Transcranial magnetic stimulation of early visual cortex interferes with subjective visual awareness and objective forced-choice performance

https://doi.org/10.1016/j.concog.2010.09.001Get rights and content

Abstract

In order to study whether there exist a period of activity in the human early visual cortex that contributes exclusively to visual awareness, we applied transcranial magnetic stimulation (TMS) over the early visual cortex and measured subjective visual awareness during visual forced-choice symbol or orientation discrimination tasks. TMS produced one dip in awareness 60–120 ms after stimulus onset, while forced-choice orientation discrimination was suppressed between 60 and 90 ms and symbol discrimination between 60 and 120 ms. Thus, a time window specific to visual awareness was found only in the orientation condition at 120 ms. The results imply that both conscious and unconscious perception depend on activity in early visual areas. On the basis of previous estimates of neural processing speed, we suggest that the late part of the activity period most likely involve local extrastriate–striate interactions which provide the contents for visual awareness but are not themselves sufficient for awareness to arise.

Introduction

Our understanding of the organization of visual cortex and its function at neuron level has grown rapidly during recent decades, but at system level it is not clear how the populations of neurons interact and give rise to conscious percepts. One important question concerns the contribution of the primary visual cortex (V1) on visual awareness. According to views based on the classical feedforward conception, visual processing occurs hierarchically starting from V1 and proceeds in feedforward manner to higher cortical areas where the necessary and sufficient activity producing awareness takes place (e.g., Crick and Koch, 1995, Zeki and Bartels, 1999). Alternative models assume that feedforward processing may only be sufficient for unconscious processes, but visual awareness would require also recurrent processing where the higher areas interact with the earlier ones (Bullier, 2001, Hochstein and Ahissar, 2002, Lamme et al., 1998, Pollen, 2008). This view assumes that early visual areas participate in conscious perception during two periods: during an “early” feedforward phase and during a later feedbackward phase. Only the late V1 activity is assumed to be specific to visual awareness (Block, 2005, Lamme, 2006). Although there is evidence from neuroimaging studies that the activity level of early visual cortex determines the perceptual contents that will enter visual awareness (Kamitani and Tong, 2005, Ress and Heeger, 2003), it remains to be explicitly tested whether the late V1 activity is specific to visual awareness or whether it is necessary also for unconscious perception.

Here we study whether there is a distinct period during which the activity of early visual cortex is specific to visual awareness by making use of transcranial magnetic stimulation (TMS). It allows one to stimulate a cortical region through the skull and to interfere with cortical processing (Walsh & Pascual-Leone, 2003). By applying single TMS pulses at varying stimulus onset asynchronies (SOA) between the visual stimulus and the TMS pulse, one can make causal inferences concerning the contribution of the stimulated area on visual processing in specific time windows. Therefore TMS is a potentially useful method for testing whether or not the early visual cortex contributes to visual awareness during one or more critical periods. In a classical study (Amassian et al., 1989) the participants were asked to identify randomly generated letter trigrams and it was found that TMS applied on occipital cortex masked perception between 60 and 120 ms after stimulus onset. At intervals 80–100 ms, a blur or nothing was seen. After these findings, the masking effect of TMS on early visual cortex around 100 ms has been replicated in many studies (e.g., Beckers and Hömberg, 1991, Camprodon et al., 2010, Corthout et al., 2002, Corthout et al., 2003, Corthout et al., 1999, Corthout et al., 1999, Kammer et al., 2005, Masur et al., 1993, Sack et al., 2009). However, evidence for two distinct critical time windows, predicted by models assuming feedback from extrastriate areas to V1, has been controversial. Corthout et al., 1999, Corthout et al., 1999 and Corthout et al., 2002, Corthout et al., 2003 report four different time periods during which TMS over the primary visual cortex impairs letter identification. However, the four dips were never observed in the same experiment or in a single participant. In addition to the typical ∼100 ms dip, some of the participants showed an early 20 ms dip (Corthout, Uttl, Walsh et al., 1999). Corthout et al. suggested that this early dip is related to feedforward sweep, whereas the later 100 ms dip reflects disturbance of recurrent processing. Later experiments, however, did not replicate the early dip (Corthout et al., 2000, Corthout et al., 2002). TMS applied at negative SOAs (that is, before the visual stimulus was presented) also induced interference which was most probably due to eye blinks (Corthout et al., 1999, Corthout et al., 2000) or disruption of pre-stimulus activity (Corthout et al., 2003). As human V1 is activated on average about 60 ms after the appearance of a stimulus (Baseler and Sutter, 1997, Foxe and Simpson, 2002, Vanni et al., 2001, Wilson et al., 1983), the early dips near the stimulus onset cannot be considered as correlates of online cortical processing as the stimulus has not even reached the visual cortex at the time when TMS is applied. In a recent study, Camprodon et al. (2010) found that TMS over occipital pole impaired categorization of animals 100 ms and 220 ms after stimulus onset. They suggested that the 100 ms dip reflected disruption of feedforward processing, whereas the 220 ms dip represented disturbance of feedback to V1.

