Shaping the Visual Mismatch Response: The Influence of Exogenous Components and Stimulus Placement

ABSTRACT

This study investigated the relationship between the pattern‐specific event‐related potential (ERP) components (C1, C2, C3) and the component specific to automatic change detection in the visual modality (visual mismatch response, vMMR) and the influence of stimulus placement on the latter. These insights would help to understand whether the emergence of vMMR is due to the modulation of exogenous activity or a specific deviant‐related activity and advance the methodology of investigating vMMR. In two experiments, we presented checkerboard patterns in passive oddball sequences consisting of frequent (standard) and rare (deviant) events to examine vMMR as a function of the stimulated half‐fields. In Experiment 1, the lower and upper half‐field stimuli were presented within the same sequences, whereas in Experiment 2, they were presented in different sequences, completed with whole‐field stimulation. As expected, we observed polarity reversal of the C1 and C2 components as a function of the stimulated half‐field. Deviant stimuli elicited negative vMMR during lower and whole‐field stimulation, but upper half‐field stimulation evoked no significant vMMR. At lower half‐field stimulation, the C2 component was larger to deviant stimuli, indicating the contribution of exogenous components to the deviant‐standard difference. The sLORETA calculations showed that similar sources, including the primary visual cortex and other visual areas, were active during all exogenous and deviant‐related components. Altogether, these results demonstrate the relationship between the pattern‐specific and vMMR components and the more dominant influence of lower half‐field stimulation in the recorded brain activity during automatic change detection.

Short abstract

In our study of stimulating different parts of the visual field independently, results indicate that the common way of presenting vMMR‐related stimuli to the whole field is mainly the effect of deviance detected from the lower part of the visual field. We found that (1) early pattern‐specific ERP components (C1 and C2) emerge in all cases but with reversed polarity in upper half‐field stimulation. Moreover, (2) lower half‐field activity has a stronger contribution to both the pattern‐specific activity and vMMR.

1. Introduction

The current study aims to address two main goals. The first is theoretical: to clarify the interpretation of the visual mismatch response (vMMR), which can be understood in various ways. Specifically, the study explores whether stimulus‐specific adaptation (SSA) of pattern‐specific ERP components to standard stimuli fully explains the emergence of vMMR or beside the rareness‐related adaptation difference (Kenemans et al. 2003; Male et al. 2020), it reflects a ‘genuine’ deviant‐related additional, activity (e.g., Kimura et al. 2010). The second goal is methodological: to examine the optimal spatial placements of stimuli for eliciting vMMR, including the effects of lower half‐field versus upper half‐field stimulation and whether the putative opposite polarities of evoked components interfere with each other.

The signature of the automatic registration of visual events violating sequential regularities is the visual mismatch response (vMMR) component of event‐related potentials (ERPs). VMMR is usually investigated in the passive oddball paradigm. In this paradigm, the stimulus sequence consists of a frequent (standard) category, and this category is infrequently violated by an infrequent (deviant) category. VMMR is the difference between the ERPs elicited by the deviant and standard events. The paradigm is considered passive because the vMMR‐related events are not part of the ongoing task. As such, these vMMR‐related events are sometimes regarded as non‐attended (for reviews see Kimura et al. 2011; Stefanics et al. 2014). The present study aimed to investigate how vMMR is related to those exogenous ERP components which are elicited by both deviant and standard stimuli.

We applied checkerboard onset stimuli, because these stimuli elicit reliable pattern‐specific ERP components, first described by Aled Jeffreys (e.g., Jeffreys and Axford 1972). This and subsequent research identified three posterior components evoked by these checkerboard stimuli: C1, C2, and C3, with approximate latencies of 60–80, 100–120, and 140–160 ms, respectively. Traditionally, these components are interpreted as signatures of subsequent steps of processing visual stimuli: C1 is related to the early processing of stimuli and is especially sensitive to salient features like contrast; C2 is associated with the further integration and more detail‐oriented processing of incoming stimuli; and C3 might indicate a more complex processing related to cognitive functions such as attention or memory. An important characteristic of the C1 and C2 (but not C3) components (e.g., Aine et al. 1996; Capilla et al. 2016; Clark et al. 1994; Di Russo et al. 2002; Fortune et al. 2009) is that they reverse polarity as a function of the stimulated part of the visual field, that is, they are positive–negative at lower half‐field stimulation and negative–positive at upper half‐field stimulation. The deviant minus standard ERP difference, that is, vMMR, is supposedly closely related to the pattern‐specific components due to their similar latency, especially to C2. In that case, a similar polarity reversal is expected in the vMMR. However, the majority of vMMR studies have reported negative deviant minus standard difference potentials, and the more common terminology for the difference potential is visual mismatch negativity (vMMN).

