GBS Somatosensory Research

Dynamics of the Thalamocortical Loop

Rats and other rodents use arrays of facial vibrissae to actively discriminate between surface features in the external world with a high degree of accuracy (Guic-Robles et al., 1989; Carvell and Simons, 1990, 1995). The resulting vibrissa deflections, underlying the animal’s tactile perception, possess correlations shaped by dynamics of active whisking, vibrissa anatomy, and the mechanics of vibrissa-surface interface. Although there is still much to learn about the mechanics of the facial vibrissa in naturalistic conditions, we have utilized deflection patterns of increasing complexity in our laboratory, as shown in the left of Figure 10, to provide building blocks for more complex, texture-like patterns.

Figure 10. Left: Vibrissa deflection with various temporal structures (A-D). Right: Response (in spikes/ms, at 2ms resolution) of cortical cell to above deflections, in microns, 1 cm from the face. From Webber & Stanley (2004).

One of the most basic elements of the response is the pair-wise interaction, as shown in the right of the figure. Cortical cells respond to a punctate stimulus with a quick initial excitatory response component, followed by a prolonged, relatively pronounced inhibitory tail, which can extend past 100 ms. Using a pair-wise tactile probe and varying the time between the deflections, we infer the level of suppression induced by the first deflection from the magnitude of the response to the second, as shown in the right of Figure 10.

Nonlinear Spatiotemporal Encoding. Temporal stimulus patterns induced nonlinear interactions between excitatory and suppressive components of the neuronal response. In a recent study, we characterized the nonlinear dynamics of cortical responses to tactile patterns. Specifically, we found that the stimulus-evoked suppression, described above, serves to “suppress the suppression” normally induced by subsequent stimuli, resulting in complex responses to stimulus patterns, as illustrated in Figure 11.

Figure 11. Stimuli presented to the primary vibrissa (PV) and PSTHs of a typical barrel neuron with 1st pulse width varying and 2nd held at 40 ms. Dashed line: time of 3rd deflection. From Webber & Stanley (2004).

When the time interval between the first and second deflection is large, as in the leftmost panel, the second deflection serves to suppress the response to the third. However, when the time interval between the first and second is slightly shorter (middle), the first suppresses the second, and effectively "releases" the third from suppressive effects. Finally, when both time intervals are small (right), the responses to the second and third deflections are suppressed. This simple phenomena is extremely robust, is predictable, and dictates how the vibrissa respond to more complex patterns of stimuli.

Understanding how spatiotemporally distributed stimuli are encoded is essential to the study of sensory processing in the vibrissa system. Thalamocortical RFs typically span multiple vibrissae. Spatiotemporally distributed deflections thus engage multiple RF subregions, producing complex temporal patterns of neuronal activity across the array. Such interactions are common to visual, auditory, and other somatosensory pathways. However, the resulting interplay between excitatory and inhibitory mechanisms is still not fully understood. In recent work, we have quantified the manner in which spatiotemporal stimuli are encoded in this pathway, using probes such as that depicted in Figure 12.

Figure 12. PV-AV pair over a corrugated surface. White arrows and the corresponding labels I-IV identify four types of pair-wise interactions that were experimentally measured.

Use of simplified punctate deflections showed that cross-vibrissa interactions strongly influence the response to spatiotemporally distributed patterns of deflections. In particular, spatial components dramatically altered the frequency response properties of single cells, as shown in Figure 13. The primary vibrissa (PV) and adjacent vibrissa (AV) were presented with a fixed frequency train of impulse deflections (period = 120ms), but the phase (or latency) between the two deflection patterns was varied. This has a dramatic effect on the corresponding cortical response (gray), as compared to stimulation of the PV alone (black).

Figure 13. Dependence of transient responses on PV-AV time delay for a stimulation period of 120ms. Results for td = 30, 75 and 90ms are shown in the left, middle and right columns, respectively. For each column, the top panel represents a schematic of the two-vibrissa stimulus (Top: PV; Bottom: AV). The firing rate averaged over all cells is shown in the middle panel. The bottom panel shows the transient fractional PV responses for the two-vibrissa case (gray), along with those resulting from the 120ms-periodic single-vibrissae stimulus (black).

Secondly, an experimentally-based model provided accurate predictions of responses to distributed deflection patterns from simpler pair-wise interactions. The prediction model further demonstrated that, over certain timescales, spatiotemporal response integration enhances the discriminability of distributed patterns. This result is consistent with earlier studies in alert animals (Carvell and Simons, 1995), which showed that active discrimination of textures with two adjacent vibrissae was significantly better than with a single whisker. Taken together, the findings demonstrate the importance of spatiotemporal interactions to the coding of distributed deflection patterns, which may be critical in the natural environment.

