Sage Journals: Discover world-class research

Abstract

Maniglia and colleagues reported a significant reduction in visual crowding following perceptual learning training on contrast detection using a lateral masking configuration with collinear flankers. They interpreted this reduction within a framework of shared cortical mechanisms between collinear inhibition, elicited by lateral masking with closely spaced flankers, and crowding. We reanalyzed their data to directly test this hypothesis by examining correlations between learning gains at short target-to-flankers separations (reduced contrast detection thresholds) and crowding reduction. Surprisingly, individual analyses revealed an inverse correlation: participants with greater reduction in collinear inhibition showed smaller reductions in crowding. We suggest that these participants exhibited separation-specific learning, which previous studies indicate may hinder effective transfer. Thus, while collinear inhibition and crowding may share mechanisms, distributed improvement across separations might be necessary to observe transfer of learning to crowding.

Keywords

collinear facilitation visual crowding perceptual learning contrast sensitivity

Introduction

Collinear modulation and visual crowding are among the most studied phenomena in visual perception. In collinear modulation, the detection of a central target, usually a Gabor patch, is affected by the presence of iso-oriented flanking elements (Polat & Sagi, 1993): at short separation between the target and flankers, (<2λ in fovea, < 6/8λ in periphery), contrast sensitivity decreases (collinear inhibition), while larger distances lead to an increase in sensitivity (collinear facilitation). In crowding, the identification of a central target, often a letter, is hindered by the presence of flanking elements placed within a so-called “critical space” (Levi, 2008).

While differing in methodology (collinear modulation is usually measured with oriented gratings, whereas crowding typically uses letters) and perceptual characteristics (collinear inhibition reduces detection, whereas crowding reduces discrimination), these two phenomena share several characteristics.

First, the strength of these effect seems to depend on the flanker separation: short distances are more deleterious for both crowding and lateral interaction, while larger separations reduce both.

Second, the inhibitory range of the flankers increases with eccentricity (Bouma, 1970; Lev & Polat, 2011; Maniglia et al., 2011, 2015a), possibly following cortical magnification. Levi and colleagues (Levi et al., 1984) suggested that crowding takes place when both target and flankers are placed within the same cortical “perceptual hypercolumn.” Similarly, Lev and Polat proposes that collinear modulation takes place between or within “perceptual fields” (Lev & Polat, 2011).

Third, the neural substrates of both phenomena, while still debated, are usually considered to be found in early visual cortex (Bringuier et al., 1999; Cass & Spehar, 2005; Gilbert & Wiesel, 1985, 1989; Grinvald et al., 1994; Levi & Carney, 2009; Pelli, 2008; Polat et al., 1998), hypothesis supported by animal electrophysiology and human neuroimaging studies(Henry & Kohn, 2022; Kapadia et al., 1995, 2000; Millin et al., 2014; Polat et al., 1998), with possibly feedback modulation from extra-striate areas (Chicherov et al., 2014; Levi, 2008; Maniglia et al., 2019; Whitney & Levi, 2011) As such, several authors have suggested that these phenomena might share, at least partially, the same neural substrates (Bonneh et al., 2007; Doron et al., 2015; Petrov et al., 2006; Polat & Sagi, 1993, 2007; Polat et al., 1997).

Fourth, learning studies show that training on collinear modulation transfers to reduction of crowding (Maniglia et al., 2011; Maniglia et al., 2020).

In particular, Maniglia and colleagues trained participants on peripheral collinear modulation and observed a reduction of critical space of crowding after training (Maniglia et al., 2011). Authors interpreted these results as supportive of the hypothesis of shared neural substrates, with reduced cortical inhibition after training contributing to the reduction of visual crowding.

However, the authors did not directly compare training gain and transfer gain to address the hypothesis.

Here, we used linear regression methods to test whether the learning gain observed in collinear inhibition is related to the transfer gain to crowding reduction. Surprisingly, results showed a significant, but inverse correlation between reduction of collinear inhibition (measured at the shortest separation of 2λ) and reduction of visual crowding, suggesting a less straightforward interaction between the two phenomena. We interpret this inverse correlation as possibly resulting from a lack of transfer when training is too specific to one of the sub-conditions. This aligns with previous findings in the lateral masking literature (Polat & Sagi, 1994) and with studies and models addressing learning and generalization (Hung & Seitz, 2014; Maniglia & Seitz, 2018; Xiao et al., 2008).

