Visual Neuroscience ~2006!, 23, 721–728. Printed in the USA. Copyright © 2006 Cambridge University Press 0952-5238006 $16.00 DOI: 10.10170S0952523806230049
The maximum range and timing of excitatory contextual modulation in monkey primary visual cortex
D.M. ALEXANDER 1 and J.J. WRIGHT 2 1 2
Faculty of Information Technology, University of Technology, Sydney, Australia The Liggins Institute, University of Auckland, Auckland, New Zealand
(Received November 30, 2004; Accepted March 27, 2006!
Abstract Contextual modulations of receptive field properties by distal stimulus configurations have been shown for a variety of stimulus paradigms. A survey of excitatory contextual modulation data for V1 shows the maximum scale of interactions, measured in terms of distance in V1, to be between 10 mm and 30 mm. Different types of excitatory contextual modulation in V1 occur throughout the interval of 40–250 ms after stimulus delivery. This window provides opportunity for global propagation of visual contextual information to a subset of V1 neurons, via several routes within the visual system. We propose a number of experiments and analyses to confirm the results from this empirical survey. Keywords: Contextual modulation, Monkey, Primary visual cortex
integrating global information with spatial precision ~Lee et al., 1998!. A number of findings have indicated that excitatory contextual modulation of V1 neurons can result from stimuli that cover extensive regions of the visual field. This paper is concerned with the maximum extent of excitatory contextual interactions, assuming there is a range of interaction scales, with most interactions being over shorter distances. A survey of the contextual modulation literature and unpublished data sets indicates that excitatory contextual modulation can be elicited in V1 neurons by distal stimulus configurations corresponding to cortical distances of up to 10 mm to 30 mm, depending on the stimulus paradigm. Some stimulus paradigms show no evidence of a decline in excitatory contextual modulation with distance. A survey of the contextual modulation literature shows that excitatory contextual modulation can occur over a range of time-scales, and is generally delayed with respect to the initial feedforward response. The modulation can extend up to 250 ms post-stimulus, providing ample opportunity for widely separated neurons in V1 to interact functionally. The experiments cited in this paper make use of a variety of definitions of neuronal activitation by localized stimuli in the visual field. The cRF is synonymous with “minimum response field,” and it is tested with a small probe ~usually a spot, or short slit of light!. The cRF of a neuron has been defined as that part of the visual field from which a stimulus can elicit strong activity. The aggregate classical receptive field ~AcRF! is that section of the visual field that influences activity recorded at multiple neurons. The cRF differs from the “summation field,” which is measured by systematically increasing the diameter of a stimulus. The summation field is generally larger than the respective cRF. The measures of excitatory contextual modulation cited in this
Introduction When a stimulus is presented outside the classical receptive field ~cRF!, no strong neuronal response is observed in the primary visual cortex ~V1! ~Hubel & Wiesel, 1962!. The cRF of V1 neurons have been described as spatially localized filters, selectively tuned for orientation and spatial frequency ~Schiller et al., 1976; Movshon et al., 1978!. The filters are regarded as spatially localized, because stimuli of small visual angle appear both necessary and sufficient to achieve an optimal response. The tuning of the localized filters in V1 has been considered largely independent of the perceptual context ~De Valois et al., 1979!. A number of excitatory contextual modulation effects, not confined to the cRF, have been found in the monkey. Excitatory contextual modulation refers to increases in the activity of neurons that depends on stimulus features outside the cRF. Contextual modulation is not dependent on specific cues, such as texture boundaries, but can be found for a diverse range of stimulus types ~Zipser et al., 1996!. These effects suggest a second set of mechanisms at work, distinct from cRF processes ~Lamme et al., 1998a!. Contextual modulation effects imply that V1 is involved in complex visual perception and participates in interpretation of the visual scene ~Li et al., 2000; Zipser et al., 1996!. In addition to serving as a way-station in which local pre-processing takes place, V1 may act as a high resolution buffer; within this framework V1 may be involved in many levels of visual processing, including
Address correspondence and reprint requests to: Dr. David Alexander, Faculty of Information Technology, University of Technology, Sydney, P. O. Box 123, Broadway, 2007, NSW, Australia. E-mail:
[email protected]
721
722
D.M. Alexander and J.J. Wright
paper encompass a wide range of stimulus types. In all cases cited, the distances refer to the radius of the stimulus ~or half-width or half-length where appropriate!.
Extent of contextual modulation in V1 Modulatory interactions can be measured in V1 in response to distal stimulus features in the visual field. Distance estimates in this paper are given in terms of distances in V1 rather than distances in the visual field because the cortical magnification factor ~CMF! varies so markedly across the extent of V1 and other visual areas. Cortical distances in V1 therefore more closely reflect anatomical constraints on cortical visual processing. The focus of this paper is the maximum of the range of modulatory influences. The bulk of interactions are much shorter than this maximum, as a matter of definition. The V1 CMF equations of Tootell et al. ~1988! were used to estimate cortical distances. Specifically, the CMF estimate of the vertical and horizontal meridian was calculated for a given eccentricity, and then the two values averaged. Perusal of Fig. 1 shows that a curve with values positioned mid-way between “Tootell VM” and “Tootell HM” constitutes a median-like compromise given the CMF estimates provided by a number of authors ~see Table 2
for actual numerical values used for the CMF estimates in this paper!. Cortical distances of excitatory contextual modulation were estimated where contextual modulation data and reliable estimates of eccentricity of the cRF were available for individual neurons. Previously published results and data supplied by researchers enabled estimates of the maximum cortical distance of excitatory contextual modulation to be estimated for six different stimulus paradigms. A summary of the survey results for the macaque is presented in Table 1. 1. Relative luminance: the luminance of a foreground square is held constant, whereas the luminance of the surround is varied; 25% of V1 neurons have tuning curves for relative luminance ~Kinoshita & Komatsu, 2001!. The foreground square can be many times RF size. 2. Surround only texture: most neurons will respond to a surround only texture at above baseline levels. Rossi et al. ~2001! have shown that this effect extends to “hole” sizes that are up to 12 ⫻ RF size. The percentage of neurons responding to surround only textures falls off roughly monotonically with stimulus size, with some suggestion of a long tail.
