CoChat
Sign Up
Cited by 257Visual perception and processing mechanismsVisual Attention and Saliency DetectionNeural dynamics and brain function

The DOI matches its Crossref record; 51 of the 51 references it cites were found in a registry or an index.

A Normalization Model of Attentional Modulation of Single Unit Responses

Open access
Lee, Joonyeol, Maunsell, John H. R.·2009·PLoS ONE
DOI 10.1371/journal.pone.0004651PMC PMC2645695
Open this paper in CoChat — free accountOpen PDF

Abstract

Although many studies have shown that attention to a stimulus can enhance the responses of individual cortical sensory neurons, little is known about how attention accomplishes this change in response. Here, we propose that attention-based changes in neuronal responses depend on the same response normalization mechanism that adjusts sensory responses whenever multiple stimuli are present. We have implemented a model of attention that assumes that attention works only through this normalization mechanism, and show that it can replicate key effects of attention. The model successfully explains how attention changes the gain of responses to individual stimuli and also why modulation by attention is more robust and not a simple gain change when multiple stimuli are present inside a neuron's receptive field. Additionally, the model accounts well for physiological data that measure separately attentional modulation and sensory normalization of the responses of individual neurons in area MT in visual cortex. The proposal that attention works through a normalization mechanism sheds new light a broad range of observations on how attention alters the representation of sensory information in cerebral cortex.

Overview

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1371/journal.pone.0004651

Cites

51 of 51 references verified0 citation errors · 0 not checkable

Statistics

Figures

Figure 1 Attentional normalization model. The black dotted ellipse shows the receptive field of the neuron being considered, and the white dotted ellipses show receptive fields of neurons that provide input to this neuron. Each set of input neurons constitutes one normalization pool. Arrows overlapped with Gabors show the motion direction of the drifting Gabors. See text for details.
Figure 1 Attentional normalization model. The black dotted ellipse shows the receptive field of the neuron being considered, and the white dotted ellipses show receptive fields of neurons that provide input to this neuron. Each set of input neurons constitutes one normalization pool. Arrows overlapped with Gabors show the motion direction of the drifting Gabors. See text for details.CC BY · PubMed Central
Figure 2 Data and simulations for contrast dependent response summation. A: Average responses of MT neurons to pairs of stimuli that had independent contrasts. Reproduced with permission from Figure 9 of Hilary W. Heuer and Kenneth H. Britten, Contrast dependence of response normalization in area MT of the rhesus monkey. J. Neurophysiol. 88:3398–3408, 2002. Each cell was normalized by its maximum firing rate, and responses were averaged across 39 MT neurons. B: A contour surface showing a simulation of this test using the normalization model. The plot is generated by averaging predicted responses of 100 model neurons.
Figure 2 Data and simulations for contrast dependent response summation. A: Average responses of MT neurons to pairs of stimuli that had independent contrasts. Reproduced with permission from Figure 9 of Hilary W. Heuer and Kenneth H. Britten, Contrast dependence of response normalization in area MT of the rhesus monkey. J. Neurophysiol. 88:3398–3408, 2002. Each cell was normalized by its maximum firing rate, and responses were averaged across 39 MT neurons. B: A contour surface showing a simulation of this test using the normalization model. The plot is generated by averaging predicted responses of 100 model neurons.CC BY · PubMed Central
Figure 3 Simulations of effects of attention. A: Simulation of an attention effect on a direction tuning function. The solid black line is the direction tuning function with attention directed away from the stimulus, and the dotted black line is the tuning function with attention directed toward the stimulus. The dashed-dotted line is the spontaneous activity of the model neuron. B: Simulation of the effects of attention on responses to single and paired stimuli. The Gabor with the up-arrow is the preferred stimulus and the Gabor with the down-arrow is the non-preferred stimulus. The black dotted ellipse shows the receptive field, and the black and white circles mean attention to the preferred stimulus and non-preferred stimuli, respectively. Grey histograms are the responses when attention is directed to a stimulus outside the receptive field (not shown). Black and white histograms are the responses when the attention is directed to the preferred stimulus or the null stimulus in the receptive field, respectively. For simulations here, we used parameters of .
Figure 3 Simulations of effects of attention. A: Simulation of an attention effect on a direction tuning function. The solid black line is the direction tuning function with attention directed away from the stimulus, and the dotted black line is the tuning function with attention directed toward the stimulus. The dashed-dotted line is the spontaneous activity of the model neuron. B: Simulation of the effects of attention on responses to single and paired stimuli. The Gabor with the up-arrow is the preferred stimulus and the Gabor with the down-arrow is the non-preferred stimulus. The black dotted ellipse shows the receptive field, and the black and white circles mean attention to the preferred stimulus and non-preferred stimuli, respectively. Grey histograms are the responses when attention is directed to a stimulus outside the receptive field (not shown). Black and white histograms are the responses when the attention is directed to the preferred stimulus or the null stimulus in the receptive field, respectively. For simulations here, we used parameters of .CC BY · PubMed Central
Figure 4 Task designs for the stimulus interactions and attention experiments. The top panels show the stimulus configurations for the two experiments and the diagrams below show the structure of individual trials. The arrows overlapped the Gabors show the directions of their motion. The height of the lines in the trial plots represents the contrast or speed of the Gabor in each stimulus location. In the stimulus interactions experiment, the contrasts of the Gabor in one location are either twice or half the contrasts of the Gabor in the other location in the receptive field. Grey digits show that the contrasts of the target stimulus are randomly selected from the contrasts of two distracters in the receptive field. In the attention experiment, the contrasts of the two stimuli in the receptive field are matched with each other (see Methods).
Figure 4 Task designs for the stimulus interactions and attention experiments. The top panels show the stimulus configurations for the two experiments and the diagrams below show the structure of individual trials. The arrows overlapped the Gabors show the directions of their motion. The height of the lines in the trial plots represents the contrast or speed of the Gabor in each stimulus location. In the stimulus interactions experiment, the contrasts of the Gabor in one location are either twice or half the contrasts of the Gabor in the other location in the receptive field. Grey digits show that the contrasts of the target stimulus are randomly selected from the contrasts of two distracters in the receptive field. In the attention experiment, the contrasts of the two stimuli in the receptive field are matched with each other (see Methods).CC BY · PubMed Central
Figure 5 Contrast response functions for two example neurons from attention and stimulus interactions experiments. A: Responses of each neuron in the attention experiment. Filled circles are responses when attention is directed to the preferred stimulus and open circles are responses when attention is directed to the null stimulus. Solid and dotted lines are the best fitting functions of the attentional normalization model. B: Responses of each neuron in the stimulus interactions experiment. Filled circles are responses when the preferred stimulus has a higher contrast than the null stimulus (2×), and open circles are responses when the null stimulus has a higher contrast than the preferred stimulus (2×). Solid and dotted lines are the best fitting functions of the model. Vertical lines are standard errors and the best fitting functions were obtained by fitting data from the attention and stimulus interactions experiments simultaneously to the attentional normalization model.
Figure 5 Contrast response functions for two example neurons from attention and stimulus interactions experiments. A: Responses of each neuron in the attention experiment. Filled circles are responses when attention is directed to the preferred stimulus and open circles are responses when attention is directed to the null stimulus. Solid and dotted lines are the best fitting functions of the attentional normalization model. B: Responses of each neuron in the stimulus interactions experiment. Filled circles are responses when the preferred stimulus has a higher contrast than the null stimulus (2×), and open circles are responses when the null stimulus has a higher contrast than the preferred stimulus (2×). Solid and dotted lines are the best fitting functions of the model. Vertical lines are standard errors and the best fitting functions were obtained by fitting data from the attention and stimulus interactions experiments simultaneously to the attentional normalization model.CC BY · PubMed Central
Figure 6 Correlations between response modulations from attention and sensory normalization. Each value in the scatter plot shows the modulation indices of individual neurons for attention and sensory normalization. Vertical and horizontal lines are 95% confidence intervals from bootstrap analyses. The regression line is the best fitting function of a linear equation (see Methods). The rho is a Pearson's correlation coefficient.
Figure 6 Correlations between response modulations from attention and sensory normalization. Each value in the scatter plot shows the modulation indices of individual neurons for attention and sensory normalization. Vertical and horizontal lines are 95% confidence intervals from bootstrap analyses. The regression line is the best fitting function of a linear equation (see Methods). The rho is a Pearson's correlation coefficient.CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Cited by