When learning is limited to a single separation (or, more precisely in the context of this paper, when one separation shows proportionally more learning effect than the others), it might disrupt the local chain of lateral interactions, thus compromising collinear facilitation for the untrained separations (or in this case, separations that show less learning effect) and preventing transfer (Polat & Sagi, 1994).

Method

Participants

Eight healthy vision participants took part in the original study. Details can be found in Maniglia et al. (2011).

Apparatus and Procedure

Participants took part in a peripheral contrast detection training using a lateral masking configuration. Before and after training, participants underwent a visual crowding task to assess transfer of learning. Further details can be found in Maniglia et al. (2011).

Data Analysis

To analyze the relationship between collinear modulation and crowding, we used Pearson's correlations and linear multiple regression gain model in R. The dependent variables of interest for the Pearson's correlations were crowding thresholds (separation between target and flankers leading to 79% correct identification (Whitney & Levi, 2011)) at pre- and post-test, separately.

The dependent variable for the multiple regression was crowding thresholds at post-test. To estimate crowding improvements, pre-test crowding thresholds were used as a covariate. This approach is considered more effective in capturing longitudinal changes as it controls for baseline performance and does not assume that the regression coefficient of pre versus post performance is 1.0 (Cohen, 2013). To investigate the relationship of crowding change and training, training threshold gains at the various flanker separations were added as covariates. Training thresholds were calculated as log[iso/ortho] flanker orientation ratios (Maniglia et al., 2011; Maniglia et al., 2015b; Shani & Sagi, 2005). Training gains were then calculated by subtracting post- from the pre-training thresholds.

Results

Correlation Between Lateral Inhibition and Critical Space of Crowding

First, we looked at the within-participant correlations between contrast threshold ratios (iso/ortho), measured at different separations, and critical space of crowding. Results showed no significant correlation between collinear inhibition (2λ, 3λ, and 4λ) or facilitation (8λ) and crowding at baseline or after training (see Figure 1).

Figure 1.

Within-participant correlations between crowding and collinear modulation. Larger values on the y-axis indicate worse crowding performance, whereas positive values on the x-axis indicate collinear inhibition (collinear contrast thresholds higher than orthogonal thresholds).

Correlation Between Training Gain and Transfer to Crowding

Next, we looked at the correlation between training gain in both tasks (Figure 2). We found no significant correlations, except for the marginally significant test at 2λ.

Figure 2.

Within-participant correlations between changes in crowding and collinear modulation training. The y-axis represents the change in visual crowding (post-training minus the baseline performance), measured in arcminutes of the visual angle. The x-axis represents the change in contrast detection performance across different flanker separations (post-training the minus baseline performance).

We then compared linear regression models predicting crowding performance from training gain at each target-to-flanker separation (Table 1). This analysis found that the best model based on parsimony and model fit (Kenny, 1979) was Model 1, in which training gains at 2λ predicted transfer to crowding. Specifically, results show a negative coefficient of 2λ, indicating that an improvement at 2λ due to training (indicated by a negative threshold difference) leads to higher (worse) post-test crowding thresholds (indicating a smaller transfer gain).

Table 1.

Regressed crowding change model comparisons. For each model, we present the relevant statistical information.

