Receptive Field Dimensions Research Articles

Alignable kernel network

To enhance the adaptability and performance of Convolutional Neural Networks (CNN), we present an adaptable mechanism called Alignable Kernel (AliK) unit, which dynamically adjusts the receptive field (RF) dimensions of a model in response to varying stimuli. The branches of AliK unit are integrated through a novel align transformation softmax attention, incorporating prior knowledge through rank ordering constraints. The attention weightings across the branches establish the effective RF scales, leveraged by neurons in the fusion layer. This mechanism is inspired by neuroscientific observations indicating that the RF dimensions of neurons in the visual cortex vary with the stimulus, a feature often overlooked in CNN architectures. By aggregating successive AliK ensembles, we develop a deep network architecture named the Alignable Kernel Network (AliKNet). AliKNet with interdisciplinary design improves the network’s performance and interpretability by taking direct inspiration from the structure and function of human neural systems, especially the visual cortex. Empirical evaluations in the domains of image classification and semantic segmentation have demonstrated that AliKNet excels over numerous state-of-the-art architectures, achieving this without increasing model complexity. Furthermore, we demonstrate that AliKNet can identify target objects across various scales, confirming their ability to dynamically adapt their RF sizes in response to the input data.

Transmission of colour and acuity signals by parvocellular cells in marmoset monkeys

The red-green axis of colour vision evolved recently in primate evolutionary history. Signals serving red-green colour vision travel together with signals serving spatial vision, in the parvocellular (PC) division of the subcortical visual pathway. However, the question of whether receptive fields of PC pathway cells are specialized to transmit red-green colour signals remains unresolved. We addressed this question in single-cell recordings from the lateral geniculate nucleus of anaesthetized marmosets. Marmosets show a high proportion of dichromatic (red-green colour-blind) individuals, allowing spatial and colour tuning properties of PC cells to be directly compared in dichromatic and trichromatic visual systems. We measured spatial frequency tuning for sine gratings that provided selective stimulation of individual photoreceptor types. We found that in trichromatic marmosets, the foveal visual field representation is dominated by red-green colour-selective PC cells. Colour selectivity of PC cells is reduced at greater eccentricities, but cone inputs to centre and surround are biased to create more selectivity than predicted by a purely 'random wiring' model. Thus, one-to-one connections in the fovea are sufficient, but not necessary, to create colour-selective responses. The distribution of spatial tuning properties for achromatic stimuli shows almost complete overlap between PC cells recorded in dichromatic and trichromatic marmosets. These data indicate that transmission of red-green colour signals has been enabled by centre-surround receptive fields of PC cells, and has not altered the capacity of PC cells to serve high-acuity vision at high stimulus contrast.

Human Umbilical Cord Blood Cells Restore Brain Damage Induced Changes in Rat Somatosensory Cortex

Intraperitoneal transplantation of human umbilical cord blood (hUCB) cells has been shown to reduce sensorimotor deficits after hypoxic ischemic brain injury in neonatal rats. However, the neuronal correlate of the functional recovery and how such a treatment enforces plastic remodelling at the level of neural processing remains elusive. Here we show by in-vivo recordings that hUCB cells have the capability of ameliorating the injury-related impairment of neural processing in primary somatosensory cortex. Intact cortical processing depends on a delicate balance of inhibitory and excitatory transmission, which is disturbed after injury. We found that the dimensions of cortical maps and receptive fields, which are significantly altered after injury, were largely restored. Additionally, the lesion induced hyperexcitability was no longer observed in hUCB treated animals as indicated by a paired-pulse behaviour resembling that observed in control animals. The beneficial effects on cortical processing were reflected in an almost complete recovery of sensorimotor behaviour. Our results demonstrate that hUCB cells reinstall the way central neurons process information by normalizing inhibitory and excitatory processes. We propose that the intermediate level of cortical processing will become relevant as a new stage to investigate efficacy and mechanisms of cell therapy in the treatment of brain injury.