There is some evidence that a late V1 activity period is necessary for visual awareness in motion perception. Pascual-Leone and Walsh (2001) found that TMS over V1 disrupted experiences of motion in phosphenes (conscious experiences of flashes of light) when applied after V5 stimulation, implying that V5/MT activation reaches awareness via V1 (see also Silvanto, Cowey, Lavie, & Walsh, 2005). In order to study awareness of real motion, Silvanto, Lavie, and Walsh (2005) applied TMS over either the early visual cortex or V5/MT at different time windows after the presentation of a moving stimulus. They found two periods of early visual cortex activity, one preceding and another postdating the V5/MT critical period, suggesting that feedback from V5/MT to early visual areas is critical for awareness of motion. This result was recently replicated with the addition that the later early visual cortex period was necessary also for unconscious perception of motion (Koivisto, Mäntylä, & Silvanto, 2010), suggesting that the late activity period of early visual cortex is not specific to visual awareness of motion.

Although these findings suggest that recurrent activity and V1 are important in conscious motion perception, causal evidence is less clear for two distinct periods of V1 activity in perception of stationary stimuli. If the early dip that is induced by TMS before or near the onset of visual stimulus is considered as a pre-stimulus effect that does not reflect online processing of the stimulus or interpreted as a non-specific effect of TMS (e.g., Sack, Kohler, Linden, Goebel, & Muckli, 2006), one is left with the classical ∼100 ms dip in majority of the studies. There is no agreement on interpretation of the ∼100 ms TMS dip: does it reflect the disruption of the feedforward sweep (Camprodon et al., 2010) or recurrent processing (Breitmeyer et al., 2004, Corthout et al., 1999)? It is also unclear whether the broad 100 ms dip is unique to tasks which require visual awareness, or whether it reflects impairments also in tasks which can be performed without awareness. From the perspective of consciousness studies, it is important to note that the majority of the earlier TMS masking studies have not measured subjective awareness but identification of letter trigrams (Amassian et al., 1989, Beckers and Hömberg, 1991, Masur et al., 1993) or forced-choice discrimination of letters (Corthout et al., 1999, Corthout et al., 1999 (Corthout et al., 2002, Corthout et al., 2003), symbols (Sack et al., 2009), or pictures (Camprodon et al., 2010). Already Amassian et al. (1989) noted that during the letter trigram identification task” at times the subject gave the correct response, but claimed to have ‘guessed’ rather than ‘seen’ the letters”. This observation suggests that unconscious as well as conscious processes were contributing to performance. More recently, Boyer, Harrison, and Ro (2005) applied TMS over the early visual cortex 100–128 ms after stimulus onset and found that the participants were able to discriminate the orientation and color of stimuli better than expected by chance even in trials in which they subjectively reported not being aware of the stimulus. Unfortunately, this study did not examine systematically the effects of TMS as a function of SOA. The important implication of the study is that “objective” discrimination tasks are not optimal measures of visual awareness, because they may be performed better than chance on the basis of unconscious processes alone. Therefore, it remains possible that the TMS studies that have found only one dip around 100 ms (by using objective measures) have missed a later dip that is specific to visual awareness. A study that uses both subjective measures of visual awareness and “objective” forced-choice measures of visual perception is needed to clarify the contribution of early visual cortex on visual awareness in different time windows.

Here we used single-pulse TMS to investigate the necessity of early visual cortex activity in visual awareness as a function of time. The observers performed a forced-choice discrimination task and provided a rating of their subjective perceptual experience of the same stimuli on each trial. Two different stimulus types, which have provided conflicting results concerning the role of early visual cortex in conscious vs. unconscious perception (Boyer et al., 2005, Sack et al., 2009), were included: the stimulus was either an arrow symbol pointing upward or downward (symbol discrimination condition) or a bar with horizontal or vertical orientation (orientation discrimination condition). While Boyer et al. (2005) found that TMS to early visual cortex impaired specifically visual awareness but not unconscious discrimination of orientation, Sack et al. (2009) did not find any evidence for a similar dissociation in perception of symbol stimuli. Thus, if there exists a late period of activity in early visual cortex that is unique to visual awareness (e.g., Boyer et al., 2005), the application of TMS during this period should impair aware visual experience but leave forced-choice discrimination performance unaffected. On the other hand, if the critical activity period of early visual cortex is central to both visual awareness and unconscious perception (e.g., Koivisto et al., 2010, Sack et al., 2009), rather than being a specific marker of visual awareness, TMS should impair both subjective awareness and objective forced-choice discrimination performance during this period.