Fig. 1. Estimates of CMF from various authors ~Brewer et al., 2002; Van Essen et al., 1984; Tootell et al., 1988; Schwartz, 1977; Gattass et al., 1981!. Mp and Me refer to estimates of the CMF along iso-polar and iso-eccentric lines, respectively; VM and HM refer to estimates of the CMF along the horizontal and vertical meridians, respectively.
Excitatory contextual modulation in monkey V1
723
Table 1. Maximum spatial extent of excitatory contextual modulation for different stimuli. Maximum extent is provided in terms of the radial distance of the distal stimulus feature. These distances in the visual field are converted into distances in V1, where appropriate data were available. Otherwise distance is provided in terms of multiples of the length of the RF. All maximum stimulus sizes are given in degrees of visual field of the radius (or half-width or half-length) of the stimulus
Effect
Eccent. of RF8
Max. stimulus8
Max. extent
Reference
Summation to low contrast gratings Summation for low contrast lines -for lines in textured background Distal modulation of color response Border ownership Response to Kanizsa squares Response to glass patterns Relative luminance Center response to texture defined band Figure versus ground enhancement Surround only texture Curve tracing Artificial scotomas
2–5 2–7 2–7 2–5 0.6– 6 0.6– 4 2–25 0.9–8 3– 4 0.5– 6 2– 6 1– 6 3–22
5 1.5 1.5 7 5 1 2.5 5 3 3.5 6 6.0 7.5
6 ⫻ RF 12 ⫻ RF 7 ⫻ RF 18.1 mm 27.9 mm 6 ⫻ RF 5 ⫻ RF 24.1 mm 6 ⫻ RF 5 ⫻ RF 27.4 mm ;30 mm 10.3 mm
Sceniak et al. ~1999! Kapadia et al. ~1999! Kapadia et al. ~1999! Wachtler et al. ~2003! Zhou et al. ~2000! Lee & Nguyen ~2001! Smith et al. ~2002! Kinoshita & Komatsu ~2001! Lee et al. ~1998! Lee et al. ~1998! Rossi et al. ~2001! Roelfsema et al. ~1998! Fiorani et al. ~1992!
3. Distal modulation of color response: distally placed patches of color can modulate the response of V1 neurons to a color stimulus. Most interactions are inhibitory, but a substantial minority is excitatory ~Wachtler et al., 2003!. 4. Border ownership: the activity of some neurons in V1 can be modulated when their RFs lie on the edge of a foreground figure ~Zhou et al., 2000!. This modulation depends on the configuration of the foreground figure compared to the background, even where the local information in the RF is identical. 5. Curve tracing: the activity of V1 neurons can be modulated when the RF of a neuron lies on a curve that the monkey is attending to ~Roelfsema et al., 1998!. The monkey is trained to fixate on one end of the curve and to ignore a similarly positioned distracter curve. 6. Artificial scotoma: very large, low contrast bars activate a neuron’s RF, even when a large mask ~5–10 times RF size! is placed over the RF ~Fiorani et al., 1992!. The data-points used to calculate maximum cortical distance are shown in Table 2. The results show that stimulus configurations ~radii! covering cortical distances in V1 of 10 mm to 30 mm can modulate activity of V1 neurons in an excitatory fashion; the estimated maximum distances vary depending on the nature of the stimulus paradigm. The six measures of cortical distance provided here involve varying combinations of sub-cortical, intra-striate, and extra-striate interactions; but all involve measurable effects in V1. Are these measures characteristic of the maxima of the range, or are they spurious outliers? Perusal of published distributions ~Fiorani et al., 1992, Fig. 5; Zhou et al., 2000, Fig. 19; Rossi et al., 2001; Table 1! and the additional data points given in Table 2 indicate that for the most of the stimulus types the calculated values cannot be mere outliers, but are characteristic of the maxima of the distributions. For the curve tracing effect, there is no evidence of a weakening of the excitatory modulation over any of