Related Publications

A. S. Boloori and G. B. Stanley. The dynamics of spatiotemporal response integration in vibrissa representation in the primary somatosensory cortex, J. Neurosci., 23767-3782, 2006. PDF

R. M. Webber and G. B. Stanley. Transient and steady-state dynamics of cortical adaptation, J. Neurophys., 95:2923-2932, 2006. PDF

R. M. Webber and G. B. Stanley. Nonlinear encoding of tactile patterns in the barrel cortex, 91: 2010-2022, J. Neurophys., 2004. PDF

G. B. Stanley and R. W. Webber. A point process analysis of sensory encoding, J. Comput. Neurosci., 15:321-333, 2003. PDF


	SOMATOSENSORY RELATED PROJECTS Overview, Spatiotemporal Encoding, Adaptation, Behavior, Trauma - Click for related publications Back to Main Research Page

Overview - Tactile Sensation: Neural Encoding of Texture Rats and other rodents have arrays of facial whiskers (vibrissae) that they use to explore the tactile world. The vibrissa are vital for survival; neonates deprived of their vibrissa exhibit severely impaired behavioral development. Surprisingly, rats can also discriminate between very similarly textured surfaces based on vibrissa exploration alone, as shown in a study by Carvell and Simons (1990), the experimental paradigm of which is illustrated on the left in Figure 9. Figure 9. Left panel shows two-alternative forced choice task of discrimination between two textures with vibrissa exploration alone (from Carvell and Simons, 1990). Right panel shows the basic anatomy of the pathway, from the vibrissa (or whiskers), to the brainstem, thalamus, and to excitatory and inhibitory neurons in cortical columns (or "barrels"). Letters and numbers indicate rows and arcs of vibrissa on the facial pad. Note the maintenance of the somatotopic map throughout the pathway. As shown on the right, there is a topographical representation of the facial vibrissa in the brainstem, thalamus, and in cortical columns containing excitatory and inhibitory cells. The strong somatotopic organization of this pathway, along with the behavioral relevance of this sensory modality, has served as motivation for our work in neuronal encoding in the vibrissa system. Importantly, we believe that many of the general principles of coding in this pathway are similar to those in other pathways, such as the human fingertip, or the visual pathway. From an engineering perspective, the vibrissa system represents an array of active mechanical sensors, the biological understanding of which might be useful in robotic applications, either to augment other sensor modalities, or for environments where, for example, artificial vision would not be feasible. Back to Main Research Page
Dynamics of the Thalamocortical Loop Rats and other rodents use arrays of facial vibrissae to actively discriminate between surface features in the external world with a high degree of accuracy (Guic-Robles et al., 1989; Carvell and Simons, 1990, 1995). The resulting vibrissa deflections, underlying the animal’s tactile perception, possess correlations shaped by dynamics of active whisking, vibrissa anatomy, and the mechanics of vibrissa-surface interface. Although there is still much to learn about the mechanics of the facial vibrissa in naturalistic conditions, we have utilized deflection patterns of increasing complexity in our laboratory, as shown in the left of Figure 10, to provide building blocks for more complex, texture-like patterns. Figure 10. Left: Vibrissa deflection with various temporal structures (A-D). Right: Response (in spikes/ms, at 2ms resolution) of cortical cell to above deflections, in microns, 1 cm from the face. From Webber & Stanley (2004). One of the most basic elements of the response is the pair-wise interaction, as shown in the right of the figure. Cortical cells respond to a punctate stimulus with a quick initial excitatory response component, followed by a prolonged, relatively pronounced inhibitory tail, which can extend past 100 ms. Using a pair-wise tactile probe and varying the time between the deflections, we infer the level of suppression induced by the first deflection from the magnitude of the response to the second, as shown in the right of Figure 10. Nonlinear Spatiotemporal Encoding. Temporal stimulus patterns induced nonlinear interactions between excitatory and suppressive components of the neuronal response. In a recent study, we characterized the nonlinear dynamics of cortical responses to tactile patterns. Specifically, we found that the stimulus-evoked suppression, described above, serves to “suppress the suppression” normally induced by subsequent stimuli, resulting in complex responses to stimulus patterns, as illustrated in Figure 11. Figure 11. Stimuli presented to the primary vibrissa (PV) and PSTHs of a typical barrel neuron with 1st pulse width varying and 2nd held at 40 ms. Dashed line: time of 3rd deflection. From Webber & Stanley (2004). When the time interval between the first and second deflection is large, as in the leftmost panel, the second deflection serves to suppress the response to the third. However, when the time interval between the first and second is slightly shorter (middle), the first suppresses the second, and effectively "releases" the third from suppressive effects. Finally, when both time intervals are small (right), the responses to the second and third deflections