Sorted by citations

See all 257

Related

See all

Metadata

Authors
Lee, J., & Maunsell, J. H. R.
Affiliations
  • Baylor College of Medicine (US)
  • Howard Hughes Medical Institute (US)
Corresponding author
Maunsell, John H. R.
ORCID
Venue
PLoS ONE
Publisher
Public Library of Science (PLoS)
Published
2009-02-27
Volume
4
Issue
2
Pages
e4651
ISSN
1932-6203 (ISSN-L)
DOI
10.1371/journal.pone.0004651
PMC
PMC2645695
OpenAlex
W2163588736
Cited by
257
References
51
Citations by year
2026: 3 · 2025: 10 · 2024: 6 · 2023: 27 · 2022: 10 · 2021: 14 · 2020: 11 · 2019: 18 · 2018: 13 · 2017: 13 · 2016: 18 · 2015: 18 · 2014: 20 · 2013: 18 · 2012: 17
FWCI
9.09
Type
journal article
Language
en
Open access
Licence
Locations
  • PLoS ONE (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • Digital Access to Scholarship at Harvard (DASH) (Harvard University) (submittedVersion, cc-by)
  • RePEc: Research Papers in Economics (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion)
Funders
  • Howard Hughes Medical Institute
  • National Institutes of Health
  • National Eye Institute (R01EY005911)
Topics
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Visual perception and processing mechanisms
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Visual Attention and Saliency Detection
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Neural dynamics and brain function
Keywords
attentional modulation, single-unit responses, gain modulation, multiple stimuli, receptive fields, area MT, visual cortex, computational model of attention, neuronal gain
MeSH terms
Attention, Cerebral Cortex, Humans, Models, Theoretical

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review