You are on page 1of 8

BR A IN RE S EA RCH 1 3 98 ( 20 1 1 ) 6 4 –71

available at

Research Report

One plus one is less than two: Visual features elicit
non-additive mismatch-related brain activity

István Sulykos⁎, István Czigler
Institute for Psychology of the Hungarian Academy of Sciences, Budapest, Hungary
Eötvös Loránd University, Budapest, Hungary


Article history: In a passive oddball task (performing in a video game), participants were presented with
Accepted 4 May 2011 sequences of either standard stimuli or patterns containing deviant orientation, deviant
Available online 12 May 2011 spatial frequency or both deviant orientation and spatial frequency. Orientation deviants
presented to the lower half of the visual field elicited a posterior negative component with a
Keywords: peak latency of 130 ms. Spatial frequency deviants elicited a similarly negative component
Event-related potential that was later followed by another negative component. Activity elicited by the double-
Visual mismatch negativity and deviant stimulus was identical to activity elicited by the orientation deviant alone. The
mismatch response subtraction difference of the peak latency and scalp distribution of the deviant minus the
Non-additivity standard difference potentials were unequal to those of the exogenous event-related
potential (ERP) components and were therefore considered visual mismatch negativities
(vMMNs). The non-additivity of the feature-related responses is interpreted as sensitivity of
the implicit change-detection system to deviant events rather than an exclusive sensitivity
to individual features. Deviant stimuli presented to the upper half of the field elicited
responses with positive polarity, but this activity was less pronounced than the vMMN.
Polarity reversal of the response to upper half-field stimulation suggests that the origin of
the activity lies in retinotopic areas. Because of the emergence of a mismatch component
with positive polarity, we propose that the term visual mismatch negativity (vMMN) be
replaced with the more general term visual mismatch response (vMMR).
© 2011 Elsevier B.V. All rights reserved.

1. Introduction irrelevant but irregular events is usually investigated with the
passive oddball paradigm. In this paradigm, stimuli with rare
Events violating regularities of environmental stimulation will deviant features (such as pitch, tone intensity, color and
elicit brain activity even when those events have no imme- orientation) are presented in sequences of frequent identical
diate relevance. This brain activity can be recorded as either (standard) stimuli, during which the participants attend to
electric or magnetic mismatch negativity (MMN) in the task-related stimuli or their attention is diverted from the
auditory (for review, see Kujala et al., 2007; Näätänen et al., standard and deviant in other ways. In our study, we
2007), visual (for review see Czigler, 2007) and somatosensory presented two deviant features in the visual modality and
(e.g. Shinozaki et al., 1998) modalities. Brain activity elicited by asked whether processing of these two features is

⁎ Corresponding author at: Institute for Psychology of the Hungarian Academy of Sciences, 1394 Budapest, P.O. Box 398, Hungary. Fax: + 36 1
354 2416.
E-mail address: (I. Sulykos).

0006-8993/$ – see front matter © 2011 Elsevier B.V. All rights reserved.

