Brain-computer interfaces (BCIs) are being developed to help paralyzed individuals engage more with life, such as by playing video games they can control with their minds. A team from Stanford and Michigan universities has successfully enabled a paralyzed person to control a quadcopter using a BCI. This study was peer-reviewed and published in the journal Nature Medicine.
BCIs are still not precise enough for many applications. In the United States alone, over five million people live with severe motor impairments. Many of these individuals turn to video games, which offer various ways to participate without using hands. BCIs are seen as the next step to allow more complex control in video games. However, many BCIs struggle with precision in individual finger movements.
The team, led by neurosurgeon Matthew Willsey, developed a BCI that reportedly achieves a higher level of movement precision and freedom than previously possible. The BCI, with its microelectrodes, was implanted in the precentral gyrus of a nearly completely paralyzed 69-year-old patient, referred to as T5. This part of the brain includes the motor cortex, responsible for controlling hand and finger movements.
Initially, the BCI needed training. The patient observed a virtual hand performing various movements and was asked to imagine making the same movements with his own hand. The corresponding neural signals were recorded and analyzed using machine learning algorithms, essentially artificial intelligence. This allowed researchers to identify which signals corresponded to which finger movements.
The difference from previous studies was that the BCI was not only designed for relatively simple motor tasks like mouse clicks or grabbing objects but also for representing complex finger movements in multidimensional space—movements needed in video games or to fly a quadcopter.
In subsequent tests, T5 was able to move three groups of fingers and the thumb of a virtual hand using only his thoughts. This virtual hand was linked to the remote control of a virtual quadcopter. With just his thumb, he could steer the quadcopter left, right, up, and down. The patient was able to navigate the quadcopter through a predetermined obstacle course and fly through randomly appearing rings. Overall, the system allowed navigation through or around 18 rings in less than three minutes, representing a sixfold improvement over previous systems. However, the system was not flawless; sometimes, different finger groups were activated simultaneously, complicating control.
The patient, who had specifically wished to fly a quadcopter again, described the experience as similar to playing a musical instrument: “Imagine playing a clarinet and then taking another person’s instrument. You immediately notice it’s different, but thanks to your experience, you adapt.” He noted that controlling the quadcopter was “more sensitive than fingers” and that he only needed to “nudge it in a direction.”
For Matthew Willsey and his colleagues, this research is a step toward involving paralyzed individuals more in recreational activities. Video games have become a type of litmus test for the capability of BCIs. For instance, Elon Musk’s startup Neuralink introduced a BCI last year that reportedly allowed a patient to play Counter-Strike. A team from the University of Texas in Austin demonstrated that brain waves needed for gaming could be analyzed non-invasively using EEG caps.