the distances tested ~Khayat et al., 2004, Fig. 2e!. The size tuning curves for relative luminance and texture-defined borders are likewise flat ~Kinoshita & Komatsu, 2001; Lee et al., 1998!. Other measures of the maximum range of excitatory contextual modulation in this survey have been expressed in units of RF size because in these cases the cortical distances could not be estimated from the available data ~see Table 1!. These measures are not always directly comparable to each other, because the definitions of RF size vary from paper to paper; we therefore use the generic term RF. The measures are provided here as supporting evidence for the cases where cortical distances were estimated, to show the generality of these effects across stimulus paradigms. The peak area summation to low contrast gratings can be as large as 6 times the peak area summation to high contrast gratings ~Sceniak et al., 1999, diagram 3!. The maximum response to Glass patterns reported in Smith et al. ~2002! was 5 ⫻ RF size. For V1 neurons that respond to the illusory contours of Kanizsa squares, the maximum distance reported was 6 ⫻ RF size ~Lee & Nguyen, 2001!. For neurons that had a center response to large, textured defined shapes, the maximum distance reported was also 6 ⫻ RF size ~Lee et al., 1998!. Figure versus ground enhancement has been demonstrated for figures up to 5 ⫻ RF size ~Lee et al., 1998; Lee & Mumford, 2003!. Summation to low contrast lines is up to 12 ⫻ RF size ~Kapadia et al., 1999!. Estimates of the cortical extent of signal processing in V1 generally use equations based on the CMF. However, the CMF is not an accurate method to estimate cortical distances for large stimuli, because the stimulus representation is highly elongated towards the fovea. For example, a circular stimulus whose center is positioned at an eccentricity of 2.38, and has a radius of 2.38, stretches all the way to centre of the field of vision, by definition. This corresponds to a cortical distance of ;20 mm in V1 ~Tootell et al., 1988!, compared to an estimate of ;10 mm based on the CMF. Two factors suggest the distance along the elongated axis may be a more appropriate estimate of the maximum cortical distance involved. First, intrinsic patchy connections are themselves elongated in the foveal direction, matching the elongation of
724
D.M. Alexander and J.J. Wright Table 2. Calculation of cortical distances given in Table 1. All data (the RF eccentricities and stimulus radii) are by personal communication, from the researcher given in the first column, except where a reference is given. The estimates of CMF are from Tootell et al. (1988), and are an average of the estimates of the horizontal and vertical meridians for the given eccentricity. A different method was used for the experimental paradigm of curve tracing. In these experiments, the curved line stretches from the fixation point to the distal RF, so the cortical distance estimate was based directly on the distance from the centre of the visual field to the RF centre, rather than the CMF Researcher
Stimulus type
Rüdiger von der Heydt
Border ownership
Hidehiko Komatsu
Relative luminance
Andrew Rossi
Surround only texture
Thomas Wachtler
Distal colour response
Pieter Roelfsema
Curve tracing
Fiorani et al. ~1992!
Artificial scotomas
the stimulus mapping to V1 ~Angelucci et al., 2002!. Second, studies looking at localization of inhibitory effects from surround stimuli do not report interactions with retinotopic angle in the effectiveness of surround inhibition ~Cavanaugh et al., 2002!, as expected if the longer cortical distances associated with the foveaward direction reduced the effectiveness of horizontal interactions. The CMF was used in Table 2 for consistency with previous studies, but it may underestimate the maximum cortical distances involved in distal excitation by as much as a factor of two. The timing of contextual modulation in V1 Presenting simple, isolated, high contrast stimuli on a grey background produces a strong initial response in V1 neurons between 40– 60 ms after stimulus onset, and it has been suggested that this initial peak carries more information than the later response components ~Heller et al., 1995; Gershon et al., 1998; Li et al., 2000; Angelucci & Bullier, 2003!. A number of experiments suggest the initial impulse response is reduced under different stimulus conditions. First, low contrast stimuli do not produce an early, strong impulse response ~Kapadia et al., 1999, figure 5c!. Instead, the neuronal response has a longer time-course, peaking about 110 ms after the stimulus. A similar effect occurs for high contrast bars embedded within a field of randomly oriented bars. The impulse response is greatly reduced and there is another, almost equal, peak in the response at around 140 ms ~Kapadia et al., 1999, figure 5d!. The reduction of the impulse response scales with the size of the surrounding field of randomly oriented bars ~Li et al., 2000, figures 8b and 9d!. Third, RF activation prior to a saccade reduces the maximum response amplitude to a stimulus entering the RF of
RF eccen.
Stimulus rad.
