Bachmann / Francis Visual Masking

Studying Perception, Attention, and Consciousness
1. Auflage 2013
ISBN: 978-0-12-800383-1
Verlag: Elsevier Science & Techn.
Format: EPUB
Kopierschutz: Adobe DRM (»Systemvoraussetzungen)

E-Book, Englisch, 116 Seiten

Visual masking is a technique used in cognitive research to understand pre-conscious processes (priming, for example), consciousness, visual limits, and perception issues associated with psychopathology. This book is a short format review of research using visual masking: how it has been used, and what these experiments have discovered.Topics covered include concepts, varieties, and theories of masking; masking and microgenetic mechanisms and stagesof visual processing; psychopharmacological and genetic factors in masking, and more. - Provides succinct information about the widely dispersed masking studies and points out some new trends in masking research - Reviews transcranial magnetic stimulation (TMS) as an alternative to the traditional psychophysical masking methods - Comments on the methodological pitfalls hidden in the practice of masking, helping to improve the quality of future research where masking is used as a tool - Informs readers about recent developments in theoretical attempts to understand masking

Talis Bachmann is a professor in the departments of law and psychology at the University of Tartu in Estonia, specializing in Cognitive and Forensic Psychology. He is also head of the Perception and Consciousness group in the Estonian Center of Behavioral and Health Sciences. He is on the Executive Board of the Union of Estonian Psychologists, and is a member of the Association for Scientific Studies of Consciousness, and Association for Psychological Science. He currently serves on the board of Consciousness and Cognition, was the former co-editor of Acta Universitatis Scientiarum Socialium et Artis Educandi Tallinnensis, and was a former board member on The European Journal of Cognitive Psychology. He is an author of 190 academic publications.Talis is regarded as one of the leading experts in masking, microgenesis, and perception of pixelated visual images.

Autoren/Hrsg.

