Change In Response Gain Research Articles

Invited Session I: 30 years of normalization in the visual system: Testing and expanding the Reynolds & Heeger Normalization Model of Attention.

I present evidence supporting key predictions of the Reynolds-Heeger Normalization Model of Attention: 1) Both exogenous and endogenous covert spatial attention, can affect performance by contrast or response gain changes, depending on the stimulus size and the relative size of the attention field. 2) Feature-based attention improves performance, consistent with a change in response gain, when the featural extent of the attention field is small (low uncertainty) or large (high uncertainty) relative to the featural extent of the stimulus. 3) An expanded model of spatial resolution and attention that distinguishes between exogenous and endogenous attention, focusing on texture-based segmentation as a model system, which has revealed a dissociation between these types of attention across eccentricity. Our model emulates sensory encoding to segment figures from their background and predict performance. To explain attentional effects, exogenous and endogenous attention require separate operating regimes across spatial frequency. The model reproduces behavioral performance across several experiments and resolves three unexplained phenomena: 1) the parafoveal advantage in segmentation, 2) the uniform improvements across eccentricity by endogenous attention, and 3) the peripheral improvements and central impairments by exogenous attention. This model provides a generalizable framework for predicting effects of endogenous and exogenous attention on perception across the visual field.

How the size of exogenous attentional cues alters visual performance: From response gain to contrast gain.

The normalisation model of attention (NMoA) predicts that the attention gain pattern is mediated by changes in the size of the attentional field and stimuli. However, existing studies have not measured gain patterns when the relative sizes of stimuli are changed. To investigate the NMoA, the present study manipulated the attentional field size, namely, the exogenous cue size. Moreover, we assessed whether the relative rather than the absolute size of the attentional field matters, either by holding the target size constant and changing the cue size (Experiments 1-3) or by holding the cue size constant and changing the target size (Experiment 4), in a spatial cueing paradigm of psychophysical procedures. The results show that the gain modulations changed from response gain to contrast gain when the precue size changed from small to large relative to the target size (Experiments 1-3). Moreover, when the target size was once again made larger than the precue size, there was still a change in response gain (Experiment 4). These results suggest that the size of exogenous cues plays an important role in adjusting the attentional field and that relative changes rather than absolute changes to exogenous cue size determine gain modulation. These results are consistent with the prediction of the NMoA and provide novel insights into gain modulations of visual selective attention.

Contrast adaptation and interocular transfer in cortical cells: A re-analysis & a two-stage gain-control model of binocular combination

How do V1 cells respond to, adapt to, and combine signals from the two eyes? We tested a simple functional model that has monocular and binocular stages of divisive contrast gain control (CGC) that sit before, and after, binocular summation respectively. Interocular suppression (IOS) was another potential influence on contrast gain. Howarth, Vorobyov & Sengpiel (2009, Cerebral Cortex, 19, 1835–1843) studied contrast adaptation and interocular transfer in cat V1 cells. In our re-analysis we found that ocular dominance (OD) and contrast adaptation at a fixed test contrast were well described by a re-scaling of the unadapted orientation tuning curve – a simple change in response gain. We compared six variants of the basic model, and one model fitted the gain data notably better than the others did. When the dominant eye was tested, adaptation reduced cell response gain more when that eye was adapted than when the other eye was adapted. But when the non-dominant eye was tested, adapting either eye gave about the same reduction in overall gain, and there was an interaction between OD and adapting eye that was well described by the best-fitting model. Two key features of this model are that signals driving IOS arise 'early', before attenuation due to OD, while suppressive CGC signals are 'late' and so affected by OD. We show that late CGC confers a functional advantage: it yields partial compensation for OD, which should reduce ocular imbalance at the input to binocular summation, and improve the cell's sensitivity to variation in stereo disparity.

Effects of affective content and motivational context on neural gain functions during naturalistic scene perception.