are suppressed. This simple phenomena is extremely robust, is predictable, and dictates how the vibrissa respond to more complex patterns of stimuli. Understanding how spatiotemporally distributed stimuli are encoded is essential to the study of sensory processing in the vibrissa system. Thalamocortical RFs typically span multiple vibrissae. Spatiotemporally distributed deflections thus engage multiple RF subregions, producing complex temporal patterns of neuronal activity across the array. Such interactions are common to visual, auditory, and other somatosensory pathways. However, the resulting interplay between excitatory and inhibitory mechanisms is still not fully understood. In recent work, we have quantified the manner in which spatiotemporal stimuli are encoded in this pathway, using probes such as that depicted in Figure 12. Figure 12. PV-AV pair over a corrugated surface. White arrows and the corresponding labels I-IV identify four types of pair-wise interactions that were experimentally measured. Use of simplified punctate deflections showed that cross-vibrissa interactions strongly influence the response to spatiotemporally distributed patterns of deflections. In particular, spatial components dramatically altered the frequency response properties of single cells, as shown in Figure 13. The primary vibrissa (PV) and adjacent vibrissa (AV) were presented with a fixed frequency train of impulse deflections (period = 120ms), but the phase (or latency) between the two deflection patterns was varied. This has a dramatic effect on the corresponding cortical response (gray), as compared to stimulation of the PV alone (black). Figure 13. Dependence of transient responses on PV-AV time delay for a stimulation period of 120ms. Results for td = 30, 75 and 90ms are shown in the left, middle and right columns, respectively. For each column, the top panel represents a schematic of the two-vibrissa stimulus (Top: PV; Bottom: AV). The firing rate averaged over all cells is shown in the middle panel. The bottom panel shows the transient fractional PV responses for the two-vibrissa case (gray), along with those resulting from the 120ms-periodic single-vibrissae stimulus (black). Secondly, an experimentally-based model provided accurate predictions of responses to distributed deflection patterns from simpler pair-wise interactions. The prediction model further demonstrated that, over certain timescales, spatiotemporal response integration enhances the discriminability of distributed patterns. This result is consistent with earlier studies in alert animals (Carvell and Simons, 1995), which showed that active discrimination of textures with two adjacent vibrissae was significantly better than with a single whisker. Taken together, the findings demonstrate the importance of spatiotemporal interactions to the coding of distributed deflection patterns, which may be critical in the natural environment. Related Publications A. S. Boloori and G. B. Stanley. The dynamics of spatiotemporal response integration in vibrissa representation in the primary somatosensory cortex, J. Neurosci., 23767-3782, 2006. PDF R. M. Webber and G. B. Stanley. Transient and steady-state dynamics of cortical adaptation, J. Neurophys., 95:2923-2932, 2006. PDF R. M. Webber and G. B. Stanley. Nonlinear encoding of tactile patterns in the barrel cortex, 91: 2010-2022, J. Neurophys., 2004. PDF G. B. Stanley and R. W. Webber. A point process analysis of sensory encoding, J. Comput. Neurosci., 15:321-333, 2003. PDF Back to Main Research Page
Adaptation & Invariant Neuronal Representations Adaptation is a ubiquitous property of all sensory pathways of the brain, and thus likely critical in the encoding of behaviorally-relevant sensory information. Despite evidence identifying specific biophysical mechanisms contributing to sensory adaptation, its functional role in sensory encoding is not well understood, particularly in the natural environment where transient rather than steady-state activity could dominate the neuronal representation. Figure 14. Adaptation dynamics. (a) Top panel: Periodic square wave stimulus at 6 Hz, beginning at rest with alternating away from rest and back to rest deflections. Middle panel: Corresponding average PSTH for all 41 recorded cells. Bottom panel: Average firing rate in a 30 ms window following each deflection of the stimulus (away from rest: black, back to rest: gray). The portion of the response prior to reaching steady-state is termed the transient portion. (b) PSTH following the first deflection (black) and at steady-state (gray) in the away and back directions. PSTH for isolated stimulus in back direction (dashed black). Arrows represent time of vibrissa deflection. In a recent study, we show that the heterogeneous transient and steady-state adaptation dynamics of single cortical neurons in the rat vibrissa system are directly predictable from temporal tuning properties that capture the time course of post-excitatory suppression following an isolated vibrissa deflection. Figure 14 shows the adaptation response to a periodic deflection pattern, when averaged over individual cells the population. Single cell transient cortical adaptation dynamics were strongly influenced by the state of the thalamocortical network at the onset of the sensory stimulus. Specifically, we altered the