Signals derived from the rat motor cortex can be used for controlling one-dimensional movements of a robot arm. It remains unknown, however, whether real-time processing of cortical signals can be employed to reproduce, in a robotic device, the kind of complex arm movements used by primates to reach objects in space. Here we recorded the simultaneous activity of large populations of neurons, distributed in the premotor, primary motor and posterior parietal cortical areas, as non-human primates performed two distinct motor tasks. Accurate real-time predictions of one- and three-dimensional arm movement trajectories were obtained by applying both linear and nonlinear algorithms to cortical neuronal ensemble activity recorded from each animal. In addition, cortically derived signals were successfully used for real-time control of robotic devices, both locally and through the Internet. These results suggest that long-term control of complex prosthetic robot arm movements can be achieved by simple real-time transformations of neuronal population signals derived from multiple cortical areas in primates.

To determine whether simultaneously recorded motor cortex neurons can be used for real-time device control, rats were trained to position a robot arm to obtain water by pressing a lever. Mathematical transformations, including neural networks, converted multineuron signals into 'neuronal population functions' that accurately predicted lever trajectory. Next, these functions were electronically converted into real-time signals for robot arm control. After switching to this 'neurorobotic' mode, 4 of 6 animals (those with > 25 task-related neurons) routinely used these brain-derived signals to position the robot arm and obtain water. With continued training in neurorobotic mode, the animals' lever movement diminished or stopped. These results suggest a possible means for movement restoration in paralysis patients.

We have used single unit recording techniques to map the representation of cutaneous and joint somatosensory modalities in the primary somatosensory (SI) cortex of both anesthetized and awake rats. The cytoarchitectonic zones within the rat SI were divided into the following main categories: (1) granular zones (GZs)--areas exhibiting koniocortical cytoarchitecture (i.e., containing dense aggregates of layer IV granule cells), (2) perigranular zones (PGZs)--narrow strips of less granular cortex surrounding the GZs, and (3) dysgranular zones (DZs)--large areas of dysgranular cortex enclosed within the SI. The narrow strip between the SI and the rostrally adjacent frontal agranular cortex was termed the "transitional zone" (TZ). Initial computer-based studies of the properties of cutaneous receptive fields (RFs) in SI showed that, although there were differences in response threshold, adaptability, frequency response, and overall RF size and shape of adjacent neurons, the size and location of the "centers" of the RFs were quite constant and were similar to those seen in multiple unit recordings. The same was true of RFs of single neurons recorded through different anesthetic states. The body representation in SI was first mapped by determining single unit and unit cluster RFs within a total of 2,170 microelectrode penetrations in barbiturate-anesthetized rats. Cutaneous RFs in the GZs were quite discrete. Thus, a single, finely detailed, continuous map of the cutaneous periphery was definable within the GZs themselves. Only the forepaw had a double representation. RFs in the PGZs were larger and more diffuse, but since they covered roughly the same skin areas as the RFs in the most closely adjacent GZs, they fit into the same body map. Neurons in the DZs were unresponsive to any sensory stimuli in the anesthetized animal. In chronically implanted, freely moving, awake animals cutaneous RFs were larger and more volatile than in the anesthetized, but the accuracy of the map was clearly preserved by the fact that the locations of the RF centers (which often must be defined quantitatively) were unchanged. The PGZs and DZs in the awake animals exhibited a multimodal convergence of cutaneous and joint movement RFs within single vertical penetrations, or even on single neurons. Directionally specific and bilateral cutaneous RFs were also observed in the DZs. It was concluded the DZs are more associational or integrative areas within the SI, but they could not be shown to comprise a distinct and separate body representation.(ABSTRACT TRUNCATED AT 400 WORDS)

Neural ensemble processing of sensorimotor information during behavior was investigated by simultaneously recording up to 48 single neurons at multiple relays of the rat trigeminal somatosensory system. Cortical, thalamic, and brainstem neurons exhibited widespread 7- to 12-hertz synchronous oscillations, which began during attentive immobility and reliably predicted the imminent onset of rhythmic whisker twitching. Each oscillatory cycle began as a traveling wave of neural activity in the cortex that then spread to the thalamus. Just before the onset of rhythmic whisker twitching, the oscillations spread to the spinal trigeminal brainstem complex. Thereafter, the oscillations at all levels were synchronous with whisker protraction. Neural structures manifesting these rhythms also exhibited distributed spatiotemporal patterns of neuronal ensemble activity in response to tactile stimulation. Thus, multilevel synchronous activity in this system may encode not only sensory information but also the onset and temporal domain of tactile exploratory movements.

No abstract