Section snippets

Participants

Fourteen healthy participants with normal or corrected-to-normal vision were tested (four males; mean age = 22.1 years, range: 19–26). The experimental condition (symbol vs. orientation) was manipulated between two groups of participants (n = 7 + 7), with gender and age balanced between them. The participants were right-handed (Oldfield, 1971) and naïve to the purpose of the experiment. The experiment was undertaken with the understanding and written consent of each subject. The study was accepted by

Awareness

First, we tested whether or not we were able to elicit retinotopic visual suppression by comparing the ratings of subjective awareness between contralateral and ipsilateral visual field (relative to coil position) (Fig. 3). The visual field (2) × SOA (10) × Condition (2) ANOVA showed that field interacted with SOA, F(9, 108) = 3.15, p < .01. As expected, this interaction was due to stronger suppression of visual awareness at 60, 90, and 120 ms SOAs in the contralateral visual field than in the

Discussion

In order to clarify the contribution of early visual cortex on visual awareness in different time windows, we studied the suppressive effects of single-pulse TMS on subjective ratings of awareness and on “objective” forced-choice discrimination performance. The aim was to examine whether there is a time window at which the early visual cortex contributes specifically to visual awareness. The results revealed only a single, relatively long TMS-induced dip in visual awareness 60–120 ms after

Acknowledgments

This work was supported by the Academy of Finland [Grant Number 125175]. Author N.S.-V. was supported by the Graduate School of Psychology in Finland. We thank Teemu Laine for technical help and Tor Lehtonen for running the TMS experiment.

References (50)

  • M. Koivisto et al.

    The role of early visual cortex (V1/V2) in conscious and unconscious visual perception

    Neuroimage

    (2010)
  • V.A.F. Lamme

    Towards a true neural stance on consciousness

    Trends in Cognitive Sciences

    (2006)
  • V.A.F. Lamme et al.

    The distinct modes of vision offered by feedforward and recurrent processing

    Trends in Neuroscience

    (2000)
  • V.A.F. Lamme et al.

    Feedforward, horizontal, and feedback processing in the visual cortex

    Current Opinion in Neurobiology

    (1998)
  • H. Masur et al.

    Suppression of visual perception by transcranial magnetic stimulation – experimental findings in healthy subjects and patients with optic neuritis

    Electroencephalography and Clinical Neurophysiology

    (1993)
  • R.C. Oldfield

    The assessment and analysis of handedness: The Edinburgh Inventory

    Neuropsychologia

    (1971)
  • T. Ro

    Unconscious vision in action

    Neuropsychologia

    (2008)
  • A.T. Sack et al.

    The temporal characteristics of motion processing in hMT/V5+: Combining fMRI and neuronavigated TMS

    Neuroimage

    (2006)
  • A.T. Sack et al.

    Symbolic action priming relies on intact neural transmission along the retino-geniculo-striate pathway

    Neuroimage

    (2009)
  • S. Vanni et al.

    Multifocal fMRI mapping of visual cortical areas

    Neuroimage

    (2005)
  • S. Zeki et al.

    Toward a theory of visual consciousness

    Consciousness and Cognition

    (1999)
  • G. Beckers et al.

    Impairment of visual perception and short-term memory scanning by transcranial magnetic stimulation of occipital cortex

    Experimental Brain Research

    (1991)
  • C.N. Boehler et al.

    Rapid recurrent processing gates awareness in primary cortex

    Proceedings of the National Academy of Sciences of the United States of America

    (2008)
  • J.L. Boyer et al.

    Unconscious processing of orientation and color without primary visual cortex

    Proceedings of the National Academy of Sciences of the United States of America

    (2005)
  • J.A. Camprodon et al.

    Two phases of V1 activity for visual recognition of natural images

    Journal of Cognitive Neuroscience

    (2010)
  • Cited by (31)

    • The conscious experience of color constancy and neural responses to subliminal deviations – A behavioral and EEG/ERP oddball study

      2020, Consciousness and Cognition
      Citation Excerpt :

      The neural mechanisms that mediate unconscious vision and their relation to the emergence of conscious perception remain elusive. Unconscious processing has been documented both in lesion (Brent, Kennard, & Ruddock, 1994; Cowey, 2010; Pöppel, Held, & Frost, 1973; Stoerig & Cowey, 1989, 1992; Weiskrantz, Warrington, Sanders, & Marshall, 1974), as in healthy human data (Breitmeyer, Ogmen, & Chen, 2004; Breitmeyer, Ro, & Singhal, 2004; Breitmeyer, Ro, Ogmen, & Todd, 2007; Hurme, Koivisto, Henriksson, & Railo, 2020; Hurme, Koivisto, Revonsuo, & Railo, 2017, 2019; Koivisto, Railo, & Salminen-Vaparanta, 2011; Railo, Andersson, Kaasinen, Laine, & Koivisto, 2014; Railo, Salminen-Vaparanta, Henriksson, Revonsuo, & Koivisto, 2012; Schmidt, 2002, 2000). The role of early visual cortex in this debate has remained controversial.

    View all citing articles on Scopus
    View full text