EEG and Electrophysiology

Effects of closure on the processing invisible figures: An ERP study

Weina Zhu and Yuanye Ma

The specific nature of closure is important for the computation of an object representation. The discrimination of closure is typically faster and easier than that of other geometrical features (Treisman and Paterson, 1984). In order to investigate whether the specific nature of closure required conscious perception, we recorded event-related potentials (ERPs) during a continuous flashed suppression paradigm (CFS), in which the target was displayed in one eye to compete against a flashed full contrast Mondrian pattern presented to the other eye. This method ensured that the target image was suppressed during the experiment. During most trails, the closed figures were more likely and faster to over come the flashed suppression than open figures (469ms vs. 488ms; 18.3% vs. 9.6%). ERPs showed invisible closed figures to evoked greater N1 amplitude than open figures on the left temporal-parietal areas (-2.07uv vs. -1.73uv; t (15) = 2.76, P = 0.015) with slightly shorter latency (121.8ms vs. 129.4ms). Visible figures induced opposite results, with closed figures evoking smaller N1 amplitude (-5.3uv vs. -8.6uv). Our results indicate that closure can be processed in the absence of awareness, and showed a perceptional advantage of closed figures, which were more rapid and lightly to overcome flashed suppression. Reference Treisman, A., Paterson, R., 1984. Emergent features, Attention, and object perception. J Exp Psychol Hum Percept Perform. 10, 12-31.

Learning new faces from multiple images

Chris Longmore, Isabel Santos, Mariana Carrito, Ana Pereira, Paulo Rodrigues and Carlos Silva

It has been suggested that familiar and unfamiliar faces employ different processing mechanisms with familiar faces recognised via an invariant structural code and unfamiliar faces from the pictorial codes contained within an image of a face [Bruce and Young, 1986, British Journal of Psychology, 77, 305-327]. Recent research has suggested however that perhaps all faces, including familiar faces are recognised through pictorial coding [Longmore, Liu and Young, 2008, Journal of Experimental Psychology: Human Perception and Performance, 34, 77-100]. We present a series of experiments showing that when recognition across viewpoints (e.g. full-face to profile) of a face is required, recognition does indeed appear to be tied to the images studied. However, using the same experimental paradigm, evidence from ERPs indicates that there might be some degree of facilitation in learning two viewpoints (frontal and profile) when recognition on a previously unseen view (three-quarter) is required compared to learning a single view. This suggests that some kind of structural information is being extracted during learning that facilitates recognition from novel views. It would appear that whilst the behavioural data indicates a predominately pictorial basis for the recognition of newly learnt faces, neuro-correlates suggest greater levels of invariance.

Active Vision Plugin: An open-source MATLAB tool for saccade- and fixation-related EEG analysis

Olaf Dimigen and Ulrich Reinacher

Although natural vision is fundamentally transsaccadic, EEG data on vision is usually recorded during fixation. An alternative approach to signal analysis, used in several recent studies (for review see Dimigen et al, 2011, JEP:General, 140, 552-572), is to time-lock the EEG to saccade- or fixation onsets during free viewing. Recording high-resolution eye movements with the EEG is also useful for controlling fixation, detecting signal distortions from microsaccades, measuring saccadic reaction times, improving EEG artifact correction, or simultaneous pupillometry. To facilitate joint analyses of oculomotor and electrophysiological data, we present a plugin to the popular open source MATLAB toolbox EEGLAB, which parses, imports, and synchronizes eye tracking data and adds it as extra channels to the EEG. Eye trackers from SMI (IView X) and SR Research (EyeLink) are currently supported. Saccades and fixations are detected with the algorithm of Engbert & Mergenthaler (2006) or imported from the raw EDF file and then added to the EEGLAB event structure. Additionally, ICA components can be objectively selected or rejected based on their covariance with the electrically independent eye track. We exemplify usage of the plugin for an experiment in which we investigated the contribution of extraretinal signals to fixation-related potentials during visual search.