Research so far investigating vMMR to lower and upper half‐field stimulation has not supported the polarity reversal of vMMR, that is, the connection between the exogenous components and the change‐related activity. Czigler et al. (2004) obtained reliable vMMR only to lower half‐field stimulation. Müller et al. (2012) found no reversed sensory (exogenous) ERP components to colored dots, and a negative difference potential appeared for both upper and lower half‐field stimulation, with the difference being larger for lower half‐field stimulation. In that study, at upper half‐field stimulation, the negative vMMR was also preceded by a positive difference potential. Amenedo et al. (2007) presented moving sinusoidal gratings to the lower or upper half‐fields, and lower half‐field stimuli elicited generally larger activity and larger negative vMMR amplitudes. Clifford et al. (2010) presented colored squares and reported negativity only to lower half‐field stimulation. Contrary to these results, Berti (2009) obtained negative vMMR only for upper half‐field stimulation in a study using deviant and standard locations. It is important to note that none of the previous studies investigated vMMR in response to stimuli directly related to pattern‐specific components (e.g., checkerboard patterns). However, these studies did not focus on the relationship between the pattern‐specific C2 polarity reversal and vMMR, which is the main topic of the present study. Thus, we hypothesize that the occurrence of a vMMR with negative polarity within the time range of an exogenous component with positive polarity (and vice versa) would support the notion of the existence of a “genuine” change‐related component. In contrast, a similar polarity reversal of the exogenous components and the vMMR suggests the need for further research into their relationship. This includes the possibility that the deviant‐minus‐standard ERP difference might be due to stimulus‐specific adaptation, where decreased activity occurs in response to the frequent (standard) event (e.g., Male 2024; Male et al. 2020).

Besides this theoretical issue, there is a methodological one, which is equally important. In the majority of vMMR studies using patterned stimulation, whole‐field stimuli have been presented. In the case of covert vMMR polarity reversal with overt negativity, the recorded effect can be biased, potentially leading to the recording of only the deviancy effect from the lower part of the stimuli. This study also addresses this methodological issue.

We conducted two experiments, both using checkerboard patterns as stimuli, with the difference between the deviant and standard being the size of the constituent squares. A previous study (Petro et al. 2021) reliably elicited vMMR using this deviancy. In Experiment 1, lower and upper half‐field checkerboard stimuli were presented within the same sequences. If larger squares were the standard in the lower half‐field, they were the deviant in the upper half‐field, and vice versa. To reduce the perceptual load (i.e., the need to simultaneously maintain two regularities; Csikós et al. 2023), Experiment 2 involved stimulating the lower and upper half‐fields in separate sequences. Additionally, Experiment 2 included whole‐field stimulation, allowing us to compare ERPs to half‐field and whole‐field stimuli. In this experiment, a denser electrode set enabled the application of the sLORETA source localization method, described in more detail in the Methods section. Using this method, we aimed to isolate the brain sources of the pattern‐specific C1 and C2 components, as well as the difference potentials.

2. Experiment 1

2.1. Methods

2.1.1. Participants

The participants were 21 paid students (17 female and 4 male, mean age: 21.95 years, SD = 1.77) with normal or corrected to normal vision (measured with Snellen cards). Due to abundant alpha activity (as a consequence of the oscillation no exogenous components were detected) or an extreme number of artifacts, the results of 3 participants out of the original 24 were discarded. The study was conducted in accordance with the Declaration of Helsinki and approved by the United Ethical Review Committee for Research in Psychology, Hungary (EPKEB). Written informed consent was obtained from all participants prior to the experimental procedure.

