Diagram is from Nature, Vol. 408, 362. Used with permission from Miguel A. L. Nicolelis
First, the general region of neurons associated with the movement of a particular body part or sensory function needs to be identified. Then, a means to decode these signals and translate them to a device that will mimic the movement or function and continue to correctly do so in the long-term needs to be determined. General algorithms or mathematical equations have been created to translate these brain signals that can predict the trajectory of the movement. But, then artificial devices need to be created. These devices have to be able to process and store the signals like a mini-mini-computer. So far, “neurochips,” little microchips used in the brain, have been created, but have yet to be as efficient and reliable as needed.
Unlike previous techniques using single movable electrodes, the team at Duke used microwire arrays to record larger populations of neurons.
So, how exactly did they do it?
Well, they hooked up a cap to a monkey which had microwires that were directly in contact with the motor neurons in the brain. This was used to collect the electrical signals from the neurons. They used computers to monitor the neural traffic and signals from the brain when a monkey reached for an object. This information was fed into the data acquisition unit, as shown in the diagram, and relayed to a computer that translated the electrical signals from the brain (through mathematical modeling) to commands that a robotic arm could use, and thus mimicking the action.
Why is mapping more neurons more effective than the previous technique with single movable electrodes? The brain is quite complex and uses multiple neurons to perform certain actions. It is necessary to observe how larger populations of neurons interact and behave during motor movements in order to get a better idea of how the brain works. This newer technique also has the benefit of monitoring populations of neurons for longer periods of times.