feature-related those reported by Jeffreys and Axford (1972). With the present study. 2004).34.9. In the auditory modality. activity in response to two visual stimulus features (orienta. By two-way ANOVA (with factors Half-field and Deviancy) no p < 0. Additivity of the responses to upper-half field stimulation was tested by ANOVA with the factors of Additivity. Takegata et al. For lower-half stimulation. For the frequency deviant/lower- in case of specific electrode locations (Paavilainen et al.and upper-half stimulus features such as pitch.6. standard subtracted from the deviant.49. additivity.30. 1 (A and B). The (Czigler et al.56. the analysis on vector-scaled values showed significant Laterality main effect. η2 = 0. It seems that the larger Additivity. (11) = 2.2. 1996) or additivity hold. with the largest peak amplitude at other cases. most likely the prestriatal amplitudes at the later range. Exogenous potentials elicited a positive–negative–positive ERP sequence. in the analyses of (Horváth et al.67–3. with latencies . albeit only anterior scalp distribution.2.1. F(2. 1 (A and B) also shows the difference potentials of the reported additive MMN (Schröger. C2 and C3 at the (vMMN).9. violation of regular stimulus du. we investigated the possibility of different scalp these two regularities is separated by 100 ms (Czigler and distributions for the modeled and observed double deviant Winkler.003 (Bonferroni criterion). but that sub-additivity reflects saturation (or limited deviancy. 153 and 244 ms.05. 1995. ε = 0. single peak. even when the detection of First. The main effects In the present study we investigated deviance-related of Additivity and Latency range were significant. Results and Laterality. Anteriority 2.45 and F(1. we estimated (modeled) and observed (actual) double-deviant sought to provide further evidence supporting this idea. or whether vMMN is generated by a unified mecha. These difference 1999) and magnetic mismatch field (Levänen et al..6.. F(2.7 ms and 89. whereas In general. As shown in Fig. in the case of triple-deviant stimuli. p < 0. η2 = 0. and 101.. 2003. 2008).22) = 3.1.g. tively.0001–0. intensity. a test showed the difference was also significant. 2004). However. MMNs in response to various stimulus features. Furthermore. deviant-related components are presented in Table 2. In this case. p < 0. η2 = 0. Furthermore. p < 0. lower-half field stimulation were 97.. whereas the Laterality main effect was 2. C2 the latency values of these components differed. Fig. 2001) or combination of features (Takegata negative peaks. These results can be attributed to the smaller activity of either the lower or upper half of the visual field. It is 5.. 2001). It should be noted that sub-additivity of the mismatch Emergence of the deviant-minus-standard difference was response to stimuli violating two (or more) deviant features reliable in the six t-tests (two half-field. three types of devi- does not prove that the individual deviant features are treated ancy each) in the earlier range of difference potentials.9 ± 2. sequence.11) = 8. p < 0. η 2 = 0. 2. Additivity × Anteriority. 1993) to potentials were markedly different for lower.. 1996). Average amplitudes and peak latencies of the sub-additivity was recorded (Paavilainen et al. pitch 2. lower-half field stimuli.26. difference potentials at lower-half field stimulation. stimulating the lower half of the field 2.2. t(11) = as integrated units (as opposed to a set of isolated features).05. indicating that the activity source is a other significant main effect revealed the generally smaller retinotopic part of the visual cortex. F(1.11) = 6.73. Additivity effect and duration) elicits only one MMN. interactions.75. whereas for upper half-field stimulation. and Additivity × Laterality.051. the sub-additivity of the deviants). These interaction effects re- significant effects were found (group averaged RT and HIT flect the distribution difference between the observed double rates were 551 ± 5. some results suggest independent The peak latencies of the ERP components are shown in Table 1. The three exogenous components were labeled C1. 1998). the non-scaled values were used. Behavioral results η2 = 0. 1 (C) illustrates the areas (Pazo-Alvarez et al. as studies using the auditory modality have shown. BR A IN RE S E A RCH 1 3 98 ( 20 1 1 ) 6 4 –7 1 65 independent or common. MMNs (Winkler and Czigler. field stimuli. F(1. The mean latencies mechanisms are responsible for visual mismatch negativity for the standards at the three conditions for C1. Typical the observed double-deviant brain response relative to the feature-related vMMN is elicited by the lower half of the field modeled response (i. respec- nism that is sensitive to regularity violation per se.. Furthermore. However. whereas the five other conditions each had a et al.. respectively). tures was larger than the MMN elicited by double-deviant the difference potentials were positive with broader and more stimuli (Czigler and Winkler. ε = 0.05. These studies Fig. Thus.05.. the three exogenous ERP components had opposite stimulating the upper half elicited a negative–positive–negative polarities for lower. respec- tion and spatial frequency of Gábor patches) presented to tively.22) = 5. For the later possible that feature violations are indeed processed sepa. Anteriority and Laterality. the sum of MMNs in response to individual fea.11) = 29. vector-scaled values ration and a sequential regularity rule elicits two distinct were tested by ANOVA with the factors being Latency range. even though their latencies differed somewhat from answer the question of whether separate. p < 0. t temporal accuracy) of the MMN-generating mechanism. the difference location and deviant inter-stimulus interval. stimulus duration. p < 0. in potentials were negative.2. deviance (midline parieto-occipital) and the additive model (right centro-parietal). There was no significant MMN amplitude resulting from double deviants is due to the interaction between Additivity and the distribution-related more frequent detection of more prominent irregular events Anteriority and Laterality factors. ERP results due to the smaller effect over the left side.79. One goal of this research was to and C3. Accordingly.1%.e. a deviant event violating two different regularities (e. occipital locations.. 1996.and upper-half field stimuli. half field stimulation the difference potentials had two Wolff and Schröger.001.2. latency range at lower-half field stimulation/spatial frequency rately. 169 and 262 ms for upper-half field stimulation. 2001).

and the additive model (orientation plus frequency difference potential) is shown on C. 9 comparisons in total).66 BR A IN RE S EA RCH 1 3 98 ( 20 1 1 ) 6 4 –71 being generally shorter with lower-half field stimulation than (Component× Deviancy. p < 0. only C2 latency differences were deemed these latency differences by t-test for each deviant condition significant. 1 – Event-related activity and deviant-standard difference potentials at the location having the amplitude of the largest difference potentials (A) and exogenous C2s (B). Voltage maps show latency ranges of amplitude analyses. spatial frequency and double deviant conditions.001–0. For validation we compared ferroni criterion.003. t(11) = 5. Fig.88.63–3. Using the Bon- for upper-half field stimuli (Table 1). Deviant-related activity in the orientation. .