Visual scene processing is modulated by semantic, motivational, and emotional factors, in addition to physical scene statistics. An open question is to what extent those factors affect low-level visual processing. One index of low-level visual processing is the contrast response function (CRF), representing the change in neural or psychophysical gain with increasing stimulus contrast. Here we aimed to (a) establish the use of an electrophysiological technique for assessing CRFs with complex emotional scenes and (b) examine the effects of motivational context and emotional content on CRFs elicited by naturalistic stimuli, including faces and complex scenes (humans, animals). Motivational context varied by expectancy of threat (a noxious noise) versus safety. CRFs were measured in 18 participants by means of sweep steady-state visual evoked potentials. Results showed a facilitation in visuocortical sensitivity (contrast gain) under threat, compared with safe conditions, across all stimulus categories. Facial stimuli prompted heightened neural response gain, compared with scenes. Within the scenes, response gain was smaller for scenes high in emotional arousal, compared with low-arousing scenes, consistent with interference effects of emotional content. These findings support the notion that motivational context alters the contrast sensitivity of cortical tissue, differing from changes in response gain (activation) when visual cues themselves carry motivational/affective relevance.

A Normalization Framework for Emotional Attention.

The normalization model of attention proposes that attention can affect performance by response- or contrast-gain changes, depending on the size of the stimulus and attention field. Here, we manipulated the attention field by emotional valence, negative faces versus positive faces, while holding stimulus size constant in a spatial cueing task. We observed changes in the cueing effect consonant with changes in response gain for negative faces and contrast gain for positive faces. Neuroimaging experiments confirmed that subjects’ attention fields were narrowed for negative faces and broadened for positive faces. Importantly, across subjects, the self-reported emotional strength of negative faces and positive faces correlated, respectively, both with response- and contrast-gain changes and with primary visual cortex (V1) narrowed and broadened attention fields. Effective connectivity analysis showed that the emotional valence-dependent attention field was closely associated with feedback from the dorsolateral prefrontal cortex (DLPFC) to V1. These findings indicate a crucial involvement of DLPFC in the normalization processes of emotional attention.

The Contributions of the on and off Gain Difference to the Contextual Effect in Macaque Monkey V1

In macaque monkey V1, the response amplitudes of ON and OFF sub-regions of simple receptive fields are about equal for cells in the input layer 4c, but the OFF responses tend to be greater than the ON responses for layer-2/3 cells. This OFF-over-ON bias is evident when receptive fields are mapped with sparse noise (a series of dark or bright squares shown in different space and time) but is weaker with Hartleys (gratings shown briefly at different orientations and spatial frequencies). One possible mechanism for the receptive-field mismatch in layer 2/3 of V1 is the difference between ON and OFF response gains caused by different degrees of sparseness between Hartleys and sparse noise. Here we manipulated the relative strength of ON and OFF sparse-noise responses until the receptive field similarity (RFS, quantified as the pixel-by-pixel correlation between two maps) between sparse-noise and Hartley maps reached the maximum. On average, the relative strength of ON/OFF responses in sparse-noise maps should be increased to approximately two times of their origins in order to match closely their Hartley compartments for layer-2/3 neurons (1.93±0.19; mean±S.E.M.). The sole change in response gain largely increased the RFS between Hartley and sparse-noise maps (on average the RFS increases from 0.35 to 0.58), and it could account for approximately 81% of the overall differences between the two maps. In summary, the difference in ON and OFF gains under different visual stimulations is one critical mechanism underlying the contextual effect in the superficial layer of V1.

Feature-based attention enhances performance by increasing response gain

Covert spatial attention can increase contrast sensitivity either by changes in contrast gain or by changes in response gain, depending on the size of the attention field and the size of the stimulus (Herrmann et al., 2010), as predicted by the normalization model of attention (Reynolds & Heeger, 2009). For feature-based attention, unlike spatial attention, the model predicts only changes in response gain, regardless of whether the featural extent of the attention field is small or large. To test this prediction, we measured the contrast dependence of feature-based attention. Observers performed an orientation-discrimination task on a spatial array of grating patches. The spatial locations of the gratings were varied randomly so that observers could not attend to specific locations. Feature-based attention was manipulated with a 75% valid and 25% invalid pre-cue, and the featural extent of the attention field was manipulated by introducing uncertainty about the upcoming grating orientation. Performance accuracy was better for valid than for invalid pre-cues, consistent with a change in response gain, when the featural extent of the attention field was small (low uncertainty) or when it was large (high uncertainty) relative to the featural extent of the stimulus. These results for feature-based attention clearly differ from results of analogous experiments with spatial attention, yet both support key predictions of the normalization model of attention.