CMF
V1 distance
18 1.68 2.18 2.58 2.38 2.48 2.58 2.68 2.68 4.248 4.58 5.28 5.758 5.78 5.68 5.58 118 128 38
48 48 58 48 68 68 68 68 68 68 68 78 5.758 5.78 5.68 5.58 7.58 6.38 2.58
6.97 mm08 5.60 mm08 4.82 mm08 4.33 mm08 4.56 mm08 4.44 mm08 4.33 mm08 4.22 mm08 4.22 mm08 3.01 mm08 2.88 mm08 2.58 mm08 N0A N0A N0A N0A 1.38 mm08 1.28 mm08 3.85 mm08
27.9 mm 22.4 mm 24.1 mm 17.3 mm 27.4 mm 26.7 mm 26.0 mm 25.3 mm 25.3 mm 18.1 mm 17.3 mm 18.0 mm ;30 mm ;30 mm ;30 mm ;30 mm 10.3 mm 8.0 mm 9.6 mm
the same neuron, post-saccade ~Richmond et al., 1999; Gawne & Woods, 2003!. Under normal viewing conditions, low contrasts, complex stimulus fields and saccade-induced changes are the norm. If the impulse response is less prominent under normal viewing conditions, then less visual information is likely to be transmitted during this window ~Richmond et al., 1999! and the relative importance of the later components is likely to be increased. Table 3 presents evidence that the relevant time-course of V1 neuronal activity extends from 40 to 250⫹ ms, with unique components appearing throughout this interval. There is some evidence that surround effects alone can drive a V1 neuron at the earliest of these latencies ~Sugita, 1999!. The more complex the stimulus feature, the more delayed the response in V1. This is apparent in the latencies to figure versus ground enhancement ~Lamme, 1995!, texture defined boundaries ~Lee et al., 1998!, border ownership ~Zhou et al., 2000!, illusory contours ~Lee & Nguyen, 2001! and relative luminance ~Kinoshita & Komatsu, 2001!. A window from 40 ms to 250 ms after stimulus onset provides ample opportunity for the widespread effects of contextual modulation. These influences occur via several different anatomical systems. For large contextual modulation distances, is it meaningful to give estimates of cortical distance in V1? It is possible that, outside the distance of surround interactions, modulatory influences in V1 are unrelated to the topography of V1 and hence to distances in the visual field. Several findings, however, suggest that delays in response onset in V1 are systematically related to distance in the visual field. A majority of V1 neurons show increasing delays in the onset of inhibition with increasing size of the high contrast grating ~Bair et al., 2003, figure 7a!. The effect has been tested to a cortical distance of 12 mm. An increase in delay with stimulus
Excitatory contextual modulation in monkey V1
725
Table 3. Summary of time-scales involved in the excitatory modulation of V1 neurons. Times in center column indicate the earliest response due to the component in averaged data, or the earliest time where a range of onsets was reported. Where a “⫹” symbol is shown, the times are the averaged responses relative to the initial response of the neuron, or other relevant measure
Event Stimulus onset Impulse response to high contrast bar Bar obscured by foreground patch Delay when stimulus enters due to saccade Texture boundaries Border ownership Structure from motion Center axis response to figure figure versus ground enhancement at border Texture figure enhanced by black background Dot and luminance mask enhancement Cue invariant boundary responses Figure versus ground enhancement over whole figure Illusory contours ~Kanizsa squares! Modulation by number of figures Response to surround only texture Peak response to low contrast gratings Curve tracing effect Second response peak to bar in field of bars Stimulus-locked gamma versus induced gamma Asymptote of mutual information post-saccade Response to relative luminance stabilizes
distance is also seen in responses to surround only textures ~Rossi et al., 2001, figure 8c and 9!, although this effect asymptotes at a stimulus radius of approximately 2.58. The modularity effects of attention on V1, measured for the curve tracing effect, reveal increases in delay with retinotopic distance ~Roelfsema & Spekreijse, 1999!. This latter experiment used simultaneous recordings from multiple electrodes, so that a range of delay versus distance measurements was available for each individual stimulus presentation ~Super & Roelfsema, 2005!. Routes of contextual modulation in V1 An important route for contextual modulation in V1 is via extrastriate areas. Extra-striate areas are able to provide widespread modulatory influence because they have large RF sizes compared to V1 and the high conduction velocities of myelinated corticocortical fibers mean the entire visual cortex is temporally compact ~Hupe et al., 2001; Girard et al., 2001!. The role of extrastriate feedback in contextual modulation is also indicated by timing of activations; the activation of extrastriate areas generally precedes the contextual modulation measured in V1 ~Zipser et al., 1996!. Feedforward connections, in this view, determine RF properties while horizontal and feedback connections mediate perceptual organization and attention ~Lamme et al., 1998a!. The timing of conscious visual perception seems to more closely match that of contextual modulation than the timing of purely feedforward events in V1 ~Lamme et al., 1998b!. Extrastriate feedback may also play a greater role in activation when the stimuli are of low salience ~Hupe et al., 2001!. The middle temporal ~MT! visual area serves here as brief illustration of the role of extra-striate feedback in V1 contextual
Time ~ms!
Reference
0 40– 60 ⫹0 ⫹3 60 60 69 70 80 80 80 100 100 100 100 105 110 130 140 ⫹100 400 500
Sugita ~1999! Gawne & Martin ~2002! Lee et al. ~1998! Zhou et al. ~2000! Lamme et al. ~1998b! Lee et al. ~1998! Lamme et al. ~1998b! Lee et al. ~1998! Li et al. ~2000! Lee et al. ~1998! Lamme ~1995! Lee & Nguyen ~2001! Lamme et al. ~2000! Rossi et al. ~2001! Kapadia et al. ~1999! Roelfsema et al. ~2003! Kapadia et al. ~1999! Juergens et al. ~1999! Richmond et al. ~1999! Kinoshita & Komatsu ~2001!