Differential Targeting of Optical Neuromodulators to Ganglion Cell Soma and Dendrites Allows Dynamic Control of Center-Surround Antagonism

Retinal degenerative diseases cause photoreceptor loss and often result in remodeling and deafferentation of the inner retina. Fortunately, ganglion cell morphology appears to remain intact long after photoreceptors and distal retinal circuitry have degenerated. We have introduced the optical neuromodulators channelrhodopsin-2 (ChR2) and halorhodopsin (NpHR) differentially into the soma and dendrites of ganglion cells to recreate antagonistic center-surround receptive field interactions. We then reestablished the physiological receptive field dimensions of primate parafoveal ganglion cells by convolving Gaussian-blurred versions of the visual scene at the appropriate wavelength for each neuromodulator with the Gaussians inherent in the soma and dendrites. These Gaussian-modified ganglion cells responded with physiologically relevant antagonistic receptive field components and encoded edges with parafoveal resolution. This approach bypasses the degenerated areas of the distal retina and could provide a first step in restoring sight to individuals suffering from retinal disease.

The control of motoneurone activity: getting all of the players on the same stage

Studies of motor control, particularly at the level of the spinal cord, have placed a high premium on the exact pathways used by segmental and descending signals en route to motoneurones. The perceived importance of the role of these pathways has inspired numerous studies that have firmly embedded the term ‘identified interneuron’ into the lexicon of spinal neuroscience (Jankowska, 2001). While not disputing the need for a precise description of the circuits linking spinal neurons, to attain a truly comprehensive understanding of the control of motor activity by these circuits requires an equally precise description of the type of sensory information that is transmitted by these circuits and the transformation of this information by the intrinsic properties of spinal neurons, especially motoneurones. Thus, the control of motoneurone activity is similar to a theatrical production in which many actors and actresses, in this case representing different circuits, types of sensory information, and intrinsic properties, occupy the same stage and engage in an interactive dialogue. When viewed from the perspective of a scientist, the challenge is to get all of these players on the same experimental stage. Electrical stimulation of peripheral nerves is a time-honoured means of defining the source and type of information delivered by muscle, joint and cutaneous afferents (Matthews, 1972). When used in conjunction with electrophysiological recordings that are synchronized to the time of the stimulus delivery, electrical stimulation of peripheral nerves offers a powerful tool to discriminate the neuronal circuits responsible for the transmission of information arising from individual classes of sensory afferents. However, this technique is ill-suited to mimic the sensory information that is generated by natural movements of the limb. Under these conditions, afferents from multiple locations and modalities will be activated together and the activity in each afferent will vary depending on the position of the limb and the kinematics of the limb movement. Thus, in the context of the desired experimental play, electrical stimulation artificially limits the number of the players that represent the sum of sensory information generated by natural movements and the multiplicity of pathways engaged in the transmission of this information. In this issue of The Journal of Physiology, Hyngstrom et al. (2008) take advantage of robotic technology to get many of the key players on the same experimental stage. In this instance, a robot was used to independently rotate the ankle, knee and hip, or all three joints at the same time. These movements generate a multimodal barrage of sensory signals that arise from sensory receptors located throughout the hindlimb. A small number of investigators have previously used robots to examine the sensory responses of ascending tract neurons in the spinal cord (e.g. Bosco et al. 1996). However, Hyngstrom and colleagues have applied this technology in a way that allows us to listen to the dialogue between the many players representing sensory information, circuits, and importantly, the intrinsic properties of motoneurones. The dendrites of motoneurones possess an array of intrinsic conductances, including a set of voltage-dependent sodium and calcium channels that generate persistent inward currents (PICs, Heckman et al. 2003). PICs are readily activated or deactivated by sensory stimuli in decerebrate animals due to the actions of monoamines released by descending pathways. However, in the acute period following spinal transection, modulation of PICs by segmental circuits is greatly diminished. Thus, by comparing the current reaching the soma of ankle or knee extensor motoneurones in decerebrate and acutely spinalized animals during robotically controlled limb movements, Hyngstrom and colleagues were able to isolate the role of PICs (i.e. one of the intrinsic property players) during activation of multiple spinal circuits by physiologically relevant sensory information (represented by several players). The results of the study are described in the form of movement related receptive fields. These fields are defined by the strength of the signals received from different joints. The data suggest that PICs reduce the dimensions of movement related receptive fields. For example, in the absence of PIC modulation by descending monoaminergic systems, rotations at the hip generate large synaptic currents in ankle extensor motoneurones. These currents are greatly reduced in the presence of PIC modulation. It is perhaps ironic, given the original goal of recruiting more key players to the stage, that the interaction between these players causes some of the new members of the cast to lose their voices. This outcome may indicate that the players representing descending monoaminergic systems and PICs are the stars in this play. It is tempting to speculate that the more potent actions of intraspinal monoaminergic systems in chronic models of spinal cord injury may raise the voices of excitatory muscle afferent circuits and thereby contribute to muscle spasms (Harvey et al. 2006). However, the words for this play have yet to be written.