2.1.2. Stimuli and Procedure

Experiments were conducted in an electrically and sound‐isolated chamber. The stimuli were presented on a 24‐in. LCD monitor (Asus VS229na) with a 60 Hz refresh rate.

The stimuli were presented within a passive oddball paradigm, which meant that the participants were instructed to perform a reaction‐time task with a centrally located task‐relevant stimulus, while task‐irrelevant vMMR‐related stimuli were presented peripherally on the screen. The centrally located task‐relevant stimulus was a small fixation cross formed by unequal vertical and horizontal lines. The lengths of these lines periodically changed (randomly ranging between 5 and 15 ms, respectively) and the participants were instructed to indicate that change by pressing the SPACE bar, thus allowing us to measure their reaction time. Behavioral data consisted of the average reaction time (RT) and the percentage of correct detections of change. During this task, we presented the vMMR‐related stimuli to either the upper or the lower part of the screen.

Figure 1 shows the sketch of the stimulus display. The vMMR‐related stimuli were checkerboard patterns presented alternately either to the lower or the upper half‐field of the display. These checkerboard patterns either consisted of smaller or larger checks. During the task, these patterns created two independent sequences of stimuli, and the probability of smaller or larger check sizes varied in both sequences, as one appeared frequently (standard events) or rarely (deviant events). When the standard on one half‐field was the checkerboard with smaller checks, on the other half‐field it was the deviant, and vice versa. Accordingly, there were four conditions: lower half‐field—larger check size (Low‐L), lower half‐field—smaller check size (Low‐S), upper half‐field—larger check size (Up‐L), upper half‐field—smaller check size (Up‐S).

FIGURE 1.

An illustration of the sequence of task‐irrelevant stimuli. (A) The sequences of stimuli were presented in the upper and lower parts of the visual field. The standards on the upper side were the deviants on the lower side, and vice versa. Stimuli were presented for 100 ms. (B) The display of stimuli on the screen. The central fixation cross is the task‐relevant stimulus; the checkerboard patterns are the task‐irrelevant vMMR‐related stimuli.

The parameters of stimuli were the following: from the 160 cm viewing distance, the stimuli of the two half‐fields were separated by a horizontal area of 2°. Stimuli were presented to both sides of fields of 6.5^o (vertical) × 10^o (horizontal). The check size was either 0.4° (Smaller) or 0.8° (Larger). The lines of the central fixation cross were either 0.34° or 0.68°. The luminance of the gray background and the horizontal line was 54 cd/m², and the bright and dark checks had 114 and 15 cd/m² luminance, respectively. Stimulus duration was 100 ms. The inter‐stimulus interval was 459 ms with ±1, 2, or 3 jitters (where 1 jitter was 16.67 ms long). The stimuli appeared randomly within the sequences, but at least 2 standards were presented between two deviants. The probability of deviants was 15% within each sequence. Within a session, there were 6 blocks, each consisting of 400 task‐irrelevant stimuli, 200 in the upper and 200 in the lower half‐field sequences. The central fixation cross changed between 20 and 26 times in a single block.

2.1.3. Measurement of Electrical Brain Activity

Electrical brain activity was recorded at 27 locations according to the extended 10–20 system (F7, F3, FZ, F4, F8, FC3, FC4, T7, C3, CZ, C4, T8, CP5, CP6, P7, P3, PZ, P4, P8, PO7, PO3, POZ, PO4, PO8, O1, OZ, O2) with BrainVision Recorder (1.21.0303), ActiChamp amplifier, Ag/AgCl active electrodes (EasyCap, Brain Products GmbH), sampling rate: 1000 Hz, DC‐70 Hz online filtering. The reference electrode was placed on the nose tip, and the ground electrode on the forehead (AFz). Horizontal and vertical electrooculograms (HEOG and VEOG) were recorded with bipolar configurations between two electrodes (placed lateral to the outer canthi of the two eyes and above and below the left eye, respectively). The impedance of electrodes was below 10 kΩ. The EEG data were processed with MATLAB R2015a (MathWorks Inc. 2015) and the EEGLAB 13.6.5b toolbox (Delorme and Makeig 2004). The EEG signal was bandpass filtered offline with a non‐causal Kaiser‐windowed Finite Impulse Response filter (low pass filter parameters: 30 Hz of cutoff frequency, beta of 12.265, a transition bandwidth of 10 Hz; high pass filter parameters: 0.1 Hz of cutoff frequency). Stimulus onset was measured by a photodiode, providing an exact zero value for averaging. Epochs ranging from −100 to 400 ms relative to the onset of stimuli were extracted for further analysis, separately for small and large standards and deviants, and upper and lower half‐field stimuli. The 100 ms pre‐stimulus interval of each epoch served as the baseline. Independent component analysis (ICA) was used to remove eye‐movement artifacts. Epochs with larger than 100 μV voltage change at any electrode site were considered artifacts and rejected from further processing. We involved only those standards that were followed by a deviant into the averaging.