modulation. Receptive field size in MT is about 10 times larger than in V1 at all eccentricities ~Albright & Desimone, 1987!. Small focal injections of tracer into V1 indicate that the AcRF sizes of the feedback fields in MT are 21-fold larger than the AcRF size of neurons at the V1 injection sites ~Angelucci et al., 2002!. These feedback connections are an obvious substrate for the integration of global signals into V1 ~Bullier et al. 2001!. There is also evidence that the timing of MT’s feedback to the upper layers of V1 is coincident with feedforward signal arriving via layer 4 of V1 ~Hupe et al., 2001!. MT receives visual information directly from the colliculus, pulvinar and possibly also the LGN ~Yukie & Iwai, 1981; Standage & Benevento, 1983; Ungerleider et al., 1984; Rodman et al., 1990; Ffytche et al., 1995!. These considerations are consistent with the timing of contextual modulation effects listed in Table 3, showing that some types of contextual modulation can occur concomitantly with the early neuronal response— even though other types are delayed substantially with respect to feedforward inputs. Some types of excitatory contextual modulation, such as flank facilitation ~Kapadia et al., 1999!, are proposed to be mediated by patchy horizontal fibers intrinsic to V1. Surround only textures show relatively long delays in onset of contextual modulation, delays that increase linearly with increasing distance between the surround edge and the RF. Patchy horizontal fibers in V1 therefore also potentially play a role in these effects ~Rossi et al., 2001!. However, the short range of the patchy fibers, in combination with the smaller RF sizes in V1 ~Hubel & Wiesel, 1968!, and slow conduction velocities of V1 patchy horizontal fibers compared to extra-striate feedback connections ~Angelucci & Bullier, 2003!, have lead most researchers to conclude that extra-striate feedback is the mechanism by which contextual modulation occurs over
726 large distances in V1 ~e.g., Lamme et al., 1998a; Roelfsema & Spekreijse, 1999; Angelucci et al., 2002; Hupe et al., 2001; Lee & Mumford, 2003!. Widespread contextual interactions can also occur prior to visual signals reaching V1. Two examples of excitatory influence from outside the classical surround will be given here by way of illustration. First, widespread excitatory interactions occur as early as the ganglion cells in the retina, demonstrated by the shift effect in the monkey ~and the related periphery effect in the cat; McIlwain, 1964; Fischer & Kruger, 1974; Kruger et al., 1975; Kruger, 1977!. The shift effect involves presentation of an annular grating to the visual field well outside the range of surround interactions—having an inner diameter of 178—that modulates RF activity ~Kruger, 1977!. Ikeda and Wright ~1972! found the periphery effect arises in retinal ganglion cells that are transient responding, with fast axonal conduction velocities, and similar retinal interactions have been measured for the shift effect ~Kruger, 1977!. Second, the RFs of monkey LGN neurons can be modulated by a rotating, radial grating whose inner radius begins just outside the extent of classical surround interactions 1 ~Marrocco et al., 1982!. Most modulations are inhibitory, but 6% are excitatory. This extra-surround modulation of LGN activations disappears during cryogenic blocking of V1, suggesting that feedback from V1 to LGN plays a role in this effect ~Marrocco et al., 1982!.
Discussion A survey of published and unpublished data shows that excitatory contextual modulation can be elicited by distal stimulus configurations at distances in V1 of up to 10 mm to 30 mm, depending on the stimulus paradigm. Some stimulus paradigms show no evidence of a maximum range of effect. The timing of the contextual modulation in V1 is generally delayed with respect to the initial feed-forward response, and can extend up to 250 ms depending on the stimulus paradigm. This window provides ample opportunity for widespread propagation of contextual information to V1 neurons via a range of anatomical routes. A number of experimental and analytic approaches are suited to the study of widespread excitatory contextual modulation in V1. In the following paragraphs we discuss manipulations of the effects of delay on contextual modulation, and stimulus paradigms under which the maximum extent of excitatory contextual modulation can be further explored. Additionally, we discuss the usefulness of Fourier-related analytic tools in detecting temporally and spatially extended patterns of activation. Contextual information that is delayed with respect to feedforward processes can play an important functional role because during normal perception, visual context gives rise to ongoing, background neuronal activity within which specific feature discriminations take place. An interesting experimental approach would therefore be to manipulate foreground stimuli such that contextual modulation in V1 that is delayed with respect to feedforward inputs to V1 by x ms has the opportunity to simultaneously interact with feedforward inputs. This effect can be achieved by delaying onset of the foreground stimulus by x ms with respect 1 This LGN effect is distinct from the shift effect, which can also be measured in LGN neurons but produces a different distribution of activations among the populations of LGN cell types ~Marrocco et al., 1982!.