C4.73. as well as a significant double deviant can be interpreted as a hierarchical depen- Component×Deviancy interaction. and Anteriority × Laterality. Furthermore.1) C3 242 (4. η2 =0.71.8) 266 (7. Discussion 2.89.8) 154 (2.0) 153 (3.7) C2 153 (3.2) 169 (4.19.4. F(1. 241 ms) was clearly absent for the double deviant.78 (0. Pz. BR A IN RE S E A RCH 1 3 98 ( 20 1 1 ) 6 4 –7 1 67 Table 1 – Mean peak latencies (ms) of the exogenous components (S.. The absence of the later In an ANOVA we compared C2 and difference potential latencies vMMN at the double deviant argues against saturation as an at lower-half field stimulation.001. ε=0. ε = 0. Oz. with a p < 0.05.11) = 35.11) = 29.3) 257 (5.9) 96 (4.22) = 32.1. following electrodes: C3.23.. F(4.001. and Heslenfeld.82. p < 0.M. 3. p < 0. whereas the latter effect reflects larger The main finding of the present study is the clear sub- amplitudes over the midline. F(2. spatial frequency Laterality main effect. The former interaction reflects the Czigler and Sulykos. amplitudes in the orientation and double-deviant conditions ally a shorter difference potential latency.22)=12.39. Anteriority and Laterality.22) −2. interactions ponent described previously (e. in parenthesis).44) = 3. η2 =0. F(2.E.50) 1.54. η2 =0. F(2.46) 1.5) 248 (5. half field stimuli was observed as a negativity (vMMN). Latency deviant conditions were similar. ε =0.3) 243 (4. 2010. Whereas the similar vMMN was significant.8) 129 (3.99 (0.25. Deviancy Orientation Spatial frequency Double Half-field Lower Upper Lower Upper Lower Upper vMMRearly Amplitude −2.22) Latency 130 (2. F(1.77.68. Scalp distribution In the case of lower-half field stimulation.39. Cz. with 14. and Deviance× half field stimuli this difference is typical of the vMMN com- Laterality. the later nega- other two effects reflect the right parieto-occipital distribution tivity to the spatial frequency deviant (with a mean latency of of the components. p < 0. and Deviancy.29.2) vMMRlate Amplitude −0.2. p < 0. the Component× Deviant and standard stimuli elicited different ERPs. cannot be interpreted in this fashion.5. p<0.M in parenthesis).5) 263 (5. η 2 = 0.05. with shorter difference potential latency). η2 = 0. t(11)= spatial frequency and double). Deviancy Orientation Spatial frequency Double Half-field Lower Upper Lower Upper Lower Upper C1 99 (2. p<0. mean latency of 130–140 ms to both the orientation and the ε = 0. A similar test can be explained by a saturation effect.75. Comparison between exogenous components and η2 =0. The absence of the later mismatch-related activity to the ε = 0. main effects.22) = 7.0) 2.2) 163 (3. and orientation.22) = 7.001.34 (0. of deviant-related negativity relative to the latency the latter factors (calculated from a 3 × 3 grid including the of the exogenous C3.57. we used ANOVA to compare the vector-scaled spatial frequency condition. To investigate the possibility of a distribution or latency dif.27) −1.001.9) 174 (4. The earlier deviant-related activity to lower- and significant Component × Anteriority.1. P4.0) 103 (4. ε = 0.57. F(2.1) .90.55 (0.9) 132 (3. Kimura et al. O2) used only for distribution analyses. For these ANOVA tests the factors were as follows: spatial frequency deviant at lower-half field stimulation con- Component (C2.E. η2 = 0. Deviancy (orientation. Deviancy effects were due to the smaller latency difference in the ence potentials.05.1.4) 138 (4. C3 latency and the latency of the amplitude values and the peak latencies of C2 and the earlier later difference potential component were compared only for the negativity..01.001.18 (0. O1. The Component effect was similar to the lower-half field difference potentials stimulation (that is. spatial frequency deviants. According to the Tukey HSD tests. were deemed significant. Accordingly.7) 98 (3.22) = 4.9. F(4. whereas the the vMMN to the double deviant. At lower- Anteriority. interactions.52. additivity of event-related brain activity to the violation of For upper-half field stimulation we observed a significant regularities in the case of two visual features. p <0.44) =3. ε = 0.9) 97 (4.0) 166 (3. p < 0. the difference potentials for the orientation and the double- 2.23.27 (0. p < 0. dence of deviance-related pre-attentive mechanisms. η2 = 0. showing gener.2.18 (0.9) 104 (4.2. difference potential).9. The Component × Anter. The main effect of Component interpretation of sub-additivity.01. η2 = 0.31) Latency 241 (4. the Deviancy and Component× ference between the exogenous components and the differ. F(2. Astikainen et al. The summed activity of the two iority interaction reflects the more anterior maximum of the individual deviant-related components was much larger than deviant-related activity (in comparison to C2). 2009 for orientation more posterior distribution of the deviant-related activity (in deviancy. The resulting t-test revealed a shorter peak latency. ditions. comparison to C2).0) 136 (4. the missing later vMMN with upper-half field stimulation yielded significant Component.31) 0.g.30. This is Table 2 – Mean amplitude (μV) and peak latencies (ms) of the mismatch responses at lower and upper half-field stimulations (S. 2008.3. 2003 for spatial frequency deviancy). P3.36.