Spatial receptive field properties of lateral geniculate cells in the owl monkey (Aotus azarae) at different contrasts: a comparative study

Several physiological properties of owl monkey lateral geniculate nucleus (LGN) cells were studied to verify whether its nocturnal habit has an influence on the organization of its subcortical visual system. Receptive field (RF) dimensions were measured using drifting gratings and bipartite field stimuli. We found that owl monkey cells LGN have larger RFs and were on average more non-linear than those of diurnal monkeys. But, as in other anthropoids, there is an increase in RF centre size with increasing eccentricity, and there is a limited correlation between these centre sizes and retinal ganglion cell dendritic tree sizes. The influence of contrast on sizes and peak sensitivities of RF centres and surrounds and on the response phases was studied. Both the sizes and peak sensitivities of the RF centres and surrounds decrease as contrast increases. As a result, the responses to low spatial frequency stimuli saturate with increasing contrast. Estimates of contrasts at half-maximal responses confirm the presence of saturation. It was found that the magnocellular cells saturate more strongly than parvocellular cells. The response phase increases with increasing contrast. These data resemble those obtained in the common marmoset, indicating that these are basic features of the primate visual system. We conclude that during evolution and as an adaptation to a nocturnal lifestyle, cells in the owl monkey LGN display an increased spatial integration in comparison with diurnal primate species, without a change in the basic organization common to the primate subcortical visual system.

Dendritic impulse collisions and shifting sites of action potential initiation contract and extend the receptive field of an amacrine cell

We evaluated the contributions of somatic and dendritic impulses to the receptive field dimensions of amacrine cells in the amphibian retina. For this analysis, we used the NEURON simulation program with a multicompartmental, multichannel model of an On-Off amacrine cell with a three-dimensional structure obtained through computer tracing techniques. Simulated synaptic inputs were evenly spaced along the dendritic branches and organized into eight annuli of increasing radius. The first set of simulations activated each ring progressively to simulate an area summation experiment, while a second approach activated each annulus individually. Both sets of simulations were done with and without the presence of Na channels in the dendrites and soma. Unexpectedly, the receptive field dimensions observed in the area summation simulations was often smaller than that predicted from the summation of the annular simulations. Collisions of action potentials moving in opposite directions in the dendrites largely accounted for this contraction in receptive field size for the area summation studies. The presence of dendritic Na channels increased the size of the receptive field beyond that achieved in their absence and allowed the physiological size of the receptive field to approximate the physical dimensions of the dendritic tree. This receptive field augmentation was the result of impulse generating ability in the dendrites which enhanced the signal observed at the soma. These simulations provide a plausible mechanistic explanation for physiological recordings from amacrine cells that show similar phenomena.

Eye Position Compensation Improves Estimates of Response Magnitude and Receptive Field Geometry in Alert Monkeys

Studies of visual function in behaving subjects require that stimuli be positioned reliably on the retina in the presence of eye movements. Fixational eye movements scatter stimuli about the retina, inflating estimates of receptive field dimensions, reducing estimates of peak responses, and blurring maps of receptive field subregions. Scleral search coils are frequently used to measure eye position, but their utility for correcting the effects of fixational eye movements on receptive field maps has been questioned. Using eye coils sutured to the sclera and preamplifiers configured to minimize cable artifacts, we reexamined this issue in two rhesus monkeys. During repeated fixation trials, the eye position signal was used to adjust the stimulus position, compensating for eye movements and correcting the stimulus position to place it at the desired location on the retina. Estimates of response magnitudes and receptive field characteristics in V1 and in LGN were obtained in both compensated and uncompensated conditions. Receptive fields were narrower, with steeper borders, and response amplitudes were higher when eye movement compensation was used. In sum, compensating for eye movements facilitated more precise definition of the receptive field. We also monitored horizontal vergence over long sequences of fixation trials and found the variability to be low, as expected for this precise behavior. Our results imply that eye coil signals can be highly accurate and useful for optimizing visual physiology when rigorous precautions are observed.

