Elsevier

NeuroImage

Volume 115, 15 July 2015, Pages 17-29
NeuroImage

Sharpened cortical tuning and enhanced cortico-cortical communication contribute to the long-term neural mechanisms of visual motion perceptual learning

https://doi.org/10.1016/j.neuroimage.2015.04.041Get rights and content

Highlights

  • Motion perceptual learning improves the neural selectivity in V3A.

  • Motion perceptual learning enhances the effective connectivity from V3A to IPS.

  • A combination of the selectivity and connectivity increases explains motion learning.

Abstract

Much has been debated about whether the neural plasticity mediating perceptual learning takes place at the sensory or decision-making stage in the brain. To investigate this, we trained human subjects in a visual motion direction discrimination task. Behavioral performance and BOLD signals were measured before, immediately after, and two weeks after training. Parallel to subjects' long-lasting behavioral improvement, the neural selectivity in V3A and the effective connectivity from V3A to IPS (intraparietal sulcus, a motion decision-making area) exhibited a persistent increase for the trained direction. Moreover, the improvement was well explained by a linear combination of the selectivity and connectivity increases. These findings suggest that the long-term neural mechanisms of motion perceptual learning are implemented by sharpening cortical tuning to trained stimuli at the sensory processing stage, as well as by optimizing the connections between sensory and decision-making areas in the brain.

Introduction

Training can improve performance for many visual tasks, which is referred to as visual perceptual learning (VPL) (Sagi, 2011, Watanabe and Sasaki, 2014). The neural basis of VPL has generated a lot of interest in the past decades and at the same time is highly controversial. VPL is commonly characterized by its specificity to the trained stimulus, leading to the hypothesis that the underlying neural changes occur in early visual areas (Fahle and Poggio, 2002). However, this hypothesis was mainly based on psychophysical data and was supported only inconsistently from neurophysiological studies (Schoups et al., 2001, Ghose et al., 2002, Yang and Maunsell, 2004, Furmanski et al., 2004, Hua et al., 2010). Even within psychophysics, however, several recent studies show that the specificity of VPL can be eliminated by training with an easy task (Liu, 1999, Liu and Weinshall, 2000), training with a different task (Xiao et al., 2008), or mere exposure to a different stimulus (Zhang et al., 2010b), suggesting that VPL is mediated by higher cortical areas. So far, this alternative hypothesis has been little tested with neurophysiological methods, though researchers have found learning-related neural changes in decision-making and attention-related areas such as intraparietal sulcus (IPS) and anterior cingulate cortex (Lewis and Van Essen, 2000, Law and Gold, 2008, Kahnt et al., 2011).

Another important, but unanswered question in VPL is its long-term neural mechanisms. To date, almost all VPL studies have focused on immediate neural changes after training. However, persistency, a hallmark feature of VPL, remains largely unknown inside the brain. Yotsumoto et al. (2008) tracked the neural changes induced by a texture discrimination task, and found a transient activity enhancement in V1. This enhancement receded after two weeks, while the behavioral improvement persisted. Parallel findings have been reported in auditory and motor modalities. For example, cortical expansion was detected immediately after training in both modalities, but faded out weeks later (Molina-Luna et al., 2008, Reed et al., 2011). These findings raised questions on previous claims that are based on neural changes immediately after training, and prompted us to investigate what longer-lasting neural changes may be associated with VPL.

In our current investigation, we studied visual motion perceptual learning, with two specific aims: whether the neural modifications occurred at low- or high-level, and what neural modifications may be longer-lasting. Human subjects were trained in a motion direction discrimination task. Their behavioral performance and BOLD signals were measured before, immediately after, and two weeks after training. We examined not only how learning affected the local representation of the trained motion direction within individual visual cortical areas and IPS, a motion decision-making area and homologue of monkey LIP (lateral intraparietal area) (Kayser et al., 2010), but also how learning changed the effective connectivities between the visual areas and IPS. Law and Gold, 2008, Law and Gold, 2009 modeled the learning process as a high-level decision unit refining its connectivities to sensory neurons tuned to a specific motion direction through response reweighting (Poggio et al., 1992, Dosher and Lu, 1998, Bejjanki et al., 2011). However, there is no empirical evidence yet directly supporting this hypothesis.