Parameter	Coefficient	Std. error	t-Value	p
Model 0 (Crowding Post ∼ Crowding Pre)
Intercept	−8.21	35.47	−0.23	.82
Crowding Pre	0.9	0.24	3.74	.009
F(1,6)= 13.98, p = .01		Adjusted R² = .65
AICc = 79.22
BIC = 73.46
Model 1 (Crowding Post ∼ Crowding Pre + 2λ)
Intercept	−65.04	33.57	−1.94	.11
Crowding Pre	1.24	0.22	5.72	.002
2λ	−32.04	12.37	−2.58	.048
F(2,5)= 16.99, p = .006		Adjusted R² = .82
AICc = 81.75
BIC = 68.74
Model 2 (Crowding Post ∼ Crowding Pre + 3λ)
Intercept	−27.58	43.17	−0.64	.55
Crowding Pre	0.98	0.26	3.7	.014
3λ	−26.03	31.18	−0.83	.44
F(2,5) = 6.98, p = .036		Adjusted R² = .63
AICc = 87.51
BIC = 74.49
Model 3 (Crowding Post ∼ Crowding Pre + 4λ)
Intercept	−9.67	38.69	−0.25	.81
Crowding Pre	0.92	0.27	3.43	.018
4λ	8.96	27.02	0.33	.75
F(2,5)= 6.01, p = .046		Adjusted R² = .58
AICc = 88.38
BIC = 75.36
Model 4 (Crowding Post ∼ Crowding Pre + 8λ)
Intercept	−13.47	40.83	−0.33	.75
Crowding Pre	0.94	0.28	3.35	.02
8λ	23.31	62.33	0.37	.72
F(2,5)= 6.06, p = .046		Adjusted R² = .59
AICc =88.33
BIC = 75.32

Looking at individual lambda gain as a predictor of learning transfer to crowding, the model that includes only 2λ as a predictor has the lowest AICc (small-sample corrected Akaike Information Criterion) aside from the model using only pre-test crowding scores, and the lowest BIC (Bayesian Information Criterion). This suggests that the 2λ-only model provides the best balance between fit and parsimony. In contrast, the more complex model including all lambda separations as predictors was excluded due to high variance inflation factors (>7) and overparameterization relative to the sample size for a reliable AICc calculation, as the number of predictors (6) approached the number of participants (8).

Discussion

The relationship between visual crowding and lateral masking has been discussed in vision science for decades. On one hand, several researchers propose that the two phenomena are independent (Chakravarthi & Cavanagh, 2009; Pelli, 2008; Pelli et al., 2004), on the other hand, others believe the two to be related (Chung et al., 2001; Doron et al., 2015; Petrov et al., 2006; Polat & Sagi, 1993). The latter suggests that crowding and lateral masking share some basic features, hinting at common neural substrates. These are usually identified in the long-range horizontal connections of V1, linking units with similar orientation tuning (Bringuier et al., 1999; Cass & Spehar, 2005; Gilbert & Wiesel, 1985, 1989; Grinvald et al., 1994; Polat et al., 1998). This hypothesis is supported by electrophysiology (Polat et al., 1998) and neuroimaging (Millin et al., 2014) evidence for lateral masking and crowding, respectively (but see (Chicherov et al., 2014; Henry & Kohn, 2022; Motter, 2006) for a higher-level hypothesis on the neural substrates of crowding). Additionally, both phenomena appear modulated by feedback from extrastriate areas (Hupé et al., 1998; Levi, 2008; Maniglia et al., 2019)

However, these two phenomena differ in several aspects. In crowding, the target does not become less “visible,” as in collinear inhibition (Polat & Sagi, 2007), but rather unidentifiable (Levi, 2008). Electrophysiological evidence further suggests that crowding may occur in extrastriate regions, such as V4 (Chicherov et al., 2014).

Here, we reanalyzed data from Maniglia et al. (2011), a study showing that training contrast detection with lateral masking led to reduction of crowding. Authors interpreted this as supporting evidence of (at least partially) shared neural substrates. However, the authors did not directly test the relationship between learning in the collinear task and transfer to crowding. Here, we offer further evidence to support this relationship, although in a less intuitive way: we found a significant, but inverse correlation between post-training reduction of collinear inhibition and critical space of crowding, so that larger reductions of collinear inhibition (at short separations) led to smaller reductions of crowding.

