Center-surround inhibition in expectation and its underlying computational and artificial neural network models
- Key Laboratory of Brain, Cognition and Education Sciences, Ministry of Education, South China Normal University, China
- School of Psychology, Center for Studies of Psychological Application, and Guangdong Provincial Key Laboratory of Mental Health and Cognitive Science, South China Normal University, China
- Brain Health Institute, National Center for Mental Disorders, Shanghai Mental Health Center, Shanghai Jiao Tong University School of Medicine and School of Psychology, China
- Donders Institute for Brain, Cognition and Behaviour, Radboud University, Netherlands
Figures
Figure 1
Stimuli and protocols of the profile experiment.
(a) Left: the auditory cue, comprising either a low- or high-frequency tone, predicted the orientation of the first grating with equal validity in the baseline experiment. B20°: Baseline 20°; B70°: Baseline 70°. Right: in the main experiment, the low- or high-frequency tone predicted 20° or 70° (expected) orientation of the first grating with 75% validity. In the remaining 25% of trials, this orientation was chosen randomly and equally from four non-predicted orientations (30°, 40°, 50°, and 60°). There were two types of expected conditions: Expect 20° (E20°) and Expect 70° (E70°), and for both conditions, there were five possible distances in orientation space between the expected and test gratings, ranging from Δ0° through Δ40° with a step size of 10°. (b) In both baseline and main experiments, each trial began with an auditory cue, followed by an 1800 ms fixation interval. Then, two consecutive gratings were each presented for 150 ms and separated by a 300 ms blank interval. Participants were first asked to make a 2AFC judgment of either the orientation (clockwise or anticlockwise) or the spatial frequency (lower or higher) of the second grating relative to the first on orientation discrimination (OD, purple) and spatial frequency discrimination (SFD, blue) tasks, respectively. Then, participants were asked to make another 2AFC judgment on the tone of auditory cue, either low or high. CW: clockwise; CCW: counterclockwise; HF: higher frequency; LF: lower frequency; HT: high tone; LT: low tone. (c) Left: expectation operates by the sharpening model with suppressing unexpected information, under this configuration, the profile of expectation could display as a center-surround inhibition, with an inhibitory zone surrounding the focus of expectation. Right: expectation operates by the cancellation model with highlighting unexpected information. Under this configuration, the profile of expectation could display as a monotonic gradient, without the inhibitory zone.
Figure 2 with 1 supplement
Results of the profile experiment.
The discrimination thresholds of OD (top) and SFD (bottom) tasks during baseline (a) and main (b) experiments. In the baseline experiment, discrimination thresholds did not differ across orientations in either OD or SFD tasks, as confirmed by non-significant one-way repeated-measures ANOVAs (all p>0.18). B20°: Baseline 20°; B70°: Baseline 70°; E20°: Expect 20°; E70°: Expect 70°. (c) The averaged discrimination sensitivity (DS) of each distance on OD (top) and SFD (bottom) tasks, and the best fitting Gaussian and Mexican-hat functions to these DSs across distances. In both tasks, DS varied significantly across distances (OD: F(4,92) = 3.739, p=0.010, 𝜂p2=0.140; SFD: F(4,92) = 2.822, p=0.042, 𝜂p2=0.109), and Post hoc paired t tests revealed that, for both tasks, the DSs of Δ20° were significantly lower than those of both Δ0° and Δ40°, consistent with the classical center-surround inhibition profile. G, Gaussian model; M, Mexican-hat model. (d) R2 of the best fitting Gaussian and Mexican-hat functions for individual participants in OD (top) and SFD (bottom) tasks. For both tasks, most dots located in the upper-left zone demonstrated that the Mexican-hat model was favored over the Gaussian model. Open symbols represent the data from each participant and filled colored dots represented the mean across participants. Error bars indicate 1 SEM calculated across participants (N = 24).
Figure 2—figure supplement 1
Accuracies of auditory tone reports in the profile experiment.
Mean accuracies of auditory tone reports in OD (a) and SFD (b) tasks during the baseline (left) and main (right) experiments, plotted across different distances in orientation space (Δ0° - Δ40°). In both the baseline and main experiments, a one-way repeated-measures ANOVA with the distance (Δ0°-Δ40°) as within-participants factor showed that the main effect of distance was not significant for either OD (baseline: F(4,92) = 1.281, p=0.283, ηp2 = 0.053; main: F(4,92) = 2.412, p=0.082, ηp2 = 0.095) or SFD (baseline: F(4,92) = 0.631, p=0.594, ηp2 = 0.027; main: F(4,92) = 0.963, p=0.414, ηp2 = 0.040) tasks. Error bars indicate 1 SEM calculated across participants (N = 24).
