Scientists from the California Institute of Technology and the National Institute of Health and Medical Research (Inserm) France have achieved a significant breakthrough in brain-machine interfaces. They have successfully demonstrated the potential of using non-invasive functional ultrasound (FUS) technology to create a real-time computer control system based on thoughts.
This development is based on an innovative method of scanning brain activity using ultrasound, eliminating the need for surgical implantation of devices into the brain. This reduces health risks associated with traditional interfaces.
FUS technology allows for the capture of three-dimensional images of neural network activity in different areas of the brain using short ultrasound pulses directed through a specialized “window” in the skull.
In their experiments, researchers scanned a region of the brain responsible for planning movements in primates. The obtained data was processed by a decoder trained with machine learning, which converted it into commands to control a computer cursor.
The results showed that using FUS, monkeys could accurately position the cursor on the screen almost in real time, simply by “thinking” about moving their eyes or hands. The system did not require calibration and remained stable over a long period.
“Functional ultrasound is a completely new tool in the arsenal of brain-computer interfaces that can provide assistance to individuals with paralysis,” commented Professor Richard Andersen, one of the study’s authors.
The scientists believe that their breakthrough paves the way for the development of fully non-invasive computer control systems and even thought-controlled robotic prostheses. This advancement has the potential to greatly improve the quality of life for paralyzed patients.