November 15, 2012
Researchers Make Music From Brain Waves
Brett Smith for redOrbit.com — Your Universe Online
Chinese researchers at the University of Electronic Science and Technology in Chengdu, China have developed a method to transform brain waves into music that closely resembles something a human composer would write.
According to their report in the open access journal PLOS ONE, they were able to use an electroencephalogram (EEG) to create the pitch and duration of a note, and functional magnetic resonance imaging (fMRI) to control its intensity.
The music created by the remixing of these brain waves "embodies the workings of the brain as art, providing a platform for scientists and artists to work together to better understand the links between music and the human brain,” the researchers wrote in their report.
At first, the researchers used only the EEG to create music based on a series of algorithms. However, the intensity of the music changed quickly and abruptly, making it very different from man-played music. To correct for the rapid changes in note intensity, the researchers augmented the information from their EEG signal with information based on an accompanying fMRI signal, which they used to modulate the intensity of the music.
In their report, the authors said they used a separate signal because the pitch and intensity of notes in music are independent from each other and the EEG music they created had linked the two factors together.
The researchers also needed to find a way to creat music that made sense, as some algorithms could have turned the brain wave signals into a hyper-jumbled mess.
To produce somewhat recognizable sounds that were still based in physiology, the scientists decided to accept that their music would have to be scale-free. Most scales have less than 12 notes and placing this metric against the brain wave signal could be construed as misrepresentative.
The scientists also had to grapple with the temporal and spatial activities of the brain. The fMRI tends to have a higher spatial resolution for brain signals, while the EEG collects information on the surface of the brain and therefore is better at reflecting the brain´s temporal activities.
By combining EEG and fMRI data, the study´s authors said the music they produce better reflects the functional activity of the human brain. They said this could lead to future improvements for clinical diagnosis.
Scientists with a flair for the artistic have been trying to derive music from naturally occurring sources for decades. In 1934, Edgar Adrian and B.H.C. Matthews used an amplified speaker in an attempt to hear brain waves directly.
Ever since the invention of the EEG, scientists, probably noticing a similarity between brain waves and sound waves, have been trying to use the machine to create music, but the principles behind creating it and the output of an EEG did not make for an easy translation. Starting in the 1990s, rules and algorithms were created to translate the output of an EEG into music.
Scientists from other disciplines have also tried to make music using natural sources, including DNA, proteins, and electromyograms (EMGs).