An Electrophysiological Investigation of Attentional Bias Towards Fearful and Angry Faces in a PTSD Population

Dhrasti K Shah, Colin Cameron, Dylan Smith, Natalia Jaworska, Crystal Blais, Derek Fisher, Verner Knott and Charles A Collin

Evidence from probe detection tasks suggests that anxious individuals exhibit biased (enhanced) selective attention to visual threat stimuli, such as angry and fearful faces. Research has demonstrated that redirection of attentional focus can reduce fear response, suggesting that competing cognitive processes may interfere with or diminish facial expression processing. The current studyused event-related potential (ERPs) to examine the effects of cognitive load on attentional bias towards ecologically threatening stimuli (emotional faces) in 20 posttraumatic stress disorder (PTSD) patients and 20 healthy controls (HC) in two tasks: a) dot-probe visual (DPV); and b) dot-probe divided-attention (DPD). DPV task required participants to respond to the presentation of the visual target (following a pair of faces), while ignoring the auditory tones. The DPD task required participants to respond to the presentation of visual and auditory targets. Both groups showed early attentional bias to angry and fearful faces as indexed by P100 and face-specific N170 visual ERP components. DPD acted to diminish attentional bias in the PTSD group. The results provide electrophysiological support for early visual attentional biases toward angry and fearful faces in PTSD and possible moderating effects of divided attention.

Electroencephalographic evidence of normal inversion effect even with impaired holistic processing in high-functioning adults with autism.

Paula Tavares, Susana Mouga and Miguel Castelo-Branco

To determine if face processing impairment (recognition of identity, emotion, gender and age) observed in people with Autism Spectrum Disorders (ASD) result from early dysfunction of the visual ventral pathway, we measured event-related potentials in 16 typically developing controls and 7 subjects with high-functioning ASD performing a face decision task on a wide range of face and non-face stimuli. With exception of the Mooney faces, ASD subjects performed identically to controls, in the identification of photographic and schematic faces. ASD subjects showed the normal face inversion-effect, with longer latencies and higher N170 amplitudes over occipital electrodes to inverted rather than upright photographic faces. However, a lateralization effect was found, with right hemispheric N170 amplitudes in controls higher in all face categories than in ASD subjects. The results can be interpreted in the light of current models of face processing as sparing of the first-order configural level and impairment of the higher holistic level of face processing (Maurer et al, 2002, Trends in Cognitive Sciences, 6(6), 255-260). These results suggest that early-intermediate anomalies of the visual ventral pathway during face processing might underlie the face processing deficits detected in ASD.

The face sensitive N170 component in developmental prosopagnosia: effects of stimulus type and inversion

John Towler, Angie Gosling, Bradley Duchaine and Martin Eimer

Individuals with developmental prosopagnosia (DP) show severe face recognition deficits in the absence of any history of neurological damage. To ascertain the time-course of atypical face processing in DP, we measured the face-sensitive N170 component of the event-related brain potential (ERP) in a group of sixteen adult participants with DP and sixteen control participants. Reliable enhancements of N170 amplitudes in response to upright faces relative to houses were found for the DP group. This effect was equivalent in size to the effect observed for controls, demonstrating face-sensitivity of the N170 component in DP. In marked contrast, most DPs showed atypical N170 face inversion effects. In the control group, N170 components were reliably enhanced and delayed for inverted faces compared to upright faces, indicative of processing differences for these stimuli. In the DP group, the difference between upright and inverted faces was absent. The absence of typical N170 face inversion effects in the DP group suggests that deficits in the structural encoding of faces are a common source of face recognition impairments in individuals with DP.

Evidence for a unified representation of faces from EEG frequency tagging stimulation

Renaud Laguesse, Adriano Boremanse, Anthony M. Norcia and Bruno Rossion

Behavioral studies have shown that the processing of the top half of a face is no longer influenced by the identity of the bottom half when they are spatially misaligned. However, evidence for the integration of face parts into a unified ('holistic') representation when the two halves are aligned vs. misaligned is still lacking. To address this issue, we used the frequency-tagging stimulation technique during the EEG recording (Regan & Heron, 1969). Ten participants were presented with faces composed of top/bottom parts flickering at different frequencies (5.88 or 7.14 Hz). The parts could either be aligned, or spatially misaligned (5 levels of misalignment). The contiguous border between the top/bottom halves was kept constant across conditions by increasing the size of the stimulus with misalignment. The recorded EEG was transformed to the frequency domain. For the fundamental frequencies of stimulation, frequency-specific increases were found in the medial-occipital region for all conditions. Most importantly, over the right occipito-temporal channels, EEG increases were found at intermodulation terms (i.e., 5.88 + 7.14 = 13.01 Hz) for the aligned condition only. This study provides objective evidence for a unified representation of faces in the right occipito-temporal cortex.