Figure 3 with 2 supplements
Results of computational modeling.
(a) Illustration of the Tuning sharpening model (left) and the Tuning shift model (right). The Tuning sharpening model postulates that expectation sharpens the tuning of individual neurons (thick curves) towards the expected orientation, which results in a center-surround population response profile (black curve) centered at the expected orientation. The Tuning shift model postulates that expectation attracts the tuning of individual neurons (thick curves) from unexpected orientation towards the expected orientation, which also results in a center-surround population response profile. (b) The fitted discrimination thresholds on OD (left) and SFD (right) tasks in the baseline (top) and main (bottom) experiments. (c) The averaged DSs using Tuning sharpening model on OD (left) and SFD (right) tasks. (d) R2 of the best fitting Gaussian and Mexican-hat functions for individual participants based on the fitted DSs using Tuning sharpening model on OD (left) and SFD (right) tasks. Open symbols represent the data from each participant and filled colored dots represented the mean across participants. (e–g) The results from the Tuning shift model, see caption for (b–d) for a description of each type of graph. The amplitude A (vertical stripes) and width σ (diagonal stripes) differences between the baseline and main experiments using Tuning sharpening model in Δ0° (h) and Δ10°-Δ40° (i) conditions, on OD (left) and SFD (right) tasks. The location x0 differences between the baseline and main experiments using Tuning shift model in Δ0° (j) and Δ10°-Δ40° (k) conditions, on OD (left) and SFD (right) tasks. Statistical comparisons were performed using t-tests against zero. Open symbols represent the data from each participant, and error bars indicate 1 SEM calculated across participants (N = 24). B20°: Baseline 20°; B70°: Baseline 70°; E20°: Expect 20°; E70°: Expect 70° (*p<0.05; **p<0.005; ***p<0.001).
Figure 3—figure supplement 1
The calculated parameters of the population responses on OD and SFD tasks.
(a) Amplitude A (vertical stripes) and Width σ (diagonal stripes) of the Tuning sharpening model for the baseline (left) and main (right) experiments in the Δ0° (top) and Δ10°-Δ40° (bottom) conditions. (b) Location x0 of the Tuning shift model for the baseline (left) and main (right) experiments in the Δ0° (top) and Δ10°-Δ40° (bottom) conditions. Note that the data were presented as the bias between x0 and their hypothesized channel location, i.e., 20°, 30°, 40°, 50°, 60°, and 70°. Error bars indicate 1 SEM calculated across participants (N = 24). (c and d) Computational modeling for the SFD task, see caption for (a and b) for a description of each type of graph.
Figure 3—figure supplement 2
Results of combined computational modeling.
(a) Illustration of the Combined model. The Combined model incorporates sharpening of the expected orientation channel (center channel) together with shifting of the unexpected orientation channels (surround channels) from unexpected toward the expected orientation, resulting in a center–surround population response profile (black curve). (b) The fitted discrimination thresholds on OD (left) and SFD (right) tasks in the baseline (top) and main (bottom) experiments. (c) The averaged DSs using Combined model on OD (left) and SFD (right) tasks. (d) R2 of the best fitting Gaussian and Mexican-hat functions for individual participants based on the fitted DSs on OD (left) and SFD (right) tasks. Open symbols represent the data from each participant and filled colored dots represented the mean across participants. The amplitude A (vertical stripes) and width σ (diagonal stripes) differences between the baseline and main experiments in Δ0° (e) and Δ10°-Δ40° (f) conditions, on OD (left) and SFD (right) tasks. The location x0 differences between the baseline and main experiments Δ10°-Δ40° (g) conditions, on OD (left) and SFD (right) tasks. For the expected orientation (Δ0°), results showed that the amplitude change was significantly higher than zero on both OD (t(23) = 2.582, p=0.017, Cohen’s d=0.527) and SFD (t(23) = 2.078, p=0.049, Cohen’s d=0.424) tasks (e, vertical stripes); the width change was significantly lower than zero on both OD (t(23) = –2.438, p=0.023, Cohen’s d=0.498) and SFD (t(23) = –2.578, p=0.017, Cohen’s d=0.526) tasks (e, diagonal stripes). For unexpected orientations (Δ10°-Δ40°), however, the amplitude and width changes were not significant with zero on either OD (amplitude change: t(23) = 0.443, p=0.662, Cohen’s d=0.091; width change: t(23) = –1.819, p=0.082, Cohen’s d=0.371) or SFD (amplitude change: t(23) = 1.130, p=0.270, Cohen’s d=0.231; width change: t(23) = –1.710, p=0.101, Cohen’s d=0.349) tasks (f). In the meantime, the location shift was significantly different than zero for unexpected orientations (Δ10°-Δ40°), OD task: t(23) = 3.611, p=0.001, Cohen’s d=0.737; SFD task: t(23) = 2.418, p=0.024, Cohen’s d=0.493 (g). Open symbols represent the data from each participant and error bars indicate 1 SEM calculated across participants (N = 24). B20°: Baseline 20°; B70°: Baseline 70°; E20°: Expect 20°; E70°: Expect 70° (*p<0.05; **p<0.005; ***p<0.001).
