Despite brain being a complex organ, Neuroscientists from Carnegie Mellon University have learned a way to decode our thoughts. It’s a pretty big deal to read mind. Isn’t it? Scientists though haven’t learned to practice telepathy completely. Our exact thoughts can’t be traced yet, but the types of thoughts our brain process can be scanned.
The objective behind this is not to uselessly provide humans the power of mind-reading, but to help people, who are unable to communicate due to some medical reasons. For example — people in a coma or the ones suffering from paralysis.
Though science has progressed a lot, still mind-reading isn’t an easy task since neural responses differ from person to person. However, techniques and procedures like functional magnetic resonance imaging, electroencephalography, and lesion studies have helped in interpreting mind to an extent.
For this research where scientists have come a step closer to reading minds, a newly designed algorithm is used. When we communicate, sound waves are produced which work on a specific set of nerve cells. These nerve cells then transmit related sensory information to the brain which translates it into words. But speaking out loud and speaking inside our brains activate similar neurons?
A new research found out brain patterns that form sentences in our head. To study the pattern, seven volunteers were made to work with an algorithm of 239 sentences. They were then asked to decode the 240th thought. And for every single time, there was a different sentence decoded.
Scientists pointed that algorithm could predict how brain put together complex thoughts and how complex thoughts are put together in a brain. The process helped in identifying the building blocks of our thoughts which was never done before. Further, researchers reported that these building blocks were same for all individuals.
Simple brain patterns like yes or no was possible to examine with our earlier scanning technologies. But this latest research is a step ahead in finding more complex thought process. Their model was able to predict characters with 87% accuracy.
The lead of the research Marcel Just says, “A next step might be to decode the general type of topic a person is thinking about, such as geology or skateboarding. We are on the way to making a map of all the types of knowledge in the brain.”