Chromatic and spatial properties of parvocellular cells in the lateral geniculate nucleus of the marmoset (Callithrix jacchus).

The parvocellular (PC) division of the afferent visual pathway is considered to carry neuronal signals which underlie the red-green dimension of colour vision as well as high-resolution spatial vision. In order to understand the origin of these signals, and the way in which they are combined, the responses of PC cells in dichromatic ('red-green colour-blind') and trichromatic marmosets were compared. Visual stimuli included coloured and achromatic gratings, and spatially uniform red and green lights presented at varying temporal phases and frequencies.The sensitivity of PC cells to red-green chromatic modulation was found to depend primarily on the spectral separation between the medium- and long-wavelength-sensitive cone pigments (20 or 7 nm) in the two trichromatic marmoset phenotypes studied. The temporal frequency dependence of chromatic sensitivity was consistent with centre-surround interactions. Some evidence for chromatic selectivity was seen in peripheral PC cells. The receptive field dimensions of parvocellular cells were similar in dichromatic and trichromatic animals, but the achromatic contrast sensitivity of cells was slightly higher (by about 30%) in dichromats than in trichromats. These data support the hypothesis that the primary role of the PC is to transmit high-acuity spatial signals, with red-green opponent signals appearing as an additional response dimension in trichromatic animals.

Olfactory Bulb Mitral-Tufted Cell Plasticity: Odorant-Specific Tuning Reflects Previous Odorant Exposure

Olfactory system second-order neurons, mitral-tufted cells, have odorant receptive fields (ORFs) (molecular receptive ranges in odorant space for carbon chain length in organic odorant molecules). This study quantified several dimensions of these excitatory odorant receptive fields to novel odorants in rats and then examined the effects of passive odorant exposure on the shape of the ORF-tuning curve. ORFs for carbon chain length of novel ethyl esters (pure odorants that the animals had not been exposed to previously) were determined before and after a 50 sec prolonged exposure to one of the odorants. In response to novel odorants, quantitative analysis of mitral-tufted cell excitatory ORFs revealed that the median ORF width spanned 3-4 carbons, generally with a single-most excitatory odorant. Exposure to either the most excitatory odorant (ON-PEAK) or an odorant that was two carbons longer (OFF-PEAK) for 50 sec produced whole ORF suppression immediately after the end of the prolonged exposure, with the ON-PEAK exposure producing the greatest suppression. These results are consistent with a feature-detecting function for mitral-tufted cells. Redetermination of the ORF 15 and 60 min after the exposure revealed that OFF-PEAK exposure produced a reduction in responsiveness to the best odorant and an increase in responsiveness to the exposed odorant. In contrast, exposure to the ON-PEAK odorant or no odorant did not affect ORFs. Given that mitral-tufted cells receive exclusive excitatory input from olfactory receptor neurons expressing identical receptor proteins, it is hypothesized that experience-induced mitral-tufted cell ORF changes reflect modulation of lateral and centrifugal olfactory bulb circuits.

Spatial properties of koniocellular cells in the lateral geniculate nucleus of the marmoset Callithrix jacchus.

1. The receptive field dimensions, contrast sensitivity and linearity of spatial summation of koniocellular (KC), parvocellular (PC) and magnocellular (MC) cells in the lateral geniculate nucleus (LGN) of 11 adult marmosets were measured using achromatic sinusoidal gratings. 2. The receptive field centre diameter of cells in each (PC, KC and MC) class increases with distance from the fovea. There is substantial overlap in centre size between the three cell classes at any eccentricity, but the PC cells have, on average, the smallest centres and the KC cells have the largest. Some PC and KC cells did not respond at all to the grating stimulus. 3. The contrast sensitivity of the receptive field centre mechanism in KC cells decreases in proportion to the centre area. A similar trend was seen for the surround mechanism. These characteristics are common to PC and MC cells, suggesting that they originate at an early stage of visual processing in the retina. 4. The KC cells showed, in general, lower peak evoked discharge rates than PC or MC cells. The spontaneous discharge rate of KC cells was lower than that of PC cells and similar to that of MC cells. 5. The majority of cells in all divisions of the LGN show linear spatial summation. A few cells did show non-linear spatial summation; these cells were predominantly located in the MC and ventral KC layers. 6. The ventral KC layers below and between the MC layers contain cells with larger and more transiently responding receptive fields than cells in the more dorsal KC layers. 7. We conclude that many of the contrast-dependent spatial properties of cells in the marmoset LGN are common to PC, MC and KC cells. The main difference between KC cells and the other two classes is that there is more variability in their response properties, and they are less responsive to high spatial frequencies.