D.M. Alexander and J.J. Wright to the contextual stimulus. The effects of retinotopic distance on delay of contextual modulation in V1 could also be incorporated into this stimulus manipulation. It would also be of interest to verify the effects of distance on delay for a wider range of stimulus paradigms, including relative luminance, distal modulation of the color response, and artificial scotoma. Testing the maximum extent of horizontal interactions in V1 requires somewhat different experiments to the standard paradigms ~e.g. small, quasi-static, high contrast gratings!. Tuning properties of neurons can be further explored using very large, low contrast, low spatial frequency, and texture-defined stimuli. “Very large” here means up to the size of the entire visual field. Low contrast stimuli are known to induce larger summation fields ~Sceniak et al., 1999! and texture defined figures ~e.g. Lee et al., 1998! have maxima in size-tuning curves that are many times RF size, and are size-invariant with respect to their border response. We suggest that under some combination of these conditions, the maximum extent of excitatory contextual modulation in V1 will be found to be much greater than is currently assumed. These properties will be sub-optimal for many neurons, as measured according to the definition of the cRF. However, novel approaches to studying tuning properties under conditions where contextual information is present may yield a different picture of neuronal selectivity. Highly apposite in this regard are the extra-RF modulations of orientation preference have been demonstrated in V1 of the ferret ~Basole et al., 2003!. In particular, line-length ~tested up to 108! and speed of motion ~up to 10080s! can strongly alter the orientation preference of V1 cells when activity is measured over 2 seconds of the stimulus presentation. Related findings have been made in monkey V1 for speed of motion ~Geisler et al., 2001!, for changes between the early and late components of the orientation response to large gratings ~Ringach et al., 1997! and for texture defined boundaries ~Lee et al., 1998!. The findings made for the ferret should therefore generalize to the monkey, and show that widespread contextual influences are capable of strongly altering the orientation preference of neurons in monkey V1. The relevance of cortical dynamics is a controversial topic in visual neuroscience, but findings in relation to spatial and temporal integration offer further novel avenues to analyze experimental data. The majority of V1 studies measure stimulus locked activity, which is distinct from signals that are stimulus induced but not phase-locked to the stimulus. Measurement of stimulus-induced activity requires activity statistics to be calculated in the frequency domain ~i.e., via Fourier and related methods! because in the time domain non-phase-locked activity is removed by averaging. We have recently proposed that wave propagation, via polysynaptic routes within V1, offers a complementary mechanism for widespread horizontal propagation of information ~Alexander et al., 2004; Wright et al., 2006!. Stimulus induced gamma activity peaks some 100 ms after the peak in stimulus locked gamma ~Juergens et al., 1999!. Whereas the duration and amplitude of gamma-band episodes decrease after this time, the episodes continue for as long as the stimulus is present; this has been measured for up to 3 seconds ~Rols et al., 2001!. Traveling waves measured in response to large grating stimuli typically exceed distances of 8 mm in V1 ~the maximum extent of the array; Eckhorn et al., 2001! and have a modal velocity of 0.4 m0s ~Eckhorn et al., 2004!. Similar findings in the rabbit primary visual cortex are reported ~Freeman & Barrie, 2000!. To the extent that this dynamical activity is functionally important to vision, then averaging activity statistics when not in the frequency
Excitatory contextual modulation in monkey V1 domain may underestimate the latency, duration, spatial extent, and horizontal velocity of the neuronal response.
Acknowledgments The authors thank Valentino Braitenberg and Almut Schüz for their helpful comments in relation to cortical anatomy and Rüdiger von der Heydt, Hidehiko Komatsu, Andrew Rossi, Thomas Wachtler and Pieter Roelfsema who kindly provided data used in this paper.
References Albright, T.D. & Desimone, R. ~1987!. Local precision of visuotopic organization in the middle temporal area ~MT! of the macaque. Experimental Brain Research 65, 582–592. Alexander, D.M., Bourke, P.D., Sheridan, P., Konstandatos, O. & Wright, J.J. ~2004!. Intrinsic connections in tree shrew V1 imply a global to local mapping. Vision Research 44, 857–876. Angelucci, A. & Bullier, J. ~2003!. Reaching beyond the classical receptive field of V1 neurons: Horizontal or feedback axons? The Journal of Physiology ~Paris! 97, 141–154. Angelucci, A., Levitt, J.B., Walton, E.J., Hupe, J.M., Bullier, J. & Lund, J.S. ~2002!. Circuits for local and global signal integration in primary visual cortex. Journal of Neuroscience 22, 8633–8646. Bair, W., Cavanaugh, J.R. & Movshon, J.A. ~2003!. Time course and time-distance relationships for surround suppression in Macaque V1 neurons. Journal of Neuroscience 23, 7690–7701. Basole, A., White, L.E. & Fitzpatrick, D. ~2003!. Mapping multiple features in the population response of visual cortex. Nature 423, 986–990. Brewer, A.A., Press, W.A., Logothetis, N.K. & Wandell, B.A. ~2002!. Visual areas in Macaque cortex measured using functional magnetic resonance imaging. Journal of Neuroscience 22, 10416–10426. Bullier, J., Hupe, J.M., James, A.C. & Girard, P. ~2001!. The role of feedback connections in shaping the responses of visual cortical neurons. Progress in Brain Research 134, 193–204. Cavanaugh, J.R., Bair, W. & Movshon, J.A. ~2002!. Selectivity and spatial distribution of signals from the receptive field surround in Macaque V1 neurons. Journal of Neurophysiology 88, 2547–2556. De Valois, K.K., De Valois, R.L. & Yund, E.W. ~1979!. Responses of striate cortex cells to grating and checkerboard patterns. Journal of Physiology 291, 483–505. Eckhorn, R., Bruns, A., Saam, M., Gail, A., Gabriel, A. & Brinksmeyer, H.J. ~2001!. Flexible cortical gamma-band correlations suggest neural principles of visual processing. Visual Cognition 8, 519–530. Eckhorn, R., Gail, A., Bruns, A., Gabriel, A., Al-Shaikhli, B. & Saam, M. ~2004!. Neural mechanisms of visual associative processing. Acta Neurobiologiae Experimentalis 64, 239–252. Ffytche, D.H., Guy, C.N. & Zeki, S. ~1995!. The parallel visual motion inputs into areas V1 and V5 of human cerebral cortex. Brain 118, 1375–1394. Fiorani, Junior M., Rosa, M.G., Gattass, R. & Rocha-Miranda, C.E. ~1992!. Dynamic surrounds of receptive fields in primate striate cortex: A physiological basis for perceptual completion? Proceedings of the National Academy of Sciences of the United States of America 89, 8547–8551. Fischer, B. & Kruger, J. ~1974!. The shift-effect in the cat’s lateral geniculate neurons. Experimental Brain Research 21, 225–227. Freeman, W.J. & Barrie, J.M. ~2000!. Analysis of spatial patterns of phase in neocortical gamma EEGs in rabbit. Journal of Neurophysiology 84, 1266–1278. Gattass, R., Gross, C.G. & Sandell, J.H. ~1981!. Visual topography of V2 in the Macaque. Journal of Comparative Neurology 201, 519–539. Gawne, T.J. & Martin, J.M. ~2002!. Responses of primate visual cortical neurons to stimuli presented by flash, saccade, blink, and external darkening. Journal of Neurophysiology 88, 2178–2186. Gawne, T.J. & Woods, J.M. ~2003!. The responses of visual cortical neurons encode differences across saccades. Neuroreport 14, 105–109. Geisler, W.S., Albrecht, D.G., Crane, A.M. & Stern, L. ~2001!. Motion direction signals in the primary visual cortex of cat and monkey. Visual Neuroscience 18, 501–516.
