Image above: A student in Adrian Owen’s lab places an EEG cap on another student.
Imagine sitting at the computer and singing David Bowie’s “Let’s Dance” in your head, then waiting as your computer reads the electrical activity in your brain and plays the song.
Koerner Fellow Adrian Owen (Western University) has imagined this and more. Owen, the co-director of the Azrieli Program in Brain, Mind & Consciousness at CIFAR, is pursuing an emerging line of research known as music imagery information retrieval (MIIR).
But he’s less interested in helping people access songs, and more interested in how computer interfaces that combine music and recordings of electrical activity in the brain could help patients with severe brain injuries communicate. This could work in many ways. Owen gives the example of a wheelchair.
“If you were a paralyzed patient and you were operating your own wheelchair, for example, you would sing one song in your head to make it go left and a different song in your head to make it go right,” Owen says.
Researchers must resolve many challenges for this vision of the future to become reality. As a step forward, Owen’s lab has created an open access dataset for researchers all over the world to collaborate on advancing the field. OpenMIIR is a compilation of electroencephalography (EEG) recordings taken while people were either listening to music or singing a song in their heads. Western University has some of the most sophisticated EEG equipment in the world, which allows for high quality recordings.
So far the database contains recordings from 10 people listening to music from different genres, at different tempos, with and without lyrics. The recordings have revealed that when someone imagines a song, the electrical patterns in their brain look relatively similar to when they are actually hearing the song.
“We obviously need to go much further than that because I want to be able to just work out what you’re singing in your head right now,” Owen says. He says his team is focusing on music because humans are very good at imagining songs. We do it naturally, sometimes even when we don’t want to, if there is a tune stuck in our heads.
His team has had success asking subjects to imagine other things. In a remarkable breakthrough reported in 2010, he communicated with a vegetative patient using functional magnetic resonance imaging (fMRI) for the first time by asking him to imagine playing tennis. But tennis doesn’t work as well with EEG recordings, which might be a more promising technology. MRI machines are not portable and they are very expensive.
“We’ve been working hard to try and come up with an EEG-based method for both identifying consciousness in these patients and for communicating with some of them,” Owen says.
He says the question requires the efforts of scientists in many disciplines, from neuroscientists to experts in information retrieval, mathematicians, engineers and more. Ethicists, lawyers and spiritual leaders are important for grappling with what to do when vegetative patients tell us what they want.
“It’s multidisciplinary like you couldn’t imagine.” He says that’s why he approached CIFAR to create a new research program, now named the Azrieli Program in Brain, Mind & Consciousness, to sustain long-term collaboration between experts in many fields on big questions around consciousness.