Dynamics of spatial summation in primary visual cortex of alert monkeys.

One of the fundamental tasks of the visual cortex is to integrate input from different parts of the retina, parsing an image into contours and surfaces, and then assembling these features into coherent representations of objects. To examine the role of the primary visual cortex in the integration of visual information, we measured the response properties of neurons under different stimulus conditions. Surprisingly, we found that even the most conventional measures of receptive field (RF) size were not fixed, but could vary depending on stimulus contrast and foreground-background relationships. On average, the length of the excitatory RF was 4-fold greater for a low-contrast stimulus than for a stimulus at high contrast. Embedding a high-contrast stimulus in a textured background tended to suppress neuronal responses and produced an enlargement in RF size similar to that observed by decreasing the contrast of an isolated stimulus. The results show that RF dimensions are regulated in a dynamic manner that depends both on local stimulus characteristics, such as contrast, and on global relationships between a stimulus and its surroundings.

Receptive field dimensions of lateral geniculate cells in the common marmoset ( Callithrix jacchus)

We measured the spatial receptive field dimensions of cells in the lateral geniculate nucleus (LGN) of the common marmoset ( Callithrix jacchus) using a bipartite field stimulus in which the two halves of the field were modulated identically but in countrphase. Horizontal and vertical edges between the two fields were positioned at different locations in the receptive field. By assuming that centers and surrounds have gaussian profiles, we were able to obtain a satisfactory mathematical description of the data. Receptive field centers were about a factor 1.6 larger than those of macaque LGN cells, in accordance with the smaller marmoset eye. There was a limited correspondence with dendritic tree dimensions of marmoset retinal ganglion cells. We further found that center and surround gaussians were not always concentric, and that the centers of some cells were elongated. This might allow some direction or orientation biases in LGN cells.

Spatial correlation of suppressive and excitatory receptive fields with direction selectivity of complex cells in cat striate cortex

The aim of the study was to account for direction selectivity of visual cortical neurons through systematic positional mismatch between excitatory and suppressive influences on each neuron. Direction-selective complex neurons were therefore recorded from striate cortex of cats lightly anaesthetized with halothane in nitrous oxide-oxygen. All but small residual eye movements were eradicated with intravenous gallamine triethiodide. Excitatory receptive field (ERF) dimensions and centring were quantified with optimal sine-wave grating stimuli, appropriately windowed to limit them to variable locations along and across the receptive field (RF) centre. Related suppressive receptive fields (SRFS) were similarly mapped during binocular conditioning, induced by an optimal grating applied to the other eye and drifting continuously in each neuron's preferred direction. Its purpose was to elevate ongoing levels of discharge to reveal often concealed null suppression. ERF and SRF profiles were systematically offset, especially along the line of preferred direction such that, for stimuli moving in the non-preferred direction, the SRF lay ahead of the ERF. Derivations of ERFS and SRFS during conditioning, within a single batch of trials, excluded eye movements as a source of positional mismatch. It is concluded that this mismatch may provide the basis for direction selectivity and the emergence of null suppression.

Physiological effects of chronic and acute application of N-methyl- d-aspartate and 5-amino-phosphonovaleric acid to the optic tectum of Rana pipiens frogs