We calculated the amplitude and latency of the ERPs from the standard stimuli and collapsed the responses to the smaller and larger check sizes. The latency and amplitude values of exogenous components C1, C2, C3/P2 were calculated from the PO3, PO4, O2, and O1 electrodes. Table 1 shows the mean number of averaged epochs for each event. Peak amplitudes were measured as the ±10 ms range around the amplitude maxima in the ERPs. To compare the latency values of the exogenous components at upper and lower half‐field stimulation, we calculated parieto‐occipital (PO3, PO4) and occipital (O1, O2) ROIs for each component and conducted ANOVAs with factors of Half‐field (upper and lower) and ROI (PO, O). In post hoc comparisons, we applied the Tukey HSD calculation. Due to the reversed polarity of C1 and C2 components, comparison of amplitude values between the upper and lower half‐field stimulation was irrelevant.

TABLE 1.

Epochs averaged together for the lower half‐field and upper half‐field events with mean, S.E.M., and range values.

	Lower half‐field		Upper half‐field
	Standard	Deviant	Standard	Deviant
Mean	159.42	159.85	156.76	156.19
S.E.M	2.99	2.48	3.39	3.06
Range	136–183	136–181	128–176	126–176

VMMR was expected to emerge within the range of appr. 100–200 ms. We collapsed the ERPs to the smaller and larger check‐sizes again for both standard and deviant events separately for upper and lower half‐field stimuli so we could calculate ERP activity from physically identical stimuli by subtracting the standard events from the deviant ones in three 40 ms epochs within this range, that is, epochs of 90–130, 130–170, and 170–210 ms, calculated for parieto‐occipital (PO3, PO4) and occipital (O1, O2) locations, respectively. In order to test our theoretical hypothesis, as the first step of analyses, we calculated vMMR‐related ERP activity at the ±10 ms range around the amplitude maxima of the C1 and C2 exogenous components, so in the same range used to calculate exogenous components‐related activity. Then, as the second step of analyses, we calculated the vMMR‐related activity in the aforementioned three time windows. Analyzing deviant‐standard differences, we applied ANOVAs with factors of Stimulus (standard, deviant) and ROI (PO, O). The effect size was calculated as partial eta square (η _p ²). In post hoc comparisons, we applied the Bonferroni calculation. In these tests, the significance level was at least p < 0.05. Statistics were calculated by using the JASP package (Version 0.19.2).

2.2. Results

2.2.1. Behavioral Results

The average hit rate of the change of fixation was 95.31% (SD = 5.14); and it was similar in the conditions Up‐L—Low‐S and Up‐S—Low‐L, 96.61% (SD = 4.30) and 94.03% (SD = 9.35), respectively. The average reaction time was 462.79 ms (SD = 45.87); and it was similar in the conditions Up‐L—Low‐S and Up‐S—Low‐L, 463.31 ms (SD = 51.98) and 462.28 ms (SD = 44.46), respectively. There were no significant differences between the conditions, measured by t‐tests: t(38) = 1.11, p = 0.27; and t(40) = 0.06, p = 0.94 for the average hit rate and reaction time, respectively. Accordingly, we obtained no performance difference (The behavioral data of one participant was missing due to technical reasons, and hence it was not included in this analysis).

