloyd's work for a sound installation in Austria.

The project highlights one fact, he said: Consciousness "is a lot of things happening at once." He figured that if he could translate all those things into sound, there would be a new kind of data — a polyphonic one — for researchers to analyze.

"The ears can do that in ways that the eyes can't," he said.

Some of the scans are of Lloyd's own brain. Others are of his students who volunteered at Hartford Hospital's Institute of Living for studies on drunken driving. Volunteers entered an MRI machine that contained a virtual driving game with a simulated steering wheel.

Watching one volunteer's brain scan on his office computer at Trinity, Lloyd points to a dramatic change in pitch in the series of notes. That is a point in the game when the subject stopped driving. Theoretically, he said, the program could become so sophisticated that a researcher would know all the subject's actions just by listening.

From what we can hear, brains produce a pleasant sound, similar to the ambient music of Brian Eno. Most of the sounds are either electronic tones or MIDI (musical instrument digital interface) piano. Although the notes don't form a conventional melody, they have a musical quality to them. The pitches he assigns are based on various musical scales. Some of the results sound a little like the work of Philip Glass, and that is no coincidence — he is Lloyd's favorite composer.

Although Lloyd is trained in philosophy, he knows his way around neuroscience. Nonetheless, the hard-science aspect of the project benefited from his collaboration with Vincent Calhoun, who was at the Institute of Living when the project began.

Now the chief technology officer for the Mind Research Network in New Mexico, Calhoun says that listening to brains could open a new door in neuroscience.Besides brain scans of Lloyd and his students, the project also includes scans of patients with schizophrenia made in separate studies at the Institute of Living.

"One of the things we found with schizophrenia was that in one particular brain, there was more rapid fluctuation in the way that different brain areas are communicating with one another," Calhoun said.

The difference, he said, is "striking" between schizophrenic patients and the control subjects. Visual data doesn't always make the differences between healthy and unhealthy brains so clear."It's one thing to look at the data in a picture and another to hear it," Calhoun said. "If you have a functional MRI image data set — there's a lot [of information] in there but it's not obvious how to get it out. All these cognitive events are overlapping with one another."

There is no definitive way to know if someone will become schizophrenic, Calhoun said, so if the aural data serves in any way to predict the disorder's onset, it could have a big effect on research and treatment.

Whether it serves as a reliable tool for doctors, though, will depend on more research."It's really going to depend on whether we can show there's a consistent difference," he said. "We need to quantify what it is about the musical output that's giving us more information."