Visually elicited activity contributes to the formation of orderly connections in the optic tectum of frogs. Glutamate receptors of the N-methyl- d-aspartate class participate in this process. Blocking those receptors interferes with activity-dependent refinement of maps in normal frogs and of ocular dominance bands in surgically produced animals with three eyes. Chronic application of N-methyl- d-aspartate sharpens the bands. The possibility that 5-amino-phosphonovaleric acid depresses tectal responsiveness was motivation for studying the effects of 5-amino-phosphonovaleric acid and N-methyl- d-aspartate applied both chronically and acutely. We evaluated tectal responsiveness to visual input by presenting flashes of light to one eye and recording responses in the ipsilateral tectal lobe. This method reveals the output of the tectal cells contralateral to the stimulated eye. These cells project via the nucleus isthmi to the opposite tectal lobe. We also mapped the receptive field dimensions of the crossed isthmotectal axons. Our results show that acute topical application of 500 μM or 1 mM N-methyl- d-aspartate dramatically increases spontaneous activity, while 100 μM N-methyl- d-aspartate causes little change. Chronic treatment with N-methyl- d-aspartate at a low dose (estimated to be in the micromolar range) shown to influence retinotectal mapping, reduces response latencies but produces no statistically significant changes in tectal cell firing rates or receptive field size. Acute application of 5-amino-phosphonovaleric acid produces complex results: 10 μM produces no changes in firing, 100 μM 5-amino-phosphonovaleric acid decreases firing, and doses of 500−100 μM increase the firing. Chronic micromolar 5-amino-phosphono-valeric acid, at doses capable of blocking activity-dependent refinement of retinotectal maps, decreases spontaneous firing and the latency of visually evoked bursts but otherwise produces no notable changes in tectal cell responses. Our data support the hypothesis that chronic treatment with N-methyl- d-aspartate and 5-amino-phos-phonovaleric acid influence retinotectal development by affecting the N-methyl- d-aspartate receptor specifically and not by causing overall changes in tectal activity.

Differential effects of GABA and bicuculline on rapidly- and slowly-adapting neurons in primary somatosensory cortex of primates.

In cortical area 3b of monkeys, responses of 71 single neurons to controlled indentations of glabrous skin were recorded before and during iontophoretic application of GABA and bicuculline methiodide (BMI), a GABA receptor antagonist. Constant amplitude indentations were applied to selected sites within the receptive fields of neurons representing the glabrous skin on the digits and palm. Profiles of response magnitudes across stimulation sites were used to quantify receptive field dimensions before and during antagonism of GABAergic inhibition. During administration of BMI, the receptive fields of 26 rapidly-adapting neurons were increased by 3-4 times their original size. Response latencies were substantially longer in the region of expansion than in the original receptive field, suggesting that expansion might be mediated by intracortical connections. The expansion of RFs onto adjacent digits after blockade of GABAergic inhibition suggests that somatotopic reorganization following digit amputations may be subserved by existing excitatory connections. The responses of slowly-adapting neurons were separated into two components, a "dynamic" response corresponding to activity elicited by the initial indenting ramp and a "static" response produced by the sustained indentation. Among 8 slowly-adapting neurons tested with BMI, the receptive fields of the dynamic response component increased to an extent that was similar to the change produced in rapidly-adapting neurons. By contrast, the static response component was rarely altered by BMI. Comparison of the responses to administration of GABA revealed that only 12 of 27 slowly-adapting neurons were inhibited in a dose-dependent manner, whereas 37 of 44 rapidly-adapting neurons exhibited significant reduction of responses in the presence of GABA. Hypotheses are proposed to explain the differential effect of BMI and GABA on slowly- and rapidly-adapting cortical neurons.

Areal influences on complex cells in cat striate cortex: stimulus-specificity of width and length summation

In single neurones recorded from the striate cortex of cats anaesthetized with N2O/O2/halothane, receptive field dimensions, length specificity and areal extent of drive were assessed for different classes of visual stimuli. Receptive fields were mapped as rectangular minimum response fields (MRFs). Spatial summation along the axis of preferred orientation was assessed: for moving bars whose length was varied (length summation); and for height variation of a square-wave grating patch against a uniform grey background, or a patch of moving texture against a stationary background of similar texture. In complementary tests a moving square-wave grating background was progressively occluded by a uniform grey foreground mask of variable height; or a mask of stationary texture of variable height progressively occluded a background of moving texture. In parallel measurements, the width of grating or textured patches or masks was varied whilst maintaining height constant. Broadly speaking, the areal influence of each class of stimulus was comparable, and distinct from extra-receptive field phenomena in evoking responses from within the receptive field, but not from surrounding areas. The masking paradigm provided the most sensitive measure of receptive field height and width. However, in some neurones length summation, the degree of end-stopping, and the directional bias depended critically on the stimulus configuration used. Length summation tended to be more dramatic for short bars than for gratings. Length summation for texture was significantly more pronounced than for an oriented bar in special and in intermediate complex neurones. By contrast, endstopping was typically less intense for gratings than for bars, and least pronounced for texture. Because of stimulus specificity, complex neurones assigned to particular functional subgroups on the basis of their response to oriented bars may exhibit quite different patterns of behaviour for other classes of stimuli.

