The idea that mind-reading might someday be possible is closer to reality thanks to a new study from D'Or Institute for Research and Education.

Using magnetic resonance imaging (MRI), researchers from the institute attempted to read the minds of participants to discover what song they were listening to.

Detailing their work in Scientific Reports, six participants listened to 40 pieces of music from genres such as classical, pop, rock and jazz. With each song, researchers captured via MRI the neural fingerprints on the participant’s brain. Simultaneously, a computer identified the brain patterns evoked by each song.

Researchers believed that the computer would be able to identify the songs participants were listening to based exclusively on participant brain activity (a process called brain decoding). In a scenario where the computer was presented with two choices, the computer correctly identified the song with 85 percent accuracy.

In a second scenario, where the computer was presented with 10 different choices, the rate of accuracy was 74 percent of the time.

Encouraged by this finding, researchers expect future brain decoding and machine learning studies to help achieve communication beyond written or spoken language.

"Machines will be able to translate our musical thoughts into songs," said Sebastian Hoefle, a researcher from D’Or Institute.

For more on the research, go to Scientific Reports.