Kremláček et al. Paavilainen et al. As a conclusion the task-related features of the standard stimuli (Berti and Schröger. (1974) who investigated the selective attention to a specific some studies of change-related visual activity showed that feature in the context of another task-irrelevant feature. but there is evidence for motion-related distribution differences were found between the exogenous vMMR. 2006. in the case of spatial investigate the feature-related hierarchical relations is the frequency/lower-half field stimulation the second posterior magnocellular–parvocellular subdivision of the visual system. treatment of the various features could be considered stimu. Thus. feature. separate deviant features. mismatch responses. 1989) modalities. 2006b) and irregular repetition of stimuli (Czigler et al. 2006.g. larger responses to infrequent stimuli can be attributed frequency and orientation (i. (2010a) localized orientation- However the hierarchical relations among the automatic related vMMN into the pre-striatal (BA19) and prefrontal (BA47 processing of various features are still debated. distribution of the vMMR to upper-field ativity) were smaller or absent when another feature with stimulation also differed from the vMMR to lower-field sti- more effective processing indicated a non-target stimulus. one plus one is less than two.... Paid students (6 male and 6 female) of a combined mean age However.68 BR A IN RE S EA RCH 1 3 98 ( 20 1 1 ) 6 4 –71 similar to the processes involved in attentive target detection. only lower-half field stimuli elicited vMMN. obvious parvocellular features. Attempts to localize the precise source of vMMR Smid et al. Czigler et al.1 Another obvious line of research to the C2 exogenous component. dent. 2005. Pazo-Alvarez et al.g. ERP was less positive to the deviant). Marqui..g. even if selective refractoriness contributes to the therefore be valuable to investigate possible interactions deviant-standard difference. ed in the experiment. This finding confirms earlier audi. Smid et al.. dimensional distribution of current source density (Pascual- (2001). In addition to related components (N1/processing negativity. Another novel finding of our study is the emergence of positive deviance-related activity to upper-half field stimuli. while the results found by Harter and Aine (1984) was As an alternative interpretation of the deviant-related interpreted as a feature-related serial processing of spatial activity. Amenedo et al. Third. identical to the earliest mismatch-related activity to stimuli there is a temporal limit of these interactions (Winkler et al. simple but reasonable explanation for the difference between In both auditory (e. The study has been carried out in accordance with the Declaration of Helsinki.. Czigler et al. sensitivity to the deviant features in our task. the polarity difference. present study also seem contradictory to the refractoriness tigating integral dimensions would be a test of the generality of explanation. Näätänen et al. 2006a. It is also important to investigate the extent In conclusion. In some previous studies (Clifford et al. more anterior).g. A study inves.. a discussion of a similar controversy in auditory modality see related exception. mulation (i. First. Two features are tinct posterior activity only when frequent (standard) stimuli separable (e.. 2005. Participants lower-half field stimulation.. May and Tiitinen... Written informed consent was obtained ance effects than the term “visual mismatch negativity”. both stimulus omission (Czigler et al.3 (SD) years with either normal or corrected-to- related components makes the term “visual mismatch normal visual acuity and with normal color vision participat- response” (vMMR) more appropriate for describing the devi. with regard to visual 1998).e. and Winkler et al. 2010b) elicited vMMR. (1998). 1983) and visual our results and previous studies is the possibility of larger (e. the majority of vMMR research has focused on ponent (i.e. 2010. indicating that the mismatch Accordingly. selection neg. the stimulus location dependence of the mismatch. (1997) reported context dependent relations to other features and feature conjunctions are another possible between the task-related processing of color and shape. peak latency of the earlier vMMR preceded the present findings.. violating only a single feature. Compared to the vMMN to the 4. task and context dependent with some fixed. In a recent study using sLORETA method of cortical three- tory reports by Czigler and Winkler (1996).g. Second. Mazza et al. stimuli with two features different from of interactions between the mechanisms underlying the regular stimulation elicit a mismatch-related activity that is feature-related vMMRs. However. 2002). Hansen and Hillyard. discarded. Furthermore.e. It would Therefore. dimensions (e. cellular systems.. (e. 2008. As an example and 11) areas. 2003.. One from all participants. 2009. features does not interfere. lightness and saturation) is mutually depen.1. negativity emerged together with a positive exogenous com- To date.g. Winkler and Czigler (1998). Similar conclusion was drawn by Garner e. Several aspects of the frequency and orientation) were investigated. Kenemans et al. spatial frequency is processed to the refractoriness of exogenous activity that is specific to the faster than orientation). Due to a large number of artifacts. the mismatch-related positivity (vMMP) of the present study is a less defined component. but see Berti. for lus. while the processing of integral 2002). genuine mismatch-related activity. we consider reported effects to be between vMMRs generated in the magnocellular and parvo. Kimura et al. 1997. a deviance-related activity of the magnocellular system components and the vMMRs in cases of both visual fields. In the present study separable dimensions (spatial Kimura et al. refractoriness cannot adequately explain the emergence of Garner (1974) defined two types of relations between simulta. 2010). of 22 ± 2. data from one additional participant were 1 We thank an anonymous reviewer for this suggestion. rating vMMN are sensitive to deviant events as well as to 2004). 2007). if the processing of such were in the sequence (Astikainen et al. . Experimental procedures 2004. Wijers et al.. responses are generated in retinotopic areas (Czigler et al. deviant-standard differences. line of future researches.. color and shape)... Rare visual stimuli elicited dis- neously appearing feature dimensions.. forasmuch in the auditory modality... 4. 2004). 2006. it seems that the processes responsible for gene. Kimura et al. feature.

