An international team has created a wheelchair capable of transform its user’s thoughts into commands and, in this way, to allow them to gain mobility and with training to function safely in a natural environment with the usual obstacles.
It sounds like science fiction, but this is how it is reported in an article published in the specialized media iScience. In fact, no less than three patients received training in the use of this chair and two of them achieved good results.
Man and machine learning
Specifically, the authors provided these patients with three training sessions every week for between two and five months in total.
These sessions, they emphasize, are not only important for the wheelchair users themselves, but also for the algorithm that uses the brain-machine interface itself. In fact, they stress that this research reveals a potential method to improve the clinical translation of non-invasive brain-machine interface technology.
To move the chair, patients put on a helmet that detects their brain activity using electroencephalography (EEG). The interface then transform this brain activity in mechanical commands for the wheelchair.
Chair users control the direction of the chair by thinking about move your own body. For example, they have to think about moving both hands to turn to the left and both feet to turn to the right.
Initially, all three participants had similar accuracy, between 43 and 55%. With the passage of time, two of them came to reach accuracies greater than 95%. At the same time, the machine was learning: parallel to the improvement in the patients, the algorithm improved its discrimination of brain patterns.
The remaining patient, for his part, did not show great improvements in precision and his brain activity patterns did not change, which was the case with the two who improved their handling of the wheelchair. This makes researchers think that in order to handle a device with these characteristics, a cortical reorganization That may not be possible for everyone.
Luca Tonin, Serafeim Perdikis, Taylan Deniz Kuzu, Thoman Armin Schildhauer, Ramón Martínez-Olivera, José del R. Millán. Learning to control a BMI-driven wheelchair for people with severe tetraplegia. IScience (2022). DOI: https://doi.org/10.1016/j.isci.2022.105418