Lunghammer/TU-Graz) imagines that he is "locked" by a neurodegenerative disease, such as advanced ALS (Lou Gehrig's disease), or completely paralyzed by traumatic spinal cord injury. You can still think, dream and feel emotions, but you can't express them. For decades, researchers have been experimenting with brain-computer interface, or BCI for short, so that people with severe disabilities can only communicate with their brains.

KDSP In recent years, BCI technology has enabled people with disabilities to write information, send emails, surf the Internet, control smart homes and even move electric wheelchairs. In 20 10, a German research team used BCI to enable ALS patients to make the first "brain painting", which effectively released the creativity of paralyzed artists. Now, a team of neuroengineers in Austria has built the first one, which is made entirely by the brain. According to a paper published in the journal Public Library of Science One, the researchers tested the thinking machine technology with a group of healthy volunteers, who can imitate music melody and accurately create original music. The director of the Institute of Neuroengineering at Graz University of Technology in Austria said that he was deeply encouraged by the success of the brain painting project and the satisfaction it brought to ALS patients, many of whom were artists.

"They hold exhibitions." They sell their own works, "Mueller Putz told The Explorer. Why not try music and composition? "and

Related: using brain phones

It is easy to make music with only the brain. The latest BCI technology is different from other interfaces, allowing you to really play music with your mind. This technology does not need to use brain waves to calibrate the tone, but can perform any operation through a series of individual characters or commands.

This technique can be used to spell word for word, or to choose the size and color of the brush. Similarly, Mueller Putz believes that it can be used to select a specific type of note (for example, a whole note or a quarter note) and the tone of the note to write a musical score.

The technology behind BCI has existed for 20 years and is based on a particularly strong brain signal called P300 event-related potential. Brain signals are recorded by EEG worn on the head, just like wearing a swimming cap. Because everyone's brain is a little different, it is necessary to calibrate the system to accurately read the subject's brain.

Calibration is accomplished by having the subjects stare at a computer screen displaying a 6×6 letter grid. When the subjects stared at the screen, a series of random flashes lit up every row and column in the grid. Every flash, the subject's brain will send out a specific waveform. According to the position of the flash on the grid, each signal sent by the brain is slightly different.

Once the calibration is over, subjects should try to pick a single character with their brains. The grid continues to blink randomly, just like before. Muller Putz said that the trick is to focus on the characters you want and count them silently whenever the rows or columns of the characters flash.

"The computer algorithm behind BCI will find out where different waveforms appear and which row and column it is in," and then identify the characters selected by users, "said Muller Putz, who has been working with BCIs since 1999.

Choosing a single character with your mind is a slow process, but it is easy to learn. In Mueller Putz's experiment, a group of BCI novices were able to copy spelling words with an accuracy of over 88% in the first attempt. Their success rate in copying this score is similar to that of the popular folk song Alouette, although it took them an average of 265,438+0 minutes to pick out 25 notes of the melody.

Related: This brain implant makes telepathic typing possible.

The researchers also invited a professional musician to try to compose music with his brain. He achieved a higher accuracy (93.6%) in the copying exercise using "Alouette", but it still takes 20 minutes to complete. He is proficient in the free composition part of the experiment. In the experiment, he can write a melody with 26 notes in only 14 minutes, which is twice the average composing speed of a non-professional musician in 30 minutes 18 notes.

Speed is not a big problem, Miller Putz explained, because users of this technology are trying to help "don't live in a hurry like us". Their views are a little different.

The team's preliminary experiment proves that BCI based on P300 can be used for music creation. The next step is to try with the disabled and collect their feedback. Muller Putz said that the long-term goal is to develop from a laptop-based system to a system small enough to be installed on a smartphone. This is the best way to send it to families and musicians who want to share their songs with the world.

Originally published in explorer.