Feature-based attention enhances performance by increasing response gain

Covert spatial attention can increase contrast sensitivity either by changes in contrast gain or by changes in response gain, depending on the size of the attention field and the size of the stimulus (Herrmann et al., 2010), as predicted by the normalization model of attention (Reynolds & Heeger, 2009). For feature-based attention, unlike spatial attention, the model predicts only changes in response gain, regardless of whether the featural extent of the attention field is small or large. To test this prediction, we measured the contrast dependence of feature-based attention. Observers performed an orientation-discrimination task on a spatial array of grating patches. The spatial locations of the gratings were varied randomly so that observers could not attend to specific locations. Feature-based attention was manipulated with a 75% valid and 25% invalid pre-cue, and the featural extent of the attention field was manipulated by introducing uncertainty about the upcoming grating orientation. Performance accuracy was better for valid than for invalid pre-cues, consistent with a change in response gain, when the featural extent of the attention field was small (low uncertainty) or when it was large (high uncertainty) relative to the featural extent of the stimulus. These results for feature-based attention clearly differ from results of analogous experiments with spatial attention, yet both support key predictions of the normalization model of attention.

When size matters: attention affects performance by contrast or response gain

Covert attention, the selective processing of visual information in the absence of eye movements, improves behavioral performance. Here, we show that attention, both exogenous (involuntary) and endogenous (voluntary), can affect performance by contrast or response gain changes, depending on the stimulus size and the relative size of the attention field. These two variables were manipulated in a cueing task while varying stimulus contrast. We observed a change in behavioral performance consonant with a change in contrast gain for small stimuli paired with spatial uncertainty, but a change in response gain for large stimuli presented at one location (no uncertainty) and surrounded by irrelevant flanking distracters. A complementary neuroimaging experiment revealed that observers’ attention field was wider with than without spatial uncertainty. Our results support key predictions of the normalization model of attention, and reconcile previous, seemingly contradictory, findings on the effects of visual attention.

Independent Controls of Attentional Influences in Primary and Secondary Somatosensory Cortex

The neuronal mechanisms underlying enhanced perception of tactile stimuli with directed attention were investigated using single-unit recordings from primary (S1, n = 53) and secondary (S2, n = 50) somatosensory cortex in macaque monkeys. Neuronal responses to textures scanned under the digit tips (spatial periods, SP, of 2, 3.7 or 4.7 mm) were recorded while attention was directed either to discriminating a change in texture or to the reward and also in a neutral no-task condition. Cell discharge was quantified in three periods of the trials: salient Delta texture (directed attention), postreward, and static (both cases, attention directed to the reward). S1 texture- and non-texture-sensitive cells, as well as S2 non-texture-sensitive cells, showed a modest enhancement of discharge during the salient Delta texture period (approximately 25%) but no change in response gain, consistent with an additive increase in neuronal responsiveness with directed attention. In contrast, S2 texture-related cells showed a larger enhancement with directed attention to salient inputs (82%) and increased response gain, suggesting that directed attention produces a multiplicative increase in S2 responsiveness. During the postreward period, and also in no-task testing, S1 texture-sensitive cells preserved their sensitivity to SP. In contrast, S2 texture-, but not non-texture-, sensitive cells showed a marked suppression of discharge and decreased gain after the discrimination response. Together, the results support the notion that S2 discharge reflects stimulus parameters in relation to ongoing behavioral demands. The results also support the existence of two independent attentional mechanisms in somatosensory cortex, one generalized (S1 and S2), and the other focused on S2 texture-related cells.

Effects of viewing distance on the responses of vestibular neurons to combined angular and linear vestibular stimulation.