Figure 4
RMSDs of Tuning sharpening and Tuning shift models.
RMSDs of Tuning sharpening and Tuning shift models during the baseline (top) and main (bottom) experiments, on OD (a) and SFD (b) tasks. B20°: Baseline 20°; B70°: Baseline 70°; E20°: Expect 20°; E70°: Expect 70°. Open symbols represent the data from each participant and filled colored dots represented the mean across participants. Error bars indicate 1 SEM calculated across participants (N = 24).
Figure 5 with 1 supplement
Protocol and error distributions of the orientation adjustment experiment.
(a) The protocol of orientation adjustment experiment was similar to that of the profile experiment, except for two aspects. First, there were four possible (20°, 40°, 50°, and 70°) orientations for the first grating: 20°/70° (Δ0° deviated from the expected orientation) and 40°/50° (Δ20°/Δ30° deviated from the expected orientation). Second, in both baseline and main experiments, the second grating was set as a random orientation within the range of 0° to 90°, and participants were required to rotate the orientation of the second grating to match the first. HT: high tone; LT: low tone. (b) Three-component mixture model to the adjusted errors from baseline (left) and main (middle) experiments. In the current study, the shift towards 20° was (arbitrarily) considered to be the negative value (‘-’), whereas the shift towards 70° was thus the positive value (‘+’). The mean shift was calculated as: mean shift = (shift towards 70° - shift towards 20°)/2. The shaded error bars indicate 1 SEM calculated across participants (N = 20). B20°: Baseline 20°; B70°: Baseline 70°; E20°: Expect 20°; E70°: Expect 70°.
Figure 5—figure supplement 1
Accuracies of auditory tone report in orientation adjustment experiments.
Mean accuracies of auditory tone report during the baseline (left) and main (right) experiment. In both the baseline and main experiments, a repeated-measures ANOVA with the expected condition (B/E20° vs. B/E70°) and orientation distance (Δ0° vs. Δ20/30°) as within-participants factors showed that the interaction between these two factors was not significant for either the baseline (F(1,19) = 3.330, p=0.084, ηp2 = 0.149) or main (F(1,19) = 1.509, p=0.234, ηp2 = 0.074) experiments. Error bars indicate 1 SEM calculated across participants (N = 20).
Figure 6 with 1 supplement
Results of the orientation adjustment experiment.
(a) The adjusted orientation difference between the baseline and main experiments in both expected (20°/70°, i.e. Δ0°, middle) and unexpected (40°/50°, i.e. Δ20°/Δ30°, right) conditions. B20°: Baseline 20°; B70°: Baseline 70°; E20°: Expect 20°; E70°: Expect 70°. (b–d) The parameter estimates difference between the baseline and main experiments in both expected (Δ0°, middle) and unexpected (Δ20°/Δ30°, right) orientations. The parameter estimates were obtained by fitting a three-component mixture model to adjusted errors in different conditions. (b) mu reflects the response distribution shift away from the presented grating orientation. (c) s.d. reflects precision of responses (with higher values indicating worse precision). (d) g estimates the probability that the participant produced a random response (i.e. the guess). Statistical comparisons were performed using t-tests against zero. Open symbols represent the data from each participant and error bars indicate 1 SEM calculated across participants (N = 20; *p<0.05; **p<0.005; ***p<0.001).
Figure 6—figure supplement 1
Adjusted errors of orientation adjustment experiments and their parameter estimates with the three-component mixture model.
The adjusted errors in both baseline (left) and main (right) experiments for expected (a) 20°/70°, i.e. Δ0° and unexpected (b) 40°/50°, i.e. Δ20°/Δ30° conditions. The estimated parameter mu in both the baseline (left) and main (right) experiments for expected (c) 20°/70°, i.e. Δ0° and unexpected (d) 40°/50°, i.e. Δ20°/Δ30° conditions. The estimated parameter s.d. in both the baseline (left) and main (right) experiments for expected (e) 20°/70°, i.e. Δ0° and unexpected (f) 40°/50°, i.e. Δ20°/Δ30° conditions. The estimated parameter g in both the baseline (left) and main (right) experiments for expected (g) 20°/70°, i.e. Δ0° and unexpected (h) 40°/50°, i.e. Δ20°/Δ30° conditions. Error bars indicate 1 SEM calculated across participants (N = 20).
