Driving Brain Re-wiring using direct digital signal processing

This is a summary of an e-mail written to Professor Nicolelis at Duke University. If we get a reply we will publish the results of the conversation here.


I recently came across your work in decoding neural signals and translating them into a digital code that can drive machines.

My interest is actually to reverse that process. Specifically, to help the injured brain to re-wire.

As you may be aware, Merzenich and Taub have shown that the brain is plastic and the brain can re-wire to recruit undamaged areas of the brain. However, this type of re-wiring does not come easy. It requires concentration and willpower from the patient. Also, it requires many thousands or tens of thousands of repetitions. Passive movement does seem to produce effective rehabilitation results.

Therefore, we have turned to individuals like Neville Hogan of MIT. Neville’s work is with robotic exoskeletons which provide assistance to the patient in an effort to motivate them to continue their training. The robot might provide 90% assistance at first, hover over many repetitions the robot slowly backs off its support and the patient is doing more and more of the work on their own as the brain re-wires.

This is exciting, but the results have not been as spectacular as we had hoped. It seems that the complexities of the machine to human interface cause less then accurate reproductions of natural motor learning.

This is where your work comes in. I have recently began to wonder if a more effective training mechanism for re-wiring the brain might be to skip the mechanical interface of a robot moving the human arm and just directly tap into the nerves or directly into the brain cells themselves and drive the neurons to fire in patterns that you have determined in your work would be “normal” patterns for certain motor movements.

The advantages of such a method is that potentially this could be a much “cleaner” signal to the brain vs. using a robot or traditional therapist to help the patient move their arm. Also, because it is a recorded signal, it could be reproduced many times with little effort and perhaps at much faster speeds then could be done with physical limb movements. The question of “concentration” comes up, but Merzenich has found some chemical methods to stimulate plasticity that might be employed in a controlled setting during the re-wiring sessions.