Thanks to an unprecedented neurosurgical intervention, a quadriplegic can activate a prosthesis by thought. A small revolution, which calls for others.
It’s a hell of a breakthrough that we are witnessing. This Friday, the review Science relay a study on the man-machine interface that leaves you dreaming. Indeed, a quadriplegic patient managed to control an artificial arm through his thought, and this, thanks to an unprecedented intervention.
The brain zone of intentions
The clinical trial took place in the United States. Erik Sorto, 34, underwent surgery at Keck Hospital in Los Angeles in April 2013. For five hours, surgeons implanted a neuronal prosthesis in a specific area of the brain: the posterior parietal cortex (PPC), where intentions are formed.
Upon awakening, and after long training sessions with researchers from the California Institute of Technology (Caltech), the patient was able to perform more or less complex gestures with precision, such as shaking a hand, triggering a blender. , have a drink and have a drink or even play “Shifumi”. The signals of his thought are in fact translated by an algorithm interpreted by a computer, which itself controls a remote manipulator arm.
Fluid gestures
This is not the first time that a paralyzed person has managed to order a prosthesis by their encoded thought. However, such precision of movement had never been achieved. And for good reason: the revolutionary character of this intervention lies in the identification of a specific network of neurons, involved in the intention that guides our actions.
So far, researchers have implanted neural prostheses in the part of the brain that controls movement, with less satisfactory results, since with this technique, gestures tend to be jerky. On the other hand, by implanting in the posterior parietal cortex, the movements regain fluidity, as shown in the video.
“Develop more reliable systems”
This is explained by the chronological process that takes place in the brain during movement. “The posterior parietal cortex is located upstream in the process leading to movement, so the signals are more related to the intention to act than to the actual execution of the movement,” writes the Richard Andersen, professor of neurology at Caltech who led this research.
“This area of the cortex does not perform, it has a planning role,” specifies to Why actor Marco congedo, CNRS researcher in the VIBS team (Vision and Brain Signal Processing). It is therefore the intention that is encoded and not the movement itself. The great novelty of this study is to show that in this area there is information on both the goal and the trajectory of the desired movement. ”
Marco Congedo, CNRS researcher in the VIBS (Vision and Brain Signal Processing) team: “The idea is to use this new knowledge in the future to develop more reliable systems. “
An invasive technique
However, this device will not be implanted in all quadriplegics from tomorrow. In fact, this trial will first need to be confirmed in other studies in order to replicate these results. However, “there is little doubt about the reproducibility of these results,” emphasizes Marco Congedo.
But above all, the invasive nature of this kind of man-machine interface poses problems in terms of ethics and practicality. “This process only takes place on subjects suffering from a serious illness, who agree to have electrodes implanted in the brain at all times,” explains Marco Congedo. At least for a while, since most materials are not biocompatible. The brain rejects them after a few months, and in general, the implant has to be replaced, which is not at all practical ”.
Marco cogendo : “In the end, few patients use this type of system because of the invasiveness of the procedures. “
The challenge: to create adaptable algorithms
Finally, another obstacle could limit the immediate impact of these results, which the authors do not fail to point out. Indeed, this thought coding device (in this case, intention) must be able to evolve according to the plasticity of the brain, which is far from obvious. “Patterns change over time. As you speak, you are creating synapses and destroying others! The brain is never equal to itself. It is not enough to set up a learning algorithm once; this algorithm must be able to adapt all the time to understand these changes ”.
.