727 Gershon, E.D., Wiener, M.C., Latham, P.E. & Richmond, B.J. ~1998!. Coding strategies in monkey V1 and inferior temporal cortices. Journal of Neurophysiology 79, 1135–1144. Girard, P., Hupe, J.M. & Bullier, J. ~2001!. Feedforward and feedback connections between areas V1 and V2 of the monkey have similar rapid conduction velocities. Journal of Neurophysiology 85, 1328–1331. Heller, J., Hertz, J.A., Kjaer, T.W. & Richmond, B.J. ~1995!. Information flow and temporal coding in primate pattern vision. Journal of Computational Neuroscience 2, 75–93. Hubel, D.H. & Wiesel, T.N. ~1962!. Receptive fields, binocular interaction and functional architecture in the cat’s visual cortex. Journal of Physiology ~London! 160, 106–154. Hubel, D.H. & Wiesel, T.N. ~1968!. Receptive fields and functional architecture of monkey striate cortex. Journal of Physiology ~London! 195, 215–243. Hupe, J.M., James, A.C., Girard, P., Lomber, S.G., Payne, B.R. & Bullier, J. ~2001!. Feedback connections act on the early part of the responses in monkey visual cortex. Journal of Neurophysiology 85, 134–145. Ikeda, H. & Wright, M.J. ~1972!. The periphery effect and its relation to the receptive field organization of ‘transient’ retinal ganglion cells. Journal of Physiology 226, 81–82. Juergens, E., Guettler, A. & Eckhorn, R. ~1999!. Visual stimulation elicits locked and induced gamma oscillations in monkey intracorticaland EEG-potentials, but not in human EEG. Experimental Brain Research 129, 247–259. Kapadia, M.K., Westheimer, G. & Gilbert, C.D. ~1999!. Dynamics of spatial summation in primary visual cortex of alert monkeys. Proceedings of the National Academy of Sciences of the United States of America 96, 12073–12078. Khayat, P.S., Spekreijse, H. & Roelfsema, P.R. ~2004!. Correlates of transsaccadic integration in the primary visual cortex of the monkey. Proceedings of the National Academy of Sciences of the United States of America 101, 12712–12717. Kinoshita, M. & Komatsu, H. ~2001!. Neural representation of the luminance and brightness of a uniform surface in the Macaque primary visual cortex. Journal of Neurophysiology 86, 2559–2570. Kruger, J. ~1977!. The shift-effect in the lateral geniculate body of the rhesus monkey. Experimental Brain Research 29, 387–392. Kruger, J., Fischer, B. & Barth, R. ~1975!. The shift-effect in retinal ganglion cells of the rhesus monkey. Experimental Brain Research 23, 443– 446. Lamme, V.A. ~1995!. The neurophysiology of figure-ground segregation in primary visual cortex. Journal of Neuroscience 15, 1605–1615. Lamme, V.A., Super, H., Landman, R., Roelfsema, P.R. & Spekreijse, H. ~2000!. The role of primary visual cortex ~V1! in visual awareness. Vision Research 40, 1507–1521. Lamme, V.A., Super, H. & Spekreijse, H. ~1998a!. Feedforward, horizontal, and feedback processing in the visual cortex. Current Opinion in Neurobiology 8, 529–535. Lamme, V.A., Zipser, K. & Spekreijse, H. ~1998b!. Figure-ground activity in primary visual cortex is suppressed by anesthesia. Proceedings of the National Academy of Sciences of the United States of America 951, 3263–3268. Lee, T.S. & Mumford, D. ~2003!. Hierarchical Bayesian inference in the visual cortex. Journal of the Optical Society of America, A, Optics, Image Science, & Vision 20, 1434–1448. Lee, T.S., Mumford, D., Romero, R. & Lamme, V.A. ~1998!. The role of the primary visual cortex in higher level vision. Vision Research 38, 2429–2454. Lee, T.S. & Nguyen, M. ~2001!. Dynamics of subjective contour formation in the early visual cortex. Proceedings of the National Academy of Sciences of the United States of America 98, 1907–1911. Li, W., Their, P. & Wehrhahn, C. ~2001!. Neuronal responses from beyond the classical receptive field in V1 of alert monkeys. Experimental Brain Research 139, 359–371. Li, W., Their, P. & Wehrhahn, C. ~2000!. Contextual influence on orientation discrimination of humans and responses of neurons in V1 of alert monkeys. Journal of Neurophysiology 83, 941–954. Marrocco, R.T., McClurkin, J.W. & Young, R.A. ~1982!. Modulation of lateral geniculate nucleus cell responsiveness by visual activation of the corticogeniculate pathway. Journal of Neuroscience 2, 256–263. McIlwain, J.T. ~1964!. Some evidence concerning the physiological basis of the periphery effect in the cat’s retina. Experimental Brain Research 1, 265–271.