Here, we report that, parallel to the long-lasting motion discrimination improvement, the neural selectivity in V3A and the effective connectivity from V3A to IPS for the trained direction exhibited a persistent increase after training, as revealed by both decoding and encoding analyses and dynamic causal modeling (DCM). We found that the behavioral learning could be well explained by a linear combination of improvements from these two sources. These findings make headways towards resolving previous controversies and demonstrate that perceptual learning should be attributed to changes both in the sensory representation of trained stimuli and the transmission of sensory signals to decision circuitry.

Section snippets

Subjects

Seventeen subjects (nine female) participated in the study. They were naïve to the purpose of the study and had never participated in any perceptual learning experiment before. All subjects were right-handed with reported normal or corrected-to-normal vision and had no known neurological or visual disorders. Their ages ranged from 20 to 25 years. They gave written, informed consent in accordance with the procedures and protocols approved by the human subject review committee of Peking University.

Psychophysical results

Subjects underwent eight daily training sessions (1000 trials per session) to discriminate motion directions around a pre-specified, but randomly selected direction (hereafter the direction is referred to as 0°). In a trial, two random-dot kinematograms (RDKs) with slightly different directions were presented sequentially. Subjects were asked to make a two-alternative forced-choice (2-AFC) judgment of the direction of the second RDK relative to the first one (clockwise or counter-clockwise) (

Discussion

Our study provides the following psychophysical and neuroimaging findings. (1) Motion direction discrimination training improved behavioral performance, which was specific to the trained direction and persisted for at least two weeks. This finding replicated the work by Ball and Sekuler (1987). (2) Immediately after training, the mean BOLD signal in V3A responding to the trained direction decreased, but this decrease mostly vanished two weeks later. (3) The decoding accuracy and neural

Acknowledgments

We thank Zhong-lin Lu for his helpful discussion. This work was supported by the National Natural Science Foundation of China (31230029, 31421003, 91132302 and 90820307) and the Ministry of Science and Technology of China (2015CB351800 and 2012CB825500). Zili Liu was supported in part by a US NSF grant (BCS 0617628) and a China NSFC grant (31228009).

References (70)

  • H. Lu et al.

    Learning motion discrimination with suppressed MT

    Vision Res.

    (2004)
  • K. Molina-Luna et al.

    Motor learning transiently changes cortical somatotopy

    Neuroimage

    (2008)
  • G.A. Orban et al.

    Similarities and differences in motion processing between the human and macaque brain: evidence from fMRI

    Neuropsychologia

    (2003)
  • A. Reed et al.

    Cortical map plasticity improves learning but is not necessary for improved performance

    Neuron

    (2011)
  • D. Sagi

    Perceptual learning in vision research

    Vision Res.

    (2011)
  • C. Schiltz et al.

    Neuronal mechanisms of perceptual learning: changes in human brain activity with training in orientation discrimination

    Neuroimage

    (1999)
  • J.T. Serences et al.

    Computational advances towards linking BOLD and behavior

    Neuropsychologia

    (2012)
  • J.T. Serences et al.

    Estimating the influence of attention on population codes in human visual cortex using voxel-based tuning functions

    Neuroimage

    (2009)
  • A.M. Smith et al.

    Investigation of low frequency drift in fMRI signal

    Neuroimage

    (1999)
  • L.M. Vaina et al.

    Can spatial and temporal motion integration compensate for deficits in local motion mechanisms?

    Neuropsychologia

    (2003)
  • M.B. Wall et al.

    The representation of egomotion in the human brain

    Curr. Biol.

    (2008)
  • B.A. Wandell et al.

    Visual field maps in human cortex

    Neuron

    (2007)
  • L.Q. Xiao et al.

    Complete transfer of perceptual learning across retinal locations enabled by double training

    Curr. Biol.

    (2008)
  • Y. Yotsumoto et al.