state of the network by preceding the periodic stimulus with an additional deflection, placing the network in a suppressed state prior to stimulus onset. Despite dramatically different transient dynamics, neurons adapted to the same steady-state response regardless of the initial state of the network, consistent with attractor-like behavior of dynamical systems. Taken together, our recent results show that transient rapid adaptation dynamics are set in motion by the state of the thalamocortical network at the onset of a sensory stimulus, yet steady-state adapted responses are invariant, reflecting a possible “set-point” of the neuronal representation. Related Publications R. M. Webber and G. B. Stanley. Transient and steady-state dynamics of cortical adaptation, J. Neurophys., 95:2923-2932 2006. PDF R. M. Webber and G. B. Stanley. Post-excitatory suppression dictates the dynamics of adaptation in barrel cortex, COSYNE, Salt Lake City, 2005. PDF R. M. Webber and G. B. Stanley. Transient adaptation properties in rat barrel cortex are affected by the initial direction of vibrissa deflection, SFN, San Diego, 2004. HTML Back to Main Research Page
Behavior & Perception Many rodents actively sense the environment using an extensive whisker system on the snout, which has a large somatotopic representation in primary somatosensory cortex. Although many studies have investigated this system in anesthetized rats, there has been a growing recognition of the importance of making electrophysiological recordings in awake behaving animals. We made multiple simultaneous single unit extracellular and multiple field potential recordings from the barrel region of primary somatosensory cortex in adult female Long-Evans rats that were trained to run on a linear track. We placed several piezo-electric films on the side of the track in order to make discrete deflections of the contralateral whiskers with accompanying PSTHs, with varying time separations between each deflection. This paradigm allows us to probe the dynamics of tactile responses in the barrel-cortex network. It has been shown that behavior state modulates tactile responses. We hypothesize that the shallow, short suppression following a stimulus that is characteristic of the awake state optimizes the rat's sensory discrimination ability, while the deep, long suppression characteristic of the quiescent state optimizes the rat's ability to detect the onset of stimuli. Further work will employ a behavioral paradigm that requires the rat to make a choice after a whisker-based detection or discrimination between different piezo spacings, under varying attentional states, in order to test this hypothesis. Related Publications R. A. Jenks and G. B. Stanley. Stimulus-driven response dynamics of the whisker/barrel-cortex system in awake behaving rats, to be presented at SFN, Washington DC, 2005. PDF R. A. Jenks, T. Warren, L. M. Frank, and G. B. Stanley. Texture detection and discrimination by the whisker/barrel-cortex system in awake behaving rats, SFN, New Orleans, 2003. HTML Back to Main Research Page
Traumatic Brain Injury (TBI) Neuronal injury resulting from direct mechanical impact activates secondary processes of cytotoxicity (excitotoxicity, oxidative stress, and apoptosis), which continue to evolve over hours/days. Following injury, the electrical activity of neuronal populations within relevant brain structures has not been widely studied at the single- or multi-unit level, and the role that the electrical activity may play in the recovery process is only beginning to be addressed. In this study, Long-Evans rats were subjected to traumatic brain injury (TBI) in the barrel cortex through the use of a controlled cortical impactor (CCI) device. With the use of a paired-pulse and 3-pulse whisker deflection paradigm, TBI at the electrophysiological (single/multi-unit and EEG) and histological levels in the barrel cortex of the rat somatosensory system were monitored and evaluated at various time points following recovery. Post-stimulus time histograms (PSTH) seem to indicate that periods of enhanced excitation and cortical suppression exist following induction of injury. Excitotoxicity and cortical spreading depression (CSD) are often reported following traumatic brain injury and the resulting changes in PSTH amplitudes at multiple time points after TBI induction. The results suggest a possible perturbation in the balance of inhibitory and excitatory neuronal activity in the cortex following injury. These data provide possible insight into the time course of the development of injury and could potentially serve as indicators of outcome. Controlled modulation of inhibitory and excitatory activity at critical time points could lead to more effective interventions for enhancing recovery following TBI. Related Publications M. C. Ding, E. Tejima, E. H. Lo, and G. B. Stanley, Electrophysiology of the rat barrel cortex following traumatic brain injury, to be presented at SFN, Washington DC, 2005. PDF Back to Main Research Page

SOMATOSENSORY RELATED PROJECTS

Overview, Spatiotemporal Encoding, Adaptation, Behavior, Trauma - Click for related publications

Back to Main Research Page

Overview - Tactile Sensation: Neural Encoding of Texture

Adaptation & Invariant Neuronal Representations

Behavior & Perception

Traumatic Brain Injury (TBI)