Cortical excitability and effective connectivity in face processing: a TMS-EEG study.

Giulia Mattavelli, Leonor J Romero Lauro, Mario Rosanova, Elena Natale, Sara Romanella and Costanza Papagno

Neuroimaging and electrophysiological studies have shown that a fronto-temporo-occipital network is involved in face processing. However, the actual connectivity and the interactions between these areas remain unclear. Transcranial magnetic stimulation combined with electroencephalography (TMS-EEG) allows non-invasive and direct measurements of excitability and effective connectivity in the human cerebral cortex. This is the first study using TMS-EEG to explore local cortical excitability and long-range effective connectivity at rest and during different face processing behavioural tasks. Single-pulse TMS was delivered over the medial prefrontal cortex (mPFC) while face identity or face expression matching tasks at 100 ms from face stimulus onset, while continuous EEG was recorded using a 60-channel TMS-compatible amplifier. TMS-evoked potentials (TEPs) recorded during the behavioural tasks were compared with TEPs recorded during a passive point fixation and event related potentials (ERPs) evoked by the tasks without TMS. Preliminary results revealed that the behavioural tasks modified local TEPs produced by mPFC stimulation. Moreover, TMS interfered with the early EEG components in the temporo-occipital regions. These findings contribute to clarify task-related activity in the face cortical network and suggest that fronto-occipital connections are involved in the early stage of face processing.

Early Neural Correlates of Familiarity during Binocular Rivalry

Bradley N. Jack, Urte Roeber and Robert P. O'Shea

An infrequent visual stimulus-a 'deviant'-yields a more negative event-related potential (ERP) 200 ms after the event than frequent visual stimuli-'standards': the visual mismatch negativity (vMMN). We sought the vMMN during binocular rivalry by presenting 80% standards-an 180° phase shift in one grating to one eye every 500 ms and the same phase shift in an orthogonal grating to the other, and 20% deviants-either by swapping the phase-shifting gratings between the eyes or by rotating them 45°-to 16 participants who pressed keys to report their experiences of rivalry. We counterbalanced which stimuli were standards and deviants. We found only a more positive ERP for the standard 100 ms after the event. This is not only opposite to the vMMN, but opposite to what would be expected from adaptation. We are exploring the parameters of this finding, such as whether it is something about rivalry (by making the stimuli identical in the two eyes) or something about the task (by relieving the participant of the necessity to record rivalry). Our initial results indicate it has something to do with rivalry, possibly because its perceptual changes make deviants seem more frequent then they are.

The influence of ongoing brain oscillations on apparent motion perception

Lia Sanders, Ryszard Auskulewicz, Niko Busch and Philipp Sterzer

Several studies have demonstrated a link between ongoing oscillatory alpha activity and upcoming stimulus detection. Here we sought to investigate how prestimulus brain oscillations influence the perception of apparent motion (AM), a visual illusion that occurs when two blinking dots are perceived as one single moving stimulus. The perception of AM depends on the presentation frequency. At low presentation frequencies, the percept varies between sequential alternation and smooth movement (AM). At higher frequencies, perception varies between AM and two simultaneously flickering dots. Using stimulation frequencies at which individual subjective perception fluctuated between motion and alternation or motion and flicker, we asked participants to classify their percepts, while we recorded EEG signals. We found that the influence of prestimulus alpha power on AM perception varied with presentation frequency. At the low presentation frequency, fronto-occipital alpha power was positively associated with AM perception, whereas at the high presentation frequency, a decrease in ongoing alpha power accompanied the motion percept. In addition, prestimulus fronto-occipital phase-locking in the alpha band was positively related to AM perception irrespective of stimulation frequency. These results point to a role of ongoing alpha power and fronto-occipital phase synchronization in the perception of AM motion.

