In a novel study presented at the Society for Neuroscience’s annual meeting (Washington DC, USA; 11–15 November 2017), researchers have demonstrated that if a brain-machine interface (BMI) makes an error, the human brain is able to quickly adapt to correct the mistakes.
BMIs are devices that translate neuronal information into commands capable of controlling external software. For patients who are no longer able to move their limbs, these devices offer a chance to partially regain lost motor and communication abilities.
In this study a female patient with tetraplegia was implanted with a 96-electrode array in her anterior interparietal cortex, a region of the brain involved in planning movement. In order to examine how the anterior interparietal cortex learns to compensate for errors the researchers assigned the woman a 2D cursor control task. Once she was able to control a cursor using her thoughts alone, the researchers introduced perturbations to the BMI.
The perturbation was successful in causing an immediate and lasting decrement in performance. However, over the course of a session, it was observed that the participant could regain proficient control of the cursor for easy (short-distance) perturbations. For more complex long-distance perturbations, which produced more severe errors in cursor movement, however, she was unable to overcome the introduced impairments in cursor control.
Of note, the participant reported that she compensated for errors caused by the perturbations by re-aiming to different directions, this suggest that individuals adapt to perturbations in BMI tasks by exploring new cognitive strategies.
“If a brain area is more flexible than others and can generate more combinations of innate activity patterns, then it may be a better region for collecting neural activity to control BMI devices,” explained lead author Sofia Sakellaridi from the California Institute of Technology (CA, USA). “Understanding the limits of learning is important for rehabilitation therapies for individuals who suffer from stroke or other brain injuries. Restoring neural functions after brain trauma may also require patients to generate specific neural patterns of activity.”
Next, the researchers intend to assess the plasticity of other brain regions, such as primary motor cortex and premotor cortex, during the learning of BMI use and following the introduction of errors to BMI mapping.
Sources: Society for Neuroscience