This relationship is significant for the shortest flanker separations (2λ and 3λ), where collinear inhibition is systematically observed (see Figure 1), but not for larger separations. This is consistent with the hypothesis that crowding is related to collinear inhibition, but not to collinear facilitation, which previous authors suggest depending upon different mechanisms. Additionally, this is consistent with evidence suggesting that training on lateral masking leads mostly to reduction of inhibition at short flankers’ separations rather than increase in facilitation at larger separations. Lev and Polat (2011) propose that the relationship between crowding and lateral interactions is characterized by the perceptual field (PF), the perceptual correspondent of the physiological receptive field. In this framework, collinear inhibition emerges as a within-PF effect, with facilitation being a between PFs effect. Similarly, crowding might depend on flankers located within the same perceptual region of integration than the target, so that the configuration is grouped into a whole (Manassi et al., 2013). Indeed, both crowding and PF size increase with eccentricity (Bouma, 1970; Lev & Polat, 2011) while being negligible or absent in the fovea (Levi et al., 2002).

However, it is important to note that while the lateral masking literature [alongside other evidence, for example, Henry & Kohn (2020); Millin et al. (2014); Nandy and Tjan (2012); Shin et al. (2017)] supports a “V1-centric” interpretation of crowding, alternative explanations involving higher-level visual areas are equally prominent. In particular, evidence from neurophysiology (Motter, 2006, 2018), neuroimaging (He et al., 2019) and psychophysics suggests that multiple brain areas, including extrastriate regions (Anderson et al., 2012; Bi et al., 2009; Tyler & Likova, 2007) and beyond (e.g., Louie et al. , 2007), contribute to crowding, along with higher-level mechanisms such as visual attention (Fang & He, 2008). Thus, the current results only partially address a more complex phenomenon. As Whitney and Levi (2011) noted, “any single visual area could explain this range of effects is therefore tenuous, at best.”

Similarly, while we, along with other authors, frame our results in the context of the relationship between lateral interactions and crowding, an alternative explanation is that improvements in contrast thresholds per se had an alleviating effect on crowding. Indeed, Rodriguez and Granger (Rodriguez & Granger, 2021) posit that differences in contrast can explain key characteristics of crowding, and several studies have shown a reduction in the crowding effect for low-contrast stimuli (Moshkovitz et al., 2022; Pascal & Abadi, 1995; Simmers et al., 1999).

What may seem counterintuitive at first is the inverse relationship between collinear inhibition and crowding, so that those who improved the most after training at the shortest separations (the more inhibitory condition) showed the smallest amount of transfer to crowding.

A possible explanation is that too-specific learning limits transfer: Indeed, when looking at training data, we observe that most of the learning took place at the shortest separations (Figure 3), with 2λ and 3λ showing the largest improvement (pre vs. post t-tests showing significant learning at 3λ (t₇ = 3.33, p = .013, and, removing one outlier in the opposite direction of the overall trend, at 2λ as well, t₆ = 2.93, p = .026). Those are the separations that learned the most and possibly gatekept transfer. This result is consistent with previous evidence suggesting that lateral masking training results mostly in reduction of inhibition (i.e., improvements at short separations) rather than increase in facilitation (Maniglia et al., 2011, 2018; Polat & Sagi, 1994). Still, the progressively reversing correlation between gains as separations increase (Figure 2) hints at a more complex relationship between collinear inhibition, facilitation, and crowding.

Figure 3.

Pre- versus post-contrast thresholds (iso orientation) across lambdas.

Doron et al. (2015) characterized the relationship between crowding and lateral masking in children, reporting a progressive reduction of crowding as collinear facilitation emerges. Here, we show that training on collinear inhibition affects visual crowding. However, several methodological differences between the two studies prevents us from comparing the results more in depth, and possibly explain why we did not observe a correlation between crowding and collinear inhibition before or after training, despite their learning gain being related. This is also consistent with evidence suggesting that performance across different visual tests does not necessarily correlate within the same individuals (Cappe et al., 2014).

To conclude, by reanalyzing data from Maniglia et al. (2011), we confirmed the original hypothesis of the authors, that is, training on collinear inhibition leads to improvements in visual crowding, albeit in the opposite direction: larger training gain led to smaller transfer gain. We interpret this as indication that, to induce transfer of learning to a visual function that shares common substrates with the trained task, training effects must be distributed across separations rather than concentrated on a subset of conditions. This interpretation is consistent with both previous training studies on lateral interaction (Polat and Sagi, 1994) and current models of PL explaining generalization and transfer of learning (Maniglia and Seitz, 2018; Xiao et al., 2008). If unequal learning across separations does gatekeep transfer of learning, future training studies using lateral masking might implement algorithms that adjust the number of trials of the difficulty of the training to ensure equal or at least similar learning curves across lambdas.