ing ±70 μV on any channel were rejected from further analysis. The common parameters of the images were the Gaussian standard deviation (0. with cutoff reached the wall (640 ms timeout. The task itself was a 4. the effects session. The upper half-field presents 4. Twenty-four simultaneously dis- played images were arranged along four imagined concentric semi-circles (with six images/semi-circle). The common the canyon. with two values of orien- tation (0° and 90°) and two spatial frequencies (3 cycle/image and 7 cycle/image). Fig. The t-tests were blocks. In six ference potential was statistically significant. The radius of the smallest semi-circle was 16. ponents. channel matrices were de- the upper half. 576 standards and 128 deviants. Four variants of these Gábor patch-like images were constructed. Fig. 2004). The canyon was segmented by walls. The order of the blocks was counterbalanced fined as the regions of interest of the deviant-related com- across participants. positioned lateral to the outer canthi of the two eyes. visible throughout the block. there were three sets of four stimulus blocks each: in of physical differences were eliminated by this averaging and four blocks.g.2. The color-button assignment was randomized among extracted for each event and averaged separately for the stan- the blocks and indicated by two colored symbols on the left and dard and deviant stimuli. The reference electrode was on the right mastoid with the reference canyon (a tube with a rectangular cross-section) was sunk into recalculated (off-line) to linked mastoids. green and blue) were used.18). For lower-half field stimulation the matrix involved . the task-irrelevant stimuli appeared on the lower half applied at the location having the amplitude of the largest of the visual field and in the other six blocks.4.5. the spaceship was fixed in a stationary position amplifier. To inch LCD monitor (60-Hz refresh rate) in a dimly-lit. The task was to press the proper between electrodes placed above and below the right eye. identify change-related activities. Both values of epoch around a peak of the deviant minus the standard dif- orientation and frequency were standard (and deviant). another four blocks. the infrequent irrelevant stimuli had Deviance-related activity was considered to occur if the deviant frequency.. the surface of which was the electrode was attached to the forehead. each containing a was recorded with a bipolar configuration between electrodes colored door. Both orientations and both spatial frequencies were 704 stimuli. and in the remaining four blocks. both subtraction difference of the average amplitude of a 10-ms orientation and spatial frequency were deviant. the infrequent stimuli had deviant orientation. The stimulus duration was 80 ms. in subtracting procedure. Task-irrelevant stimuli were either were subtracted from ERPs from deviant stimuli of the respec- standards (p = 0. The task-related stimuli were displayed on the part of the screen opposite to the task-irrelevant stimuli.25). Procedure the standard and deviant stimuli from the six conditions (orientation deviant. inducing illusory movement of the spaceship.17-fold of image size). Vertical two of the colored doors were targets. which required no duration (including a 100-ms pre-stimulus interval) were response. phase (320°) and trim-value (0.6). the SOA was 560 ms and the background was dark. 2 measurements.2 m from a 17. spatial frequency deviant and double Participants were seated in a reclining chair 1. Leff. sound. The horizontal EOG horizon. In each block there tive condition. In the were presented as standards and deviants. frequencies of 0. 2 – An example of the display. 8 trials per target in a block). the aim of which was to maneuver a spaceship across a canyon. and a response button eye movements were monitored with a bipolar montage was assigned to each target. Epochs of 500 ms The third color indicated the non-target door. Synamps2 spaceship (i. BR A IN RE S E A RCH 1 3 98 ( 20 1 1 ) 6 4 –7 1 69 4. sampling rate 1000 Hz. The mean voltage during the 100-ms right sides of the screen.3. and epochs with an amplitude change exceed- shows a sample of the stimulus display. Thereafter.82) or deviants (p = 0.8°. button upon appearance of the target and before the spaceship The EEG signal was bandpass filtered offline. Three door colors (red. they appeared on difference potentials. and the ground the surface of a schematic planet. Event-related potentials (ERPs) were averaged separately for 4. deviant at upper and lower visual half-field stimulations). Task-irrelevant stimuli The irrelevant stimuli were grayscale images of Gaussian- windowed sinusoidal gratings. Measuring brain electrical activity simple video game. Task and task-related stimuli the task. ERPs were recorded for stimuli in the lower half-field. approximating the cortical magnification factor (e.1 and 30 Hz (24 dB slope). Both the radius of the semi-circles and the diameter of the images increased expo- nentially (with a quotient of 1. whereas the obstacles (which comprised the trodes placed at 61 locations according to the extended 10–20 target and non-target stimuli) moved toward the spaceship in system using an elastic electrode cap (EasyCap).. ERPs from standard stimuli attenuated chamber. therefore. NeuroScan recording system) with Ag/AgCl elec- on the screen). and the diameter of the smallest image was 0.8°.e. No target pre-stimulus interval was used as the baseline for amplitude appeared simultaneously with the task-irrelevant stimuli. The center of the animation was the EEG was recorded (DC-70 Hz.