728 Movshon, J.A., Thompson, I.D. & Tolhurst, D.J. ~1978!. Receptive field organization of complex cells in the cat’s striate cortex. Journal of Physiology 283, 79–99. Richmond, B.J., Hertz, J.A. & Gawne, T.J. ~1999!. The relation between V1 neuronal responses and eye movement-like stimulus presentations. Neurocomputing 26–27, 247–254. Ringach, D.L., Hawken, M.J. & Shapley, R. ~1997!. Dynamics of orientation tuning in macaque primary visual cortex. Nature 387, 281–284. Rodman, H.R., Gross, C.G. & Albright, T.D. ~1990!. Afferent basis of visual response properties in area MT of the Macaque. II. Effects of superior colliculus removal. Journal of Neuroscience 10, 1154–1164. Roelfsema, P.R., Khayat, P.S. & Spekreijse, H. ~2003!. Subtask sequencing in the primary visual cortex. Proceedings of the National Academy of Sciences of the United States of America 100, 5467–5472. Roelfsema, P.R., Lamme, V.A. & Spekreijse, H. ~1998!. Object-based attention in the primary visual cortex of the Macaque monkey. Nature 395, 376–381. Roelfsema, P.R. & Spekreijse, H. ~1999!. Correlates of a gradual spread of attention over a traced curve in macaque area V1. Society for Neuroscience Abstracts 25, 4. Rols, G., Tallon-Baudry, C., Girard, P., Bertrand, O. & Bullier, J. ~2001!. Cortical mapping of gamma oscillations in areas V1 and V4 of the Macaque monkey. Visual Neuroscience 18, 527–540. Rossi, A.F., Desimone, R. & Ungerleider, L.G. ~2001!. Contextual modulation in primary visual cortex of Macaques. Journal of Neuroscience 21, 1698–1709. Sceniak, M.P., Ringach, D.L., Hawken, M.J. & Shapley, R. ~1999!. Contrast’s effect on spatial summation by Macaque V1 neurons. Nature Neuroscience 2, 733–739. Schiller, P.H., Finlay, B.L. & Volman, S.F. ~1976!. Quantitative studies of single-cell properties in monkey striate cortex. V. Multivariate statistical analyses and models. Journal of Neurophysiology 39, 1362–1374. Schwartz, E.L. ~1977!. Spatial mapping in the primate sensory projection: Analytic structure and relevance to perception. Biological Cybernetics 25, 181–194. Smith, M.A., Bair, W. & Movshon, J.A. ~2002!. Signals in Macaque
D.M. Alexander and J.J. Wright striate cortical neurons that support the perception of glass patterns. Journal of Neuroscience 22, 8334–8345. Standage, G.P. & Benevento, L.A. ~1983!. The organization of connections between the pulvinar and visual area MT in the Macaque monkey. Brain Research 262, 288–294. Sugita, Y. ~1999!. Grouping of image fragments in primary visual cortex. Nature 401, 269–272. Super, H. & Roelfsema, P.R. ~2005!. Chronic multiunit recordings in behaving animals: Advantages and limitations. Progress in Brain Research 147, 263–282. Tootell, R.B., Switkes, E., Silverman, M.S. & Hamilton, S.L. ~1988!. Functional anatomy of the Macaque striate cortex. II. Retinotopic organization. Journal of Neuroscience 8, 1531–1568. Ungerleider, L.G., Desimone, R., Galkin, T.W. & Mishkin, M. ~1984!. Subcortical projections of area MT in the Macaque. Journal of Comparative Neurology 223, 368–386. Van Essen, D.C., Newsome, W.T. & Maunsell, J.H. ~1984!. The visual field representation in striate cortex of the Macaque monkey: Asymmetries, anisotropies, and individual variability. Vision Research 24, 429– 448. Wachtler, T., Sejnowski, T.J. & Albright, T.D. ~2003!. Representation of color stimuli in awake Macaque primary visual cortex. Neuron 37, 681– 691. Wright, J.J., Alexander, D.M. & Bourke, P.D. ~2006!. Contribution of lateral interactions in V1 to organization of response properties. Vision Research 46, 2703–2720. Yukie, M. & Iwai, E. ~1981!. Direct projection from the dorsal lateral geniculate nucleus to the prestriate cortex in Macaque monkeys. Journal of Comparative Neurology 201, 81–97. Zhou, H., Friedman, H.S. & von der Heydt, R. ~2000!. Coding of border ownership in monkey visual cortex. Journal of Neuroscience 20, 6594– 6611. Zipser, K., Lamme, V.A. & Schiller, P.H. ~1996!. Contextual modulation in primary visual cortex. Journal of Neuroscience 16, 7376–7389. Zufferey, P.D., Jin, F., Nakamura, H., Tettoni, L. & Innocenti, G.M. ~1999!. The role of pattern vision in the development of corticocortical connections. European Journal of Neuroscience 11, 2669–2688.