An important limitation of this study is the small number of participants, which invites caution when interpreting correlation and discussing statistical significance. While relatively common in training studies (Ball & Sekuler, 1987; Fiorentini & Berardi, 1981; Rodán et al., 2022), and specifically with lateral masking (Polat & Sagi, 1994), it still represents a major limitation to the present report's conclusions. Another limitation of this study lies in the design of Maniglia et al. (2011), which employed a yes–no paradigm. This approach is known in the collinear modulation literature to be prone to filling-in and false alarms (FAs) (Polat & Sagi, 2007). While the original paper found that the training effect did not impact the rate of FAs, suggesting that the improvement was perceptual rather than a change in response criterion, future studies measuring collinear modulation should rely on more robust 2AFC procedures.

Footnotes

Author contribution(s)

Marcello Maniglia: Conceptualization; Formal analysis; Methodology; Software; Writing – original draft; Writing – review & editing.

Russell Cohen Hoffing: Data curation; Formal analysis; Writing – review & editing.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Ethics Approval

The original study was approved by the University of Padova ethics committee.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Consent to Participate

All participants provided a written consent form and were compensated for their participation in the original study.

Consent for Publication

All participants provided a written consent form.

Availability of Data and Materials

Data will be available upon request.

Code Availability

Custom-made Matlab and R analysis codes will be available upon request.

ORCID iDs

Marcello Maniglia

Russell Cohen Hoffing

References

Anderson

E. J.

Dakin

S. C.

Schwarzkopf

D. S.

Rees

Greenwood

J. A.

(2012). The neural correlates of crowding-induced changes in appearance. Current Biology, 22, 1199–1206. https://doi.org/10.1016/j.cub.2012.04.063

Ball

Sekuler

(1987). Direction-specific improvement in motion discrimination. Vision Research, 27, 953–965. https://doi.org/10.1016/0042-6989(87)90011-3

Cai

Zhou

Fang

(2009). The effect of crowding on orientation-selective adaptation in human early visual cortex. Journal of Vision, 9, 13.1–13.10. https://doi.org/10.1167/9.11.13

Bonneh

Y. S.

Sagi

Polat

(2007). Spatial and temporal crowding in amblyopia. Vision Research, 47(14), 1950–1962. https://doi.org/10.1016/j.visres.2007.02.015

Bouma

(1970). Interaction effects in parafoveal letter recognition. Nature, 226, 177–178. https://doi.org/10.1038/226177a0

Bringuier

Chavane

Glaeser

Frégnac

(1999). Horizontal propagation of visual activity in the synaptic integration field of area 17 neurons. Science, 283, 695–699. https://doi.org/10.1126/science.283.5402.695

Cappe

Clarke

Mohr

Herzog

M. H.

(2014). Is there a common factor for vision? Journal of Vision, 14, 4. https://doi.org/10.1167/14.8.4

Cass

J. R.

Spehar

(2005). Dynamics of collinear contrast facilitation are consistent with long-range horizontal striate transmission. Vision Research, 45, 2728–2739. https://doi.org/10.1016/j.visres.2005.03.010

Chakravarthi

Cavanagh

(2009). Bilateral field advantage in visual crowding. Vision Research, 49, 1638–1646. https://doi.org/10.1016/j.visres.2009.03.026

10.

Chicherov

Plomp

Herzog

M. H.

(2014). Neural correlates of visual crowding. Neuroimage, 93, 23–31. https://doi.org/10.1016/j.neuroimage.2014.02.021

11.

Chung

S. T.

Levi

D. M.

Legge

G. E.

(2001). Spatial-frequency and contrast properties of crowding. Vision Research, 41, 1833–1850. https://doi.org/10.1016/s0042-6989(01)00071-2

12.

