Our project integrates neurophysiological research into the somatosensory and visual mechanisms that govern planning of skilled motor behaviors of the hand with development and distribution of neuroinformatics tools useful for data visualization, analysis and manipulation required for such studies. We are establishing a Digital Video Neuroinformatics Resource to distribute integrated software tools that enable quantitative analyses of spike trains obtained with digital video (DV) recordings. It was motivated by our observation that neurophysiological studies of sensorimotor integration usually employ simple behavioral tasks because only a limited range and variety of movements can be monitored accurately using force or position sensors. The most widely used systems employ lights or reflective spots placed over joints. Movements of the fingers are difficult to study with such devices, as the markers can be shadowed by the hand or grasped object, and their outputs are not easily synchronized with simultaneously recorded neural responses. We have designed a versatile conjoint neurophysiological and behavioral data acquisition and analysis system using DV multimedia hardware and software that allows non-invasive visualization of kinematics, and also provides synchronized digitized recordings of neuronal spike trains. Our informatics tools enable quantitative analyses of spike trains obtained from DV recordings, and their correlation to kinematic actions captured in matching video images. The first generation of interactive tools will be provided for 1) spike recognition and separation, 2) event-linked rasters and PSTHs, and 3) continuous firing rate analyses for predicting behavioral activity from time varying properties of spike trains. New metrics of spike synchrony will be developed for multielectrode recordings. These include cross-correlograms, JPSTHs, synchronous events/sec histograms, and automated linkage of spike trains to instrumented data acquired in parallel on other platforms.
A parallel series of neurophysiological experiments will be performed using these neuroinformatics tools to investigate visual and somatosensory mechanisms that govern skilled actions of the hand. Spike train recordings using individual microelectrodes, and multiple electrode arrays, will analyze how visual and haptic responses are processed in parallel by neuronal assemblies within posterior parietal cortex (PPC) when objects are acquired and manipulated by the hand. These experiments assess the contribution of visual and haptic information about object size, shape and location in motor planning of acquisition and manipulation by the hand. We examine whether synchronization and/or coherence of firing between PPC regions dominated by vision and touch enable a match-to-sample mode of motor control in which the eye leads the hand in skilled actions. We postulate that skilled hand behaviors require the registration and coordination of an external map that locates an object's spatial coordinates and encodes its intrinsic geometry, and an internal map of the body's own image that represents the hand posture and dynamics. These are tested with prehension tasks in which visual or haptic cues are needed to acquire the appropriate object, and to perform skilled hand manipulations during tool use. This research has important clinical implications for understanding dysmetrias, optic ataxias, grasp abnormalities and neglect syndromes resulting from neurological damage to PPC, and provides basic insights into mechanisms of visuomotor control of the hand that may prove useful for rehabilitation after stroke or injury.
The DV tools also have wider potential application for quantitative analyses of motor skills and deficits during learning and aging. The DV informatics tools will be tested and evaluated by an international advisory panel of neuroscientists studying motor and/or sensory functions of the hand and eyes. The tools will be used to correlate hand and eye movements with neuronal activity in parietal-occipital, somatosensory, motor and premotor cortex, and with electromyographic (EMG) recordings from hand and arm muscles. Future implementations will extend the DV tools to a larger repertoire of behaviors, and other electrophysiological signals.
Kinematic and electrophysiological time series data acquired from the proposed studies will be shared on the HBP-supported Cortical Neuron Database (http://neurodatabase.org/). Synchronized spike trains from real neuronal assemblies obtained with a variety of experimental paradigms will be freely available to neuroscientists, to serve as a resource for developing new algorithms for analysis of simultaneous recordings from cortical populations, and for realistic modeling of parallel distributed neural networks.
|