EEG-correlates of different degrees of visual ambiguity

Rike Wörner, Michael Bach and Jürgen Kornmeier

During observation of ambiguous stimuli our percept is instable. Disambiguated stimulus variants stabilize our percepts and evoke a huge positive event-related potential 400 ms after stimulus onset ('P400'). No such P400 had been found for the ambiguous stimuli. This parietal/central P400-effect was found for the geometric Necker cube and Borings semantic Old/Young woman. In the current study we tested for a P400 effect in von Schiller's stroboscopic alternative motion stimulus ('SAM') and studied whether the P400 amplitude follows a stepwise reduction of stimulus ambiguity. Methods: Participants compared consecutive geometric cube stimuli and SAM stimuli with different degrees of ambiguity and indicated perceptual reversals manually. We calculated separate ERPs for stimulus categories and degrees of ambiguity. Results: For both stimulus categories the P400 amplitude increased monotonically with decreasing ambiguity. An earlier, frontal/central ERP positivity ('P200') showed similar activity pattern. Discussion. Stimulus ambiguity - across very different categories and thus at a very abstract processing level - seems to be processed as early as 200 ms after stimulus onset and are reflected in the P200 amplitude. The P200 may reflect a subconscious evaluation of the perceptual stimulus interpretation whereas the P400 may reflect conscious access.

Cortical localization of second-order visual mechanisms

Denis Yavna, Vitaly Babenko and Alexey Soloviev

The aim of our investigation is to determine localization of second-order visual mechanisms. For that we used mismatches between EPs to a non-modulated texture and those textures modulated in contrast, orientation, and spatial frequency. Response to a modulated texture differs by the fact that the same operation chain is now added by the work of second-order filters. The above 'inserted operation' should introduce definite changes in the EP. Each observer (20 persons) was to identify the presenting textures. EEG was recorded by 18 leads. The first step was to obtain, for each observer, EPs for 4 textures. The next step was to obtain the differences between the responses to the non-modulated texture and each of the modulated textures. The third step was connected with obtaining averaged mismatches in the whole test group for each difference and in each lead. This enabled to establish those time periods when obtained mismatches were not accidental. We entered averaged mismatches of these time periods in the EEGLAB program and localized the sources of these mismatches using functions of DIPFIT2 plugin. It was found that the mechanisms which provide perception of different modulations have not the same localizations within dorsal visual pathway.

Fixation-related potentials to faces in a free-viewing visual search task

Lisandro Kaunitz, Juan E. Kamienkowski, Mariano Sigman, Rodrigo Quian Quiroga and Matias J. Ison

The use of event related potentials (ERPs) has been almost exclusively limited to situations where eye movements are precluded. However, when subjects explore a scene or look for a target in the real world, they systematically produce discrete sequences of fixations, gathering and integrating information in each instance of the sequence. Despite the compelling contribution of ERPs and eye movements to cognitive neuroscience, to date only a handful of experiments have exploited both techniques simultaneously. In the present study we concurrently recorded EEG and eye movements in a free-viewing visual search paradigm. Subjects were asked to find a hidden target face in a crowded scene including distractor faces. We analyzed fixation-related potentials (fERPs) to target and distractor stimuli. Additionally, we compared the fERPs obtained in our free-viewing visual search task with the ERPs in an oddball control experiment performed without eye movements. Our findings show the emergence of fERP components associated with the perception of faces and the detection of targets, contributing towards a better understanding of brain activity during free viewing in real-world-like scenarios.

Maturation of motion-onset VEPs in children

Zuzana Kubova, Jan Kremlacek, Jana Langrova, Jana Szanyi and Miroslav Kuba

The motion-onset VEPs (M-VEPs) display distinct age dependent changes. The latencies of their main negative peak N160 are steeply shortening up to 18 years of age and the VEPs are also changing their shape during this time (Langrova et al., Vision Research, 2006, 46, 536-544). The aim of our study was to determine those parameters of motion stimulation that have principal influence on shape and latencies of the M-VEPs in children. M-VEPs were elicited by translational linear movement of vertical bars or checks and by radial movement of concentric circles. We changed spatial frequency of structure (0.2 - 1 c/deg), contrast of stimulus luminance (0.1 and 0.95), its modulation (sinusoidal versus square-waved) and velocity of motion (5 - 25 deg/s). M-VEPs to linear translation were more similar to adult M-VEPs than those to radial movement. Amplitude of N160 peak increased and its latency decreased at full square-waved contrast stimulus with higher spatial frequency. Reported M-VEPs sensitivity to stimuli parameters corresponds to slow maturation of contrast sensitivity, which should be respected during electrophysiological examination of children. Acknowledgements: Supported by GAČR, project P407/12/2528 and Charles Univ.-Faculty of Medicine in Hradec Kralove (PRVOUK).