Cohen . (2013). Applied multiple regression/correlation analysis for the behavioral sciences. Routledge. https://doi.org/10.4324/9780203774441

13.

Doron

Spierer

Polat

(2015). How crowding, masking, and contour interactions are related: A developmental approach. Journal of Vision, 15, 5. https://doi.org/10.1167/15.8.5

14.

Fang

(2008). Crowding alters the spatial distribution of attention modulation in human primary visual cortex. Journal of Vision, 8, 6.1–6.9. https://doi.org/10.1167/8.9.6

15.

Fiorentini

Berardi

(1981). Learning in grating waveform discrimination: Specificity for orientation and spatial frequency. Vision Research, 21, 1149–1158. https://doi.org/10.1016/0042-6989(81)90017-1

16.

Gilbert

C. D.

Wiesel

T. N.

(1985). Intrinsic connectivity and receptive field properties in visual cortex. Vision Research, 25, 365–374. https://doi.org/10.1016/0042-6989(85)90061-6

17.

Gilbert

C. D.

Wiesel

T. N.

(1989). Columnar specificity of intrinsic horizontal and corticocortical connections in cat visual cortex. The Journal of Neuroscience, 9, 2432–2442. https://doi.org/10.1523/JNEUROSCI.09-07-02432.1989

18.

Grinvald

Lieke

E. E.

Frostig

R. D.

Hildesheim

(1994). Cortical point-spread function and long-range lateral interactions revealed by real-time optical imaging of macaque monkey primary visual cortex. The Journal of Neuroscience, 14, 2545–2568. https://doi.org/10.1523/JNEUROSCI.14-05-02545.1994

19.

Wang

Fang

(2019). The critical role of V2 population receptive fields in visual orientation crowding. Current Biology, 29, 2229–2236.e3. https://doi.org/10.1016/j.cub.2019.05.068

20.

Henry

C. A.

Kohn

(2020). Spatial contextual effects in primary visual cortex limit feature representation under crowding. Nature Communications, 11, 1687. https://doi.org/10.1038/s41467-020-15386-7

21.

Henry

C. A.

Kohn

(2022). Feature representation under crowding in macaque V1 and V4 neuronal populations. Current Biology, 32, 5126–5137.e3. https://doi.org/10.1016/j.cub.2022.10.049

22.

Hung

S.-C.

Seitz

A. R.

(2014). Prolonged training at threshold promotes robust retinotopic specificity in perceptual learning. The Journal of Neuroscience, 34, 8423–8431. https://doi.org/10.1523/JNEUROSCI.0745-14.2014

23.

Hupé

J. M.

James

A. C.

Payne

B. R.

Lomber

S. G.

Girard

Bullier

(1998). Cortical feedback improves discrimination between figure and background by V1, V2 and V3 neurons. Nature, 394, 784–787. https://doi.org/10.1038/29537

24.

Kapadia

M. K.

Ito

Gilbert

C. D.

Westheimer

(1995). Improvement in visual sensitivity by changes in local context: Parallel studies in human observers and in V1 of alert monkeys. Neuron, 15, 843–856. https://doi.org/10.1016/0896-6273(95)90175-2

25.

Kapadia

M. K.

Westheimer

Gilbert

C. D.

(2000). Spatial distribution of contextual interactions in primary visual cortex and in visual perception. Journal of Neurophysiology, 84, 2048–2062. https://doi.org/10.1152/jn.2000.84.4.2048

26.

Kenny

D. A.

(1979). Correlation and causality (First Edition). John Wiley & Sons, pp. 277.

27.

Lev

Polat

(2011). Collinear facilitation and suppression at the periphery. Vision Research, 51, 2488–2498. https://doi.org/10.1016/j.visres.2011.10.008

28.

Levi

D. M.

(2008). Crowding–an essential bottleneck for object recognition: A mini-review. Vision Research, 48, 635–654. https://doi.org/10.1016/j.visres.2007.12.009

29.

Levi

D. M.

Carney

(2009). Crowding in peripheral vision: Why bigger is better. Current Biology, 19, 1988–1993. https://doi.org/10.1016/j.cub.2009.09.056

30.