data are presented as the Horváth. Memory-based detection of task-irrelevant visual changes. 364. Jeffreys. We investigated the possibility of Czigler. 1277–1282. In: Parasuraman..2011... Laterality. 11 data points) epochs of the dif. J. 88. Katayama.009. A. (with upper. 2003. Human visual system automatically encodes sequential regularities of discrete events. Varieties of were calculated by the Tukey-HSD test. Lett. Psychophysiology 43. Winkler. Vertical Kremláček... Orlando.. difference potentials detection: visual mismatch negativity to repeated visual stimuli. Position but not color deviants result in visual Leff. Acknowledgments Exp.. 129–140.. 97. Though there were no significant differences Neuropsychologia 48. Cogn. T. Balázs. between scalp distributions of the orientation and spatial Czigler. S.. Schröger. 293–321. Exp. Pató. two stimulus 151–155.N. 2009. 2007. Lett. J. field in primary visual cortex with special reference to the neural 702–707. T. Brain mechanisms of visual selective Greenhouse–Geisser correction was applied. J.. Jacobsen. Boston. 2010. I. Amenedo. 2007. I–II. comparison with the difference potentials elicited by the double 2006b.. activation.R. E.. (both early and late negativities. Peak latencies were Erlbaum.R. Probability-independent and -dependent ERPs reflecting visual change detection. I. Vision Res. Psychol. and compared by two-way (Half-field.. The mismatch negativity for changes in orientation — a sensory negativity and clinical neuroscience: theoretical and memory-dependent response. 2006. Visual asymmetries in pre-attentive detection of changes in motion mismatch negativity elicited by magnocellular system direction. P4. Hari. A.. ERPs and deviance frequency deviants in either half-field.. Hillsdale. In the reported effects the attention. Neurosci. 9.C. were summed (as the additive models of double deviants) for Czigler. Psychol... J. cognitive systems for pre-attentive visual deviance detection: an sLORETA analysis of the data of Kimura et al. M. Sulykos. Neuroreport 17. Schröger. Deviancy (orientation and spatial frequency). 1993. 2413–2418.. features.. Berti. I. 2007. 2319–2324. H. 28. 402–409.R. J. Neurosci. Ruusuvirta. PO3. F. J. Findings. Visual mismatch Kujala. L. Pazo-Alvarez. related activity. 1985). A historical review of the representation of the visual mismatch negativity in an active oddball task. Psychophysiol. E.G. 41–59. Pató.. 46. Psychophysiology 45. M. but we obtained no significant deviance- nonattended environmental regularities. 2006. I.. I. for upper-half field Clifford. Visual distraction: a behavioral and Levänen. Visual Appendix A. I. Detection of visual This research was supported by the National Research Found change: mismatch or rareness? Neuroreport 14. E. J. the Results 2.. methodological considerations... Davies..R. I. 74. measured at the maxima of the components. tests related to these exogenous components are described in Percept. Post-hoc analyses attention. Neuroreport 7. Brain Lang. 275–282. Visual mismatch negativity — violation of anterior effects... I. Cadaveira. Latency band stimulus omission. 2008. Visual temporal window of integration as revealed by deviant. Psychophysiology. 2002. Ohira. Source locations of pattern-specific components of human evoked potentials. 