Information processing in the brain involves the coordinated activity of large networks of neurons. Over the past 15 years we have developed and utilized technologies for extracting this information from neuronal populations in behaving animals. This involves chronically implanting multi-electrode arrays in functionally connected areas of the motor and somatosensory systems of the brain. Multi-processor computer systems are used to simultaneously discriminate and record the spiking activity of large numbers of single neurons within those systems. Mathematical techniques are used to decode the information processed by these neuronal populations while the animals perform specific behavioral tasks involving somatosensory perception and/or trained limb movement.
We have recently shown that electronic decoding of neuronal population activity could be used to extract and utilize "comm" for forelimb movement from the motor cortex of rats. The rats were initially trained to use their forelimbs to press down a lever to a certain position. The lever movement controlled movement of a robot arm to obtain a drop of water from a dropper (fig. 1). When the rat released the lever the robot arm delivered the water to the rat's mouth. Next we suddenly switched the control of the robot arm away from this lever-press, and replaced it with the electronic signal carrying decoded movement commands from the motor cortex. Four of the six rats were able to use this motor command signal to control the robot arm with sufficient accuracy to reliably and repeatedly retrieve water drops. After a few days of this training, the animals were increasingly able to move the robot arm and retrieve water without the concomitant lever pressing. This suggests that motor cortical control of limb movement is modifiable. Thus, paraplegic patients might be able to use their motor cortex activity to directly control a robot arm, or their own arm using functional stimulation of the paralyzed muscles. In collaborative studies with Dr. Miguel Nicolelis at Duke University the overall feasibility of this approach has been demonstrated in monkeys. We are now further developing technology to allow up to 1000 neurons to be recorded simultaneously.
Figure 1. Experimental paradigm. A. "Lever-movement->Robot-arm" mode: Rats were initially trained to press down a spring loaded lever (B) for a water reward. The graded lever displacement was electronically translated to produce proportional displacement of a robot arm (C) from its normal rest position (protruding through a slot in a Plexiglas barrier; D) to a water dropper (E). Upon lever release the robot arm (with the water drop) moved passively to the rest position where the rat could drink it."Neuronal-population-function->Robot-arm" mode: Rats were chronically implanted with multi-electrode recording arrays in the MI cortex and VL thalamus, yielding simultaneous recordings of up to 46 discriminated single neurons. G. Superimposed waveforms of 24 such neurons. H. Sample spike trains of two neurons (N1 & N2) over 2.0 sec. I. Neuronal-population (NP) function extracting the first principal component of a 32 neuron population. This NP function was extracted electronically in realtime using an weighted 32-channel integrator network with a 20 ms time constant. J. Switch to determine the source of input signals (i.e. lever-movement or NP function) for controlling position of the robot arm. In experiments, rats typically began working in the lever-movement->robot arm. Without warning the paradigm was then switched to the NP function->robot arm mode, allowing testing of whether the animal could routinely obtain its water through direct neural control of the robot arm.
Nicolelis, M. A. L., Baccala, L. A., Lin, R. C. S., and Chapin, J. K. (1995). Synchronous neuronal ensemble activity at multiple levels of the rat somatosensory system anticipates onset and frequency of tactile exploratory movements. Science 268, 1353-1358.
Nicolelis, A. L., Ghazanfar, A. A., Stambaugh, C. R., Oliveira, L. M., Laubach, M. Chapin, J. K., Nelson, R. J., and Kaas, J. H. (1998). Simultaneous representation of tactile information by three primate cortical areas. Nature Neuroscience 1, 621-630.
Chapin, J. K., Markowitz, R. A., Moxon, K. A., and Nicolelis, M. A. L. (1999). Direct real-time control of a robot arm using signals derived from neuronal population recordings in motor cortex. Nature Neuroscience 2, 664-670.
Chapin, J. K., and Nicolelis, M. A. L. (2000). Brain control of sensorimotor prostheses. In Neural Prostheses for Restoration of Sensory and Motor Function (Chapin, J. K., and Moxon, K. A. eds.), CRC Press, Baco Raton, pp. 235-262.
Wessberg, J., Stambaugh, C. R., Kralik, J. D., Beck, P. D., Laubach, M., Chapin, J. K., Kim, J., Biggs, S. J., Srinivasan, M. A., and Nicolelis, M. A. L. (2000). Real-time prediction of hand trajectory by ensembles of cortical neurons in primates. Nature 408, 361 - 365.