Levi

D. M.

Klein

S. A.

Aitsebaomo

(1984). Detection and discrimination of the direction of motion in central and peripheral vision of normal and amblyopic observers. Vision Research, 24, 789–800. https://doi.org/10.1016/0042-6989(84)90150-0

31.

Levi

D. M.

Klein

S. A.

Hariharan

(2002). Suppressive and facilitatory spatial interactions in foveal vision: Foveal crowding is simple contrast masking. Journal of Vision, 2, 140–166. https://doi.org/10.1167/2.2.2

32.

Louie

E. G.

Bressler

D. W.

Whitney

(2007). Holistic crowding: Selective interference between configural representations of faces in crowded scenes. Journal of Vision, 7, 24.1–2411. https://doi.org/10.1167/7.2.24

33.

Manassi

Sayim

Herzog

M. H.

(2013). When crowding of crowding leads to uncrowding. Journal of Vision, 13, 10. https://doi.org/10.1167/13.13.10

34.

Maniglia

Pavan

Aedo-Jury

Trotter

(2015a). The spatial range of peripheral collinear facilitation. Scientific Reports, 5, 15530. https://doi.org/10.1038/srep15530

35.

Maniglia

Pavan

Cuturi

L. F.

Campana

Sato

Casco

(2011). Reducing crowding by weakening inhibitory lateral interactions in the periphery with perceptual learning. Plos One, 6, e25568. https://doi.org/10.1371/journal.pone.0025568

36.

Maniglia

Pavan

Trotter

(2015b). The effect of spatial frequency on peripheral collinear facilitation. Vision Research, 107, 146–154. https://doi.org/10.1016/j.visres.2014.12.008

37.

Maniglia

Seitz

A. R.

(2018). Towards a whole brain model of perceptual learning. Current Opinion in Behavioral Sciences, 20, 47–55. https://doi.org/10.1016/j.cobeha.2017.10.004

38.

Maniglia

Soler

Cottereau

Trotter

(2018). Spontaneous and training-induced cortical plasticity in MD patients: Hints from lateral masking. Scientific Reports, 8, 90. https://doi.org/10.1038/s41598-017-18261-6

39.

Maniglia

Soler

Trotter

(2020). Combining fixation and lateral masking training enhances perceptual learning effects in patients with macular degeneration. Journal of Vision, 20, 19. https://doi.org/10.1167/jov.20.10.19

40.

Maniglia

Trotter

Aedo-Jury

(2019). TMS Reveals inhibitory extrastriate cortico-cortical feedback modulation of V1 activity in humans. Brain Structure & Function, 224, 3399–3408. https://doi.org/10.1007/s00429-019-01964-z

41.

Millin

Arman

A. C.

Chung

S. T. L.

Tjan

B. S.

(2014). Visual crowding in V1. Cerebral Cortex, 24, 3107–3115. https://doi.org/10.1093/cercor/bht159

42.

Moshkovitz

Maria Lev

Polat

(2022). The central visual field might mediate night vision. Journal of Vision, 22, 4262. https://doi.org/10.1167/jov.22.14.4262

43.

Motter

B. C.

(2006). Modulation of transient and sustained response components of V4 neurons by temporal crowding in flashed stimulus sequences. The Journal of Neuroscience, 26, 9683–9694. https://doi.org/10.1523/JNEUROSCI.5495-05.2006

44.

Motter

B. C.

(2018). Stimulus conflation and tuning selectivity in V4 neurons: A model of visual crowding. Journal of Vision, 18, 15. https://doi.org/10.1167/18.1.15

45.

Nandy

A. S.

Tjan

B. S.

(2012). Saccade-confounded image statistics explain visual crowding. Nature Neuroscience, 15, 463–9, S1. https://doi.org/10.1038/nn.3021

46.

Pascal

Abadi

R. V.

(1995). Contour interaction in the presence of congenital nystagmus. Vision Research, 35, 1785–1789. https://doi.org/10.1016/0042-6989(94)00277-s

47.