1–19. Schröger. T.. Kuba.. Schröger. Heslenfeld. E. Behavioral performance was mea- (Ed.. 1996. E. mean ±SD.. Lett. REFERENCES Kimura. Localizing sensory and doi:10.. A. In this analysis.. M.. E. I. 268–278.05.J.. L. M. Czigler. Winkler. mechanisms underlying macular sparing. Hillyard. E. Exp. J. Anteriority and Czigler. The Processing of Information and Structure. alpha level was at least 0. 1972. 180–189. M. P. 401.G.. W. Detection of Change: Event-Related Potential and fMRI sured as RT and HIT rates. 64. I.. J. Czigler. Preattentive auditory change detection stimulation). O1.A. Berti.. Tervaniemi. pp.2 and 2. Oz and O2. Visual change detection: ference potentials were compared by ANOVA. Garner..A. 224–230. values of the 10-ms (i.. L..e. the Harter. Wood. the ANOVA factors were Additivity the visual mismatch negativity event-related potential to (both modeled and observed double deviant). Visual mismatch negativity to analyzed on the vector-scaled amplitude values (McCarthy and irrelevant changes is sensitive to task-relevant changes. To compare deviant effects. Kenemans. J.. Anteriority and Laterality.. D. 1–19. J. Perform. Czigler. J. Supplementary data to this article can be found online at Kimura. 485. MMN or no MMN: no magnitude of deviance effect on the MMN amplitude.. Submitted Biol.. J. 1984. Responses of the event-related brain potential study in humans.L.. Brain Res. 2009. 1239–1242. human auditory cortex to changes in one vs. R. Aine. Pz.. 1–40. D.). PO4. Hum...1016/j. pp. A... M. 177–183. Brain Res. Neurosci. (2009). nel matrices. M. Neurosci... mean amplitude 21. C. I. Deviancy) ANOVA.. Brain Res. Weisz. Florida. Schröger. (Eds. Várnagy. Lillstrang.. L. We also investigated the peak latencies and scalp distribu.. Verbaten. CPz. while vector-scaled Hansen. . of Hungary (OTKA 71600). S. L. POz. 1124–1139. T... 2010a. Where applicable. Holmes... with the latter two factors calculated from the chan- 869–873. 2006a. J. Kimura. 2004. In: Polich. M. 16. 2003..J. 178–182. Eur. tions of the exogenous components.. ANOVA multidimensional auditory stimuli. In this comparison. Ohira.. L. but only for lower-half field Czigler. Winkler. categories affect pre-attentive color perception. Kluwer Academic Publishers. Balázs.). Schröger.brainres.2. McEvoy. Murohashi. I. 1983. S.... POz and PO4. I....05. Katayama. Int. P. Jong.2. 2006. Psychol. 60–69. Astikainen.70 BR A IN RE S EA RCH 1 3 98 ( 20 1 1 ) 6 4 –71 the locations PO3. P3. Davies. 1104. 1974. Sams. Maess. B. Neuroreport 20. Kubová. Winkler.. I..G.L. Color stimulation. 85.3 sections. E. A.... Weisz. H. Winkler. I. 2004. Selective attention to values were used to analyze distribution differences. Psychophysiology 39. 2008. Langrová. 485–490. 22. H. CP4.and event-related potentials are dependent on stimulus location in lower-half fields analyzed separately) using as factors of humans. distribution differences were Czigler. I. M. D.. 149–153. Balázs.. Visual mismatch negativity. Supplementary data mismatch negativity: new evidence from the equiprobable paradigm.. J. J. Z. relies on unitary sensory memory representations. the matrix involved the locations CP3. 2010. 198–203.. Psychophysiology 46. NJ.. R. Kimura. Franklin. 184–189. Neurosci. Academic Press. 2010b. Where appropriate. Axford. Ohira. M. Biol. S.