Visual evoked potentials to peripheral stimulation

Miroslav Kuba, Jan Kremlacek, Frantisek Vit, Jana Langrova, Jana Szanyi and Zuzana Kubova

Electrophysiological examination of visual perception with the use of visual evoked potentials (VEPs) is based so far almost exclusively on a central stimulation, since standard pattern stimuli are not enough effective to produce VEPs outside central ca. 20° of the visual field. However, when a visual motion stimulation is used, reactions to peripheral stimuli are well recordable, surprisingly mainly in the fronto-central cortical brain areas. These VEPs are namely not related so much to visual information processing, they represent predominantly significant event related reactions. With adequate arrangement of the motion stimuli in periphery (in such a way that they do not interfere with normal activities of subjects), it could be possible to perform a long-term monitoring of brain activity and to evaluate a level of visual information processing or vigilance in situations where signalization of critical changes is required (mobile visual simulator is a subject of the Czech patent application No. PV 2010-545). Supported by Charles University - Faculty of Medicine in Hradec Kralove, Czech Republic (PRVOUK).

Electrophysiological study of congenital prosopagnosia

Marta Zimmer, Kornel Nemeth and Gyula Kovacs

Congenital prosopagnosia (CP) is a long-life impairment in face recognition despite normal low-level visual processing. Previously several case studies investigated different aspects of CP. However, there are no complex studies describing the whole neuropsychological, behavioural and electrophysiological profile of these subjects as of yet. The goal of the present study was to investigate the face specificity of the N170 component in CP patients (3 members of a family - father, son, and daughter) and in age-matched controls. Subjects performed a face - non-face categorization task for 5 categories of stimuli (faces, inverted faces, shoes, chairs, and Fourier images /NOISE/). When comparing the N170 component to FACES and NOISE we have found a significant amplitude reduction and a reduced inter-trial coherence (ITC) for NOISE stimuli in controls. In CP, however, no differences were found between these conditions. Surprisingly, the face evoked N170 was similar in CP and controls. Moreover, both ERP and ITC for NOISE correlated with the performance of the individuals in the face recognition tests (CFMT and PFPB). Our results suggest that it is the processing of the noisy stimuli that is altered in CP.

Temporal dynamics of responses to dark and light stimuli in visual cortex

Stanley Jose Komban, Jianzhong Jin, Yushi Wang, Reza Lashgari, Jens Kremkow, Qasim Zaidi and Jose-Manuel Alonso

Recent physiological and psychophysical studies demonstrate a dominant representation of darks at early stages of visual processing. Here we investigate the response temporal dynamics of this dark dominance at the input layers of the cat visual cortex using a 16/32 channel multi-electrode array. We measured the response time course of each cortical site using white noise (WN) stimuli (224 ON, 424 OFF). Cortical neurons in layer 4 had shorter response latency to dark than light stimuli, however, the response to lights were faster at the response decay time and at the rebound suppression that followed the decay. Finally, consistent with our human psychophysical measurements, responses for dark targets were faster than light when presented on uniform binary WN background. However, this temporal advantage is abolished when dark pixels in the WN background was increased to 70%. Overall, we demonstrate that V1 responses to dark and light stimuli differ in latency, time of response decay and rebound suppression. In addition, we find that the latency advantage for darks disappears when the proportion of background dark pixels approaches the value at which detection time is similar for darks and lights in humans.

Effect of central vs. peripheral optic flow stimuli on postural responses revealed by surface electromyography

Alessandro Piras, Milena Raffi, Michela Persiani and Salvatore Squatrito

While retinal optic flow (OF) is considered a major cue for self-motion and postural stability, it is still debated the relative importance of central vs peripheral vision in visuo-postural responses. Psychophysical studies, by assessing the body sway in response to OF stimuli give more importance to visual periphery as source of compensatory postural responses, while the central vision would play an accessory role. Aim of this study was to verify the different influence of central vs peripheral optic flow stimulation on the activation of lower limb postural muscles: bilateral tibialis anterior and gastrocnemius medialis. Surface EMG was recorded in 24 male and female participants while radial (expansion and contraction) OF stimuli were presented either in a central or peripheral region of a wide screen. The activity of the right tibialis anterior was the most affected by OF stimuli. We found an interaction of sex by muscle in favour of male subjects (78%). The majority of the subjects had a greater postural response for either central or peripheral stimulation, meaning that the influence of visual field region on visuo-postural balance may be subject-specific, probably depending on other stimulus parameters.