Quantitative measurements of receptive field changes during antagonism of GABAergic transmission in primary somatosensory cortex of cats.

In cortical area 3b of cats, responses of 76 single neurons to punctate indentations were recorded before and during iontophoretic administration of bicuculline methiodide (BMI), a GABAergic antagonist, at levels that did not affect spontaneous activity. Constant amplitude indentations were applied to selected sites along distal-proximal and radial-ulnar axes that intersected the most sensitive area in the receptive field. Profiles of response magnitudes were used to measure receptive field dimensions before and during antagonism of GABAergic inhibition. Blockade of GABAergic transmission caused receptive field dimensions of 48 rapidly-adapting neurons to increase an average 141%, or nearly 2.5 times their original size. Analysis of the spatial distribution of inhibition indicated that in-field inhibition was larger than surround inhibition. During BMI administration, response latency was significantly longer for response elicited from the expanded territory than for responses elicited from within the original receptive field, suggesting that receptive field expansion might be mediated by multisynaptic intracortical connections. The magnitude of receptive field expansion was independent of receptive field size or peripheral location. In a substantial number of neurons, however, BMI produced asymmetric expansions that extended only in the proximal direction. For 9 slowly-adapting neurons. BMI produced measurable increases in receptive field dimensions, but these changes were significantly smaller than the changes in rapidly-adapting neurons.

Directional and orientational tuning of feline striate cortical neurones: Correlation with neuronal class

In a subset of 327 simple and complex cells from the striate cortex of cats anaesthetized with N 2O/O 2/halothane, a range of receptive field properties were compared. These included directional and orientational selectivity, tuning and symmetry; endstopping; receptive field dimensions; length summation; texture sensitivity; ocular dominance; and resting discharge levels. These properties were related to neuronal class (simple or complex) and to the special, intermediate and standard subdivisions of the complex cell category. Special complex cells showed a high incidence of direction selectivity, were less sharply tuned for orientation, more commonly endstopped, more strongly binocular, tended to have higher resting discharge levels and exhibited greater sensitivity to motion of randomly textured patterns than the other classes of neurones. The remaining classes of complex cells, together with simple cells, were more commonly direction-biased or bidirectional, and more selective for orientation than special complex cells. Standard complex cells were marginally more symmetrically tuned for orientation than the other groups. Simple cells represented the most sharply orientation tuned neurones in the cortex; unlike complex cells of all groups they were insensitive to texture motion, generally had lower levels of maintained discharge, and showed least integration of inputs between the two eyes. Assessed by appropriate measures (minimum response fields in special complex cells; length summation in standard complex cells), standard complex cells had significantly larger receptive fields than special complex cells.

Uncertainty relation for resolution in space, spatial frequency, and orientation optimized by two-dimensional visual cortical filters.

Two-dimensional spatial linear filters are constrained by general uncertainty relations that limit their attainable information resolution for orientation, spatial frequency, and two-dimensional (2D) spatial position. The theoretical lower limit for the joint entropy, or uncertainty, of these variables is achieved by an optimal 2D filter family whose spatial weighting functions are generated by exponentiated bivariate second-order polynomials with complex coefficients, the elliptic generalization of the one-dimensional elementary functions proposed in Gabor's famous theory of communication [J. Inst. Electr. Eng. 93, 429 (1946)]. The set includes filters with various orientation bandwidths, spatial-frequency bandwidths, and spatial dimensions, favoring the extraction of various kinds of information from an image. Each such filter occupies an irreducible quantal volume (corresponding to an independent datum) in a four-dimensional information hyperspace whose axes are interpretable as 2D visual space, orientation, and spatial frequency, and thus such a filter set could subserve an optimally efficient sampling of these variables. Evidence is presented that the 2D receptive-field profiles of simple cells in mammalian visual cortex are well described by members of this optimal 2D filter family, and thus such visual neurons could be said to optimize the general uncertainty relations for joint 2D-spatial-2D-spectral information resolution. The variety of their receptive-field dimensions and orientation and spatial-frequency bandwidths, and the correlations among these, reveal several underlying constraints, particularly in width/length aspect ratio and principal axis organization, suggesting a polar division of labor in occupying the quantal volumes of information hyperspace.(ABSTRACT TRUNCATED AT 250 WORDS)