Effects of viewing distance on the responses of vestibular neurons to combined angular and linear vestibular stimulation. The firing behavior of 59 horizontal canal-related secondary vestibular neurons was studied in alert squirrel monkeys during the combined angular and linear vestibuloocular reflex (CVOR). The CVOR was evoked by positioning the animal's head 20 cm in front of, or behind, the axis of rotation during whole body rotation (0.7, 1.9, and 4.0 Hz). The effect of viewing distance was studied by having the monkeys fixate small targets that were either near (10 cm) or far (1.3-1.7 m) from the eyes. Most units (50/59) were sensitive to eye movements and were monosynaptically activated after electrical stimulation of the vestibular nerve (51/56 tested). The responses of eye movement-related units were significantly affected by viewing distance. The viewing distance-related change in response gain of many eye-head-velocity and burst-position units was comparable with the change in eye movement gain. On the other hand, position-vestibular-pause units were approximately half as sensitive to changes in viewing distance as were eye movements. The sensitivity of units to the linear vestibuloocular reflex (LVOR) was estimated by subtraction of angular vestibuloocular reflex (AVOR)-related responses recorded with the head in the center of the axis of rotation from CVOR responses. During far target viewing, unit sensitivity to linear translation was small, but during near target viewing the firing rate of many units was strongly modulated. The LVOR responses and viewing distance-related LVOR responses of most units were nearly in phase with linear head velocity. The signals generated by secondary vestibular units during voluntary cancellation of the AVOR and CVOR were comparable. However, unit sensitivity to linear translation and angular rotation were not well correlated either during far or near target viewing. Unit LVOR responses were also not well correlated with their sensitivity to smooth pursuit eye movements or their sensitivity to viewing distance during the AVOR. On the other hand there was a significant correlation between static eye position sensitivity and sensitivity to viewing distance. We conclude that secondary horizontal canal-related vestibuloocular pathways are an important part of the premotor neural substrate that produces the LVOR. The otolith sensory signals that appear on these pathways have been spatially and temporally transformed to match the angular eye movement commands required to stabilize images at different distances. We suggest that this transformation may be performed by the circuits related to temporal integration of the LVOR.

608 Autonomic control of perceptual sensitivity and attention

Covert spatial attention can increase contrast sensitivity either by changes in contrast gain or by changes in response gain, depending on the size of the attention field and the size of the stimulus (Herrmann et al., 2010), as predicted by the normalization model of attention (Reynolds & Heeger, 2009). For feature-based attention, unlike spatial attention, the model predicts only changes in response gain, regardless of whether the featural extent of the attention field is small or large. To test this prediction, we measured the contrast dependence of feature-based attention. Observers performed an orientation-discrimination task on a spatial array of grating patches. The spatial locations of the gratings were varied randomly so that observers could not attend to specific locations. Feature-based attention was manipulated with a 75% valid and 25% invalid pre-cue, and the featural extent of the attention field was manipulated by introducing uncertainty about the upcoming grating orientation. Performance accuracy was better for valid than for invalid pre-cues, consistent with a change in response gain, when the featural extent of the attention field was small (low uncertainty) or when it was large (high uncertainty) relative to the featural extent of the stimulus. These results for feature-based attention clearly differ from results of analogous experiments with spatial attention, yet both support key predictions of the normalization model of attention.

Adaptive changes of vertical vestibulo-ocular reflex induced by tenectomy of vertical recti muscles and neuronal behavior related to vertical eye movement in the region of the interstitial nucleus of Cajal in alert cats.

Experiments were performed in alert cats to examine adaptive changes of the vertical vestibulo-ocular reflex (VOR) induced by tenectomy of the vertical recti muscles of one eye and behavior of neurons in the region of the interstitial nucleus of Cajal (INC) during this adaptation. Continuous pitch rotations (0.11 Hz) of a constant amplitude (+/- 10 degrees) with only the operated eye uncovered produced a gradual and significant increase in VOR gain that started within 15 min, reaching a plateau within 1 h. When only the normal eye was allowed to view again, VOR gain quickly decreased to the control values within 10 min. Such reversible gain change was not observed when visual input was not allowed to the operated eye, indicating that these changes are adaptive ones to reduce visual-vestibular conflict. A total of 18 neurons were recorded from the INC region during the VOR adaptation. These include 9 eye-movement-related neurons (8 vertical burst-tonic neurons and 1 vertical tonic neuron) and 9 other neurons that did not show close correlation with eye movement but were activated by pitch rotations (tentatively called pitch cells). During adaptive gain increase of the vertical VOR, all the vertical burst-tonic and tonic neurons showed a significant gain increase, whereas a significant change in response gain was observed in only 1 of the 9 pitch cells, indicating that vertical burst-tonic neurons in the INC region are involved in the adaptive gain increase in the vertical VOR.(ABSTRACT TRUNCATED AT 250 WORDS)