Bachmann, Talis

Francis, Gregory

Leseproben

7.2 Masking Combined with EEG Recordings
An essential part of the fMRI-based masking studies is their attempt to dissociate unconscious visual processing from processing that leads to conscious perception and thereby to find brain imaging signatures that could be regarded as markers of NCC. Largely similar aims were set by the research groups using EEG/ERP combined with visual masking for these purposes (Del Cul, Baillet, & Dehaene, 2007; Fahrenfort et al., 2008; Koivisto et al., 2008; Lamy, Salti, & Bar-Haim, 2009; Railo & Koivisto, 2009). Compared to fMRI, EEG affords much better temporal resolution, which is important in a fast transient paradigm such as masking. Moreover, sometimes ERPs may tap consciousness-related processes better than fMRI. Schoenfeld, Hassa, Hopf, Eulitz, & Schmidt (2011) recorded ERPs and fMRI activity in a patient with hysterical blindness whose “scotoma” was spatially limited to the left upper and right lower visual quadrant. (Other two intact quadrants could be therefore used as controls.) Importantly, while the fMRI responses to visual stimulation were normal in all quadrants and did not discriminate awareness from unawareness, ERP/N1 was sensitive to the seeing versus not seeing contrast. After successful treatment for the hysterical blindness, the N1 amplitude in response to the stimulus became equal between all quadrants. High-density recordings of ERPs allowing cortical source reconstruction were used in combination with a metacontrast like task by Del Cul et al. (2007). Objective and subjective (seen vs. guessed) reports were used. By varying target-mask SOA researchers manipulated objective and subjective visibility and compared its dynamics with ERP recordings. They contrasted ERPs obtained from mask-only control conditions with ERPs associated with objective and subjective performance in target masking conditions. An ERP difference between mask-only and unseen/objective masked target discrimination conditions observed with less than 250 ms after target onset indicated a considerable amount of subliminal processing (attributed to the activity in the occipitotemporal pathway). Conscious reportability was associated with activity after about 270 ms and interpreted as a result of highly distributed fronto-parieto-temporal activation. Unfortunately, the interpretations of this study (Del Cul et al., 2007) cannot be conclusive. By subtracting the ERP activity evoked by the mask only from the ERP obtained in the target-plus-mask conditions, the authors aimed to isolate the entire sequence of target-evoked ERPs. For this, they first aligned the ERPs on the mask onset, then subtracted the ERP evoked by the mask alone from each of the other target-plus-mask conditions, and then realigned the subtracted data on target onset. However, this method lies on a questionable implicit assumption that mask-evoked and target-evoked components in ERP are independent and additive. However, because responses to a mask can be considerably modulated by the processes evoked by the preceding target (see Bachmann, 1994, 2009b; Scharlau, 2007) the analysis by Del Cul and colleagues may be misleading. Moreover, it is not clear from an ERP whether the change in the expression of its components between visibility conditions signifies changed target perception or changed mask perception. For example, if the mask response is amplified (and we cannot unambiguously attribute some ERP component to target or mask processing), this means also that the target response may be relatively weakened and vice versa. Furthermore, the total energy contribution of the brief small target (16 ms, one small letter) and the long larger mask (250 ms, four letters) suggests that the ERPs are sculpted mostly not by the sensory-perceptual factors, but by the response-related factors helping subjects to prepare and execute responses. Thus, the signatures are more likely signatures of the NCCpr and/or NCCae but not the direct NCC (see Aru et al., 2012 and de Graaf, Hsieh, et al., 2012 for the relevant discussion). In a different ERP/masking study Koivisto et al. (2008) recorded brain potentials while the observers had to detect a pattern-masked stimulus-dot (83 ms). Parameters were set to leave the stimulus near the subjective threshold. Subjects were instructed not to guess and tried to respond on the basis of their conscious visual experience. Target hits as compared to target misses showed an ERP negativity around 180–350 ms (occipital and posterior temporal sites) and a positivity at 400–500 ms (peaking at parietal sites). Importantly, these correlates of successful perception were independent from manipulation of attention. Koivisto and collegues called this early negativity visual awareness negativity (VAN), hypothesized to be a primary ERP-correlate of visual awareness as different from signatures of preconscious processing or attention. In a subsequent study Railo and Koivisto (2009) combined metacontrast with ERP recordings. Targets (17 ms) were followed by an effective mask (17 ms) or a similar but ineffective pseudomask after a varying SOA (0–130 ms). Again, the so-called VAN was found to associate with target visibility. Again, the results of this study, while extending the evidence for an interesting ERP signature associated with electrophysiological negativity, are inconclusive because of several methodological problems common to most of the NCC studies using masking and brain imaging (see Bachmann, 2009b, for a more detailed discussion of this issue). For example, varying levels of awareness associated with variations in ERP were covarying with objective stimulation parameters such as SOAs and kinds of stimuli. It is not clear whether ERP signatures differentiating between aware and not aware trials were caused by the awareness-related processes or by the differences in the processing of physically different stimulation. The too long delay from the stimulus until VAN (more than 330 ms) exceeds estimates of the delay with which a phenomenal percept emerges after stimulus onset as found in many other studies (100–200 ms). Furthermore, timing the emergence of a subjective conscious percept is complicated due to interactivity between the target and mask, including the putative modulations of the ERP-response to the mask by the effects of the preceding target. Finally, because of the ambiguity of the concept of NCC (Aru et al., 2012; de Graaf, Hsieh, et al., 2012) it is not specified whether the correlate is a direct correlate of awareness intimately associated with the processes that are necessary for awareness occurring only during the episode of target experience but not before and/or after it. The pattern masking experiment by Lamy et al. (2009) introduced an important advantage for masking studies of awareness. They avoided the counterproductive confound between stimulation variability and target awareness variability by comparing ERPs associated with aware and not aware responses. Both objective sensitivity to target location and subjective evaluation of visibility were used while data was collected from trials that were identical in terms of physical stimulus, exposure time, and level of objective performance (target correct responses only). Objective as well as subjective performance measures were associated with the amplitude of the P300+ component of the ERP. Because the P300+ amplitude from the trials with unaware/correct responses was considerably higher than that from the unaware/incorrect trials, preconscious above-chance discriminability of targets was demonstrated (the effect was primarily associated with parietal electrodes). However, as the P300+ associated with aware/correct trials was clearly more pronounced than with unaware/correct trials, the NCC could be extracted. In the latter case, the modulation due to awareness was widely spread across most of the scalp locations. Lamy et al. (2009) argue that signatures of awareness are related to late ERP waves. However, as seen from their Figure 3, some difference in favor of aware conditions can be observed already at shorter ERP components. The weak statistical effect in comparing the early ERP components between awareness conditions may be due to the fact that a big part the sensory-evoked potentials up to about 200 ms reflect not only target processing but also mask processing. If, for example, the target is only a couple of tens of ms and the mask, which occupies a spatial area about 10 times larger than target, is also about 10 times longer than target, then the contribution of the target-related perceptual processes to the ERP is negligible. Instead, and artificially, the ERP dynamics reflects differences in response-level processing. Masking in combination with ERP recordings has also been used by Victor Lamme and his coworkers in the context of the role of reentrant corticocortical processes in constituting NCC. Fahrenfort et al. (2007) compared detection of a texture-defined square under nonmasked (seen) and masked (unseen) conditions. An EEG analysis was undertaken to single out signatures hypothetically associated with reentrant processing. These signatures were absent in the masked condition. Extrastriate visual areas were activated early in time by both seen and unseen stimuli, which the authors interpret as showing that feedforward processing is preserved, even though subject performance is at chance objective measures. The authors conclude that masking is caused by later processes that disrupt reentrant processing and, in essence, masking is a...