Pelli

D. G.

(2008). Crowding: A cortical constraint on object recognition. Current Opinion in Neurobiology, 18, 445–451. https://doi.org/10.1016/j.conb.2008.09.008

48.

Pelli

D. G.

Palomares

Majaj

N. J.

(2004). Crowding is unlike ordinary masking: Distinguishing feature integration from detection. Journal of Vision, 4, 1136–1169. https://doi.org/10.1167/4.12.12

49.

Petrov

Verghese

McKee

S. P.

(2006). Collinear facilitation is largely uncertainty reduction. Journal of Vision, 6(2), 170–178. https://doi.org/10.1167/6.2.8

50.

Polat

Mizobe

Pettet

M. W.

Kasamatsu

Norcia

A. M.

(1998). Collinear stimuli regulate visual responses depending on cell’s contrast threshold. Nature, 391, 580–584. https://doi.org/10.1038/35372

51.

Polat

Sagi

(1993). Lateral interactions between spatial channels: Suppression and facilitation revealed by lateral masking experiments. Vision Research, 33, 993–999. https://doi.org/10.1016/0042-6989(93)90081-7

52.

Polat

Sagi

(1994). Spatial interactions in human vision: From near to far via experience-dependent cascades of connections. Proceedings of the National Academy of Sciences of the United States of America, 91, 1206–1209. https://doi.org/10.1073/pnas.91.4.1206

53.

Polat

Sagi

(2007). The relationship between the subjective and objective aspects of visual filling-in. Vision Research, 47, 2473–2481. https://doi.org/10.1016/j.visres.2007.06.007

54.

Polat

Sagi

Norcia

A. M.

(1997). Abnormal long-range spatial interactions in amblyopia. Vision Research, 37, 737–744. https://doi.org/10.1016/s0042-6989(96)00154-x

55.

Rodán

Candela Marroquín

Jara García

L. C.

(2022). An updated review about perceptual learning as a treatment for amblyopia. Journal of Optometry, 15, 3–34. https://doi.org/10.1016/j.optom.2020.08.002

56.

Rodriguez

Granger

(2021). On the contrast dependence of crowding. Journal of Vision, 21, 4. https://doi.org/10.1167/jov.21.1.4

57.

Shani

Sagi

(2005). Eccentricity effects on lateral interactions. Vision Research, 45, 2009–2024. https://doi.org/10.1016/j.visres.2005.01.024

58.

Shin

Chung

S. T. L.

Tjan

B. S.

(2017). Crowding, visual awareness, and their respective neural loci. Journal of Vision, 17, 18. https://doi.org/10.1167/17.5.18

59.

Simmers

A. J.

Gray

L. S.

McGraw

P. V.

Winn

(1999). Contour interaction for high and low contrast optotypes in normal and amblyopic observers. Ophthalmic & Physiological Optics, 19, 253–260. https://doi.org/10.1046/j.1475-1313.1999.00416.x

60.

Tyler

C. W.

Likova

L. T.

(2007). Crowding: A neuroanalytic approach. Journal of Vision, 7, 16.1–16.9. https://doi.org/10.1167/7.2.16

61.

Whitney

Levi

D. M.

(2011). Visual crowding: A fundamental limit on conscious perception and object recognition. Trends in Cognitive Sciences, 15, 160–168. https://doi.org/10.1016/j.tics.2011.02.005

62.

Xiao

L.-Q.

Zhang

J.-Y.

Wang

Klein

S. A.

Levi

D. M.

(2008). Complete transfer of perceptual learning across retinal locations enabled by double training. Current Biology, 18, 1922–1926. https://doi.org/10.1016/j.cub.2008.10.030

A bridge between collinear inhibition and visual crowding: Hints from perceptual learning

Abstract

Keywords

Introduction

Method

Participants

Apparatus and Procedure

Data Analysis

Results

Correlation Between Lateral Inhibition and Critical Space of Crowding

Correlation Between Training Gain and Transfer to Crowding

Discussion

Footnotes

Author contribution(s)

Declaration of Conflicting Interests

Ethics Approval

Funding

Consent to Participate

Consent for Publication

Availability of Data and Materials

Code Availability

ORCID iDs

References