Kaneko. I.-J. Jacobsen. Winkler. Processing of auditory deviants with changes in deviance-elicited auditory deflection. Rinne. 109–112... Sarlo. Paavilainen. Neurosci.. 1995.. E. Percept. double. R. 5. 349.. 1999. 2001. Tapola. Lett. H. 1989. 2010. 1998. BR A IN RE S E A RCH 1 3 98 ( 20 1 1 ) 6 4 –7 1 71 May. P. 301. R... 19. 2004. 2001. R. Cognitive Brain Res. Psychophysiology 26. K. Winkler. Evidence for the different additivity characteristics of acoustic regularities: a mismatch negativity of temporal and frontal generators of mismatch negativity: a study.. Schröger. Lehtinen. 1996. Clin.. M. 25–32. A. P. change detection with single. McCarthy. Paavilainen.. Paavilainen. Winkler... H. R. Rare stimuli or rare changes: Schröger. Czigler. R. V.. Mulder.. Näätänen. I. Clin. a review... 311. evidence..J. 96. I. The mismatch and feature conjunctions in humans as indexed by the negativity (MMN) in basic research of central auditory processing: mismatch negativity. Yabe. 37–40.. E. 66–122. revealed by mismatch negativity additivity. 24D.. 165–171.. tromagnetic tomography (sLORETA): technical details.. Neuroreport 9. Mulder. explained. 62. Wolff. I. 2544–2590. Automatic 1998. A. Memory-based or 693–713. P.. Kilpeläinen. Scalp distribution of event-related Smid. Lett. R.G. The additivity of Event-related potentials during memory search and selective the auditory feature analysis in the human brain as indexed by attention to letter size and conjunction of letter size and color. E... afferent processes in mismatch negativity (MMN): a review of Takegata. 191–198. 1997. N. S. Neurophysiol. the mismatch negativity: 1 + 1 ≈ 2 but 1 + 1 + 1 < 3. H.. Wood.. T... 2005. E. R. I... Heinze. Winkler. Jakob. .. 495–499. Lett. or separated processing? Hear. Mazza. Näätänen. Takegata. Mismatch negativity: deviance Pascual-Marqui. 179–182. temporal. Mikkonen. 1985. and structural Saarela.. L. Czigler. Turatto. Interaural time and level difference: integrated what really matters for the brain? Neuroreport 16. Näätänen. Jaramillo.C. C. Human pre-attentive auditory Shinozaki. S. Psychophys. Näätänen.. Paavilainen. Methods 3809–3813.. Alho. Findings Experimental Clinical Pharmacology.. Psychophysiology 47. H. R.J. L. 79–82... 7. one versus two stimulus dimensions. M.. detection or the maintenance of the ‘standard’. T. evidence from ERPs. Independent processing of changes in auditory single features Näätänen. event-related brain potential study. 266. Res. The organization of potentials: an ambiguity associated with analysis of variance multidimensional selection on the basis of color and shape: an models. P. 2001. 2007. Paavilainen. Psychophysiology 38. 118. Winkler.. Neurosci. Tiitinen. Näätänen. Psychophysiology 32. Neurophysiol.. 2005. human event-related potential study.. M. Lett. R.... M. Temporal constraints of auditory event synthesis: detection of motion direction changes in the human brain. R. 203–208.. Standardized low-resolution brain elec. Wijers. Electroencephalogr. Neuroreport 9.C. T... 1998. 2002. 5–12. I. C.D. P. Hiruma. Okita. Sutoh.M. and triple deviations as Somatosensory automatic responses to deviant stimuli. F. Eur. Psychophysiology 42. M. G. Neurosci. J.. Paavilainen. Valppu. T. M. Neurosci.. P. T... G. P. Preattentive processing of spectral.. 529–547. P... 1061–1064. Pazo-Alvarez. Mismatch negativity (MMN). the Schröger.. 92–98. 55–65. Cadaveira. 2003. I... Neurosci.A.. Amenedo. 1978–1986.