    Different dynamics of performance and brain activation in the time course of perceptual learning

    Neuron

    (2008)
  • M. Bao et al.

    Perceptual learning increases the strength of the earliest signals in visual cortex

    J. Neurosci.

    (2010)
  • A. Bartels et al.

    Natural vision reveals regional specialization to local motion and to contrast-invariant, global flow in the human brain

    Cereb. Cortex

    (2008)
  • V.R. Bejjanki et al.

    Perceptual learning as improved probabilistic inference in early sensory areas

    Nat. Neurosci.

    (2011)
  • G.M. Boynton et al.

    Linear systems analysis of functional magnetic resonance imaging in human V1

    J. Neurosci.

    (1996)
  • G.J. Brouwer et al.

    Decoding and reconstructing color from responses in human visual cortex

    J. Neurosci.

    (2009)
  • G.J. Brouwer et al.

    Cross-orientation suppression in human visual cortex

    J. Neurophysiol.

    (2011)
  • B.A. Dosher et al.

    Perceptual learning reflects external noise filtering and internal noise reduction through channel reweighting

    Proc. Natl. Acad. Sci. U. S. A.

    (1998)
  • S.A. Engel et al.

    Retinotopic organization in human visual cortex and the spatial precision of functional MRI

    Cereb. Cortex

    (1997)
  • M. Fahle et al.

    Perceptual Learning

    (2002)
  • K.J. Friston

    Dynamic causal models for fMRI

  • G.M. Ghose et al.

    Physiological correlates of perceptual learning in monkey V1 and V2

    J. Neurophysiol.

    (2002)
  • Cited by (44)

    • Boosting visual perceptual learning by transcranial alternating current stimulation over the visual cortex at alpha frequency

      2022, Brain Stimulation
      Citation Excerpt :

      VPL is usually viewed as a manifestation of experience-dependent neural plasticity in the human brain [2,3], and has been used as a rehabilitation therapy for visually impaired people, such as patients with amblyopia [4], cortical blindness [5], and macular degeneration [6]. Numerous studies have been conducted to explore the neural substrates of VPL, primarily focusing on the cortical sites of occurrence and manifestation of VPL [7–13]. Neural oscillations, characterized by rhythmic changes in neural activity in a wide range of frequencies, play a critical role in various cognitive functions [14].

    • Recurrent Processing Drives Perceptual Plasticity

      2020, Current Biology
      Citation Excerpt :

      These results suggest that learning-dependent changes in orientation-specific representations in superficial V1 layers are independent of task context, consistent with previous neurophysiology results [9] showing learning-dependent changes in neural tuning in primary visual cortex during a fixation task. We next considered learning-dependent changes in decision-making-related regions [6, 38] in the parietal cortex. In particular, we focused on IPS1 and IPS2 that have been implicated in perceptual decision making [38].

    • Effects of regular aerobic exercise on visual perceptual learning

      2018, Vision Research
      Citation Excerpt :

      VPL is specific to the trained stimulus configuration (Gold & Ding, 2013) and has been linked to changes within areas of the visual cortex, such as V1, V2, V4 and the middle temporal area (MT), that contain populations of neurons tightly tuned to stimulus features including orientation, spatial location and motion (Ball & Sekuler, 1987; Gilbert, Sigman, & Crist, 2001; Raiguel, Vogels, Mysore, & Orban, 2006; Schoups, Vogels, Qian, & Orban, 2001; Zohary, Celebrini, Britten, & Newsome, 1994). Despite a high affinity of VPL to the trained stimulus properties, transfer of learning to untrained stimuli has also been observed (Lagas et al., 2016; Liu, 1999; Liu & Weinshall, 2000), and is thought to rely on higher-level cortical areas, such as those involved in decision-making (Chen, Cai, Zhou, Thompson, & Fang, 2016; Chen et al., 2015). Thus, VPL is likely to involve neural plasticity at multiple stages of visual processing.

    • Towards a whole brain model of Perceptual Learning

      2018, Current Opinion in Behavioral Sciences
    View all citing articles on Scopus
    1

    These authors contributed equally to this work.

    View full text