Figure 7 with 1 supplement
Protocol and results of the orientation discrimination experiment.
(a) The protocol of orientation discrimination experiment was similar to that of the orientation adjustment experiment, except for two aspects. First, there were three possible (20°, 45°, and 70°) orientations for the first grating: 20°/70° (Δ0° deviated from the expected orientation) and 45° (Δ25° deviated from the expected orientation). Second, in both baseline and main experiments, the second grating was 1°, 3°, 5°, 7°, and 9° deviated from the first grating, either clockwise (CW) or counterclockwise (CCW). Participants were asked to make a 2AFC judgment of the orientation of the second grating relative to the first, either clockwise or anticlockwise. HT: high tone; LT: low tone. Psychometric functions showing orientation judgements in each condition for Δ0° (b) and Δ25° (c). Data points averaged across participants were fit using a cumulative normal function. The abscissa refers to 10 orientation differences between the first and second gratings. The ordinate refers to the percentage of trials in which participants indicated the orientation of the second grating that was anticlockwise or clockwise to the first for expected 20° (left) and 70° (right) conditions, respectively. The slope (an index for the Tuning sharpening model), (d) and shift (an index for the Tuning shift model), (e) differences between the baseline and main experiments for expected 20° and 70° conditions. Statistical comparisons were performed using t-tests against zero. Negative: shift to the left; Positive: shift to the right. Open symbols represent the data from each participant and error bars indicate 1 SEM calculated across participants (N=18). B20°: Baseline 20°; B70°: Baseline 70°; E20°: Expect 20°; E70°: Expect 70° (*p<0.05; **p<0.005).
Figure 7—figure supplement 1
Accuracies of auditory tone reports in orientation discrimination experiments.
Mean accuracies of auditory tone report during the baseline (left) and main (right) experiment. In both the baseline and main experiments, a repeated-measures ANOVA with the expected condition (B/E20° vs. B/E70°) and orientation distance (Δ0° vs. Δ25°) as within-participants factors showed that the interaction between these two factors was not significant for either the baseline (F(1,17) = 0.481, p=0.497, ηp2 = 0.028) or main (F(1,17) = 0.288, p=0.599, ηp2 = 0.017) experiments. Error bars indicate 1 SEM calculated across participants (N = 18).
Figure 8
Results of artificial neural networks.
(a) Model structure and stimulus examples for deep predictive coding neural network (DPCNN) and standard feedforward CNN, on both OD (purple) and SFD (blue) tasks. DPCNN consisted of six feedforward encoding layers (e1–e6), five generative feedback decoding layers (d1–d5), and three fully connected (fc) layers. The reconstruction error (E1–E5) is computed and used for the proposed predictive coding updates, denoted by P.C. loops. The CNN is the same as DPCNN but removes feedback predictive coding iterations. The accuracy of each distance during the pre- (b) and post- (c) training for DPCNN (left) and CNN (right), on the OD task. (d) The training effect (i.e. the ACC difference between pre- and post-training) of each distance in DPCNN (left) and CNN (right), and the best fitting Mexican-hat and Gaussian functions to these training effects across distances, on the OD task. M, Mexican-hat model; G, Gaussian model. (e) R2 of the best fitting Mexican-hat and Gaussian functions from individual data in DPCNN (left) and CNN (right) on the OD task. Open symbols represent individual data and filled colored dots represent the mean across data. Error bars indicate 1 SEM calculated across data (N = 12). (f–i) The results from the SFD task, see caption for (b–e) for a description of each type of graph.
Additional files
-
Supplementary file 1
Supplementary materials containing all additional data, analyses, and supporting results for the study.
- https://cdn.elifesciences.org/articles/107301/elife-107301-supp1-v1.docx
-
MDAR checklist
- https://cdn.elifesciences.org/articles/107301/elife-107301-mdarchecklist1-v1.docx
Download links
A two-part list of links to download the article, or parts of the article, in various formats.
Downloads (link to download the article as PDF)
Open citations (links to open the citations from this article in various online reference manager services)
Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)
Center-surround inhibition in expectation and its underlying computational and artificial neural network models
eLife 14:RP107301.
https://doi.org/10.7554/eLife.107301.3
