Scientists hear the music of your brain

Science and art often go hand in hand. M.C. Escher played with geometry in his famous prints of impossible realities, and anatomist Gunther von Hagenst captivated millions of people with a display of preserved human corpses and body parts. More recently, a popular installation at the London Barbican Gallery offers art lovers the unique experience of walking in the rain without getting wet.

Science lets art push boundaries, and so is increasingly used by modern artists to shock and awe. But this is a two way street.

Researchers from the University of Electronic Science and Technology in China have now used recordings from brain scans to create music, and they hope that listening to the brain will give new insights into how it works.

This unusual way of blending science and art was first introduced in 1965 by the American composer Alvin Lucier in the piece Music for Solo Performer, but it was not until the 1990s that 'brainwave music' boomed with the development of powerful computers. Scientists and musicians now use computational models to convert data from brain scan technology into music played by electronic synthesizers.

In the new study, published in PLoS ONE, Jing Lu and colleagues developed a new method to combine readings from electroencephalograms (EEGs) and functional MRI (fMRI) brain scans to produce music that better mimics contemporary classical music. The amplitude and the frequency of the EEG signal were used to create the pitch and the duration of each musical note. This was then remixed with an fMRI signal, which set the intensity of the notes played.

The authors acknowledge that the sources of the two signals are, however, unrelated. "There is something a little arbitrary about putting EEG and fMRI together in this way" says philosopher Dan Lloyd from Trinity College, Connecticut, who was the first person to create music from fMRI scans. "But you have to start somewhere, learning scales before you play a sonata, and this is a good start" he adds.

EEGs measure the electrical activity of the brain. Brain cells, or neurons, communicate with one another through electrical signals. During an EEG, electrodes are attached to the scalp and plugged into a computer that converts the electrical signals to waves. EEGs are used to diagnose epilepsy and sleep disorders for instance.

fMRI, or functional magnetic resonance imaging, on the other hand measures brain activity by detecting changes in blood flow, and is mostly used in research. Neurons need oxygen to make energy, so when a brain area is more active, oxygen-rich blood surges. The pattern of brain activity across the brain is then represented in a color code.

The authors of the study plan to improve their EEG-fMRI method so it may be used for clinical diagnosis, for example, if the music could produce audible differences between healthy and sick brains. Lloyd says

"With the proper sonification [conversion to sound], something like a 'brain stethoscope' could be developed as a clinical tool for detecting clues to a variety of brain conditions".

But there are skeptics. David Sulzer, a neurophysiologist at Columbia University and jazz musician, thinks that brainwave music can only detect very significant changes in brain activity, such as an epileptic seizure. He says "If you cannot diagnose an illness through a chart recorder readout, I do not understand how it can be done sonically".

Sulzer believes music made from brain activity is an art and a science didactic tool. For instance, he uses The Brainwave Music Project to teach the public about brain function before his performances of brainwave music.

Whether brainwave music will be useful for science or medicine remains an open question, but it can surely be said that it has become an art form of its own. It might not be for everyone's taste but brainwave music certainly causes an impression, which is the very definition of modern art.

For the complete article, and to hear some samples of "brain music", please go to:

http://www.plosone.org/article/info:doi/10.1371/journal.pone.0049773