2.2.2. Event‐Related Potentials

2.2.2.1. Exogenous Components

Figure 2 shows the ERPs to the standard events of lower and upper half‐field stimuli, and the surface distribution of these components in the peak latency ranges. Lower half‐field stimuli elicited a positive component (C1), followed by a negativity (C2). The latency of the subsequent positivity was longer than the usual C3 range, so P2 is an interchangeable label for this component. Contrary to the lower half‐field stimulation, in upper half‐field stimulation, C1 was negative, C2 was positive, and these components were followed by a second positivity, C3/P2 that is obviously smaller for upper half‐field stimulation than for lower half‐field stimulation. Table 2 shows the amplitude and latency values of the exogenous components.

FIGURE 2.

Exogenous components and the surface distribution of the exogenous components C1, C2, C3/P2. (A) Exogenous components of Upper and Lower half‐field stimuli presented on the PO3, O1, PO4, O2 electrodes with 95% CI. Red: Standard lower half‐field stimuli; blue: Standard upper half‐field stimuli. (B) The surface distribution of exogenous components.

TABLE 2.

Amplitude (μV; upper part) and latency (ms; lower part) values of the exogenous components (C1, C2, C3/P2) at lower half‐field and Upper half‐field stimulation.

Stim	Lower half‐field				Upper half‐field
	Standard		Deviant		Standard		Deviant
ROI	PO	O	PO	O	PO	O	PO	O
Amplitude (μV)
C1	5.05 0.59 11.16	2.22 0.83 16.80	5.56 0.74 14.53	2.52 0.96 19.84	−1.56 0.54 9.24	−1.73 0.45 7.74	−1.86 0.61 10.70	−2.07 0.50 6.96
C2	0.32 0.89 17.18	−1.37 0.92 19.63	−0.67 1.00 20.11	−2.37 1.07 22.86	3.02 0.46 10.72	3.25 0.42 9.61	3.38 0.45 7.87	3.42 0.40 8.55
C3/P2	7.90 0.71 13.56	8.51 0.66 10.89	7.43 0.69 12.31	7.68 0.65 10.87	2.03 0.56 10.82	2.59 0.42 7.16	1.18 0.52 8.35	1.91 0.38 6.42
Latency (ms)
C1	65.24 1.40 30.00	61.67 4.61 118.00	65.04 1.46 31.00	66.00 5.46 103.00	52.81 3.55 63.00	56.86 3.06 67.00	52.76 2.86 53.00	55.23 2.90 50.00
C2	98.33 2.64 44.00	87.33 2.89 64.00	99.47 2.37 37.00	88.52 3.78 79.00	102.76 2.64 45.00	111.19 2.16 37.00	104.76 2.90 46.00	112.04 2.38 42.00
C3/P2	203.57 2.90 43.00	203.43 2.81 42.00	208.47 3.64 77.00	203.47 3.01 44.00	212.90 2.89 47.00	205.29 2.74 50.00	206.71 3.72 61.00	203.90 3.41 55.00

Note: Bold: mean, middle row: S.E.M, bottom row: range.

ANOVAs of the latency values with factors of Half‐field and ROI produced a significant main effect of Half‐field in C1, F(1,20) = 6, 10, p < 0.05, η _p ² = 0.23, showing the shorter latency for upper half‐field stimulation. For C2, the main effect of Half‐field and the interaction were significant, F(1,20) = 18.40, p < 0.001, η _p ² = 0.50 and F(1,20) = 25.71, p < 0.0001, η _p ² = 0.56, respectively. According to the post hoc Tukey HSD tests, the interaction was due to the lack of significant latency difference at the PO ROI between the upper and lower half‐field stimulation. At the O ROI, C2 latency was shorter at lower half‐field stimulation. For the C3/P2 component we obtained significant ROI main effect and interaction, F(1,20) = 6, 11, p < 0.05, η _p ² = 0.23 and F(1,20) = 6, 77, p < 0.05, η _p ² = 0.26, respectively. According to the post hoc Tukey HSD tests, there were no ROI differences at lower half‐field stimulation, but at upper half‐field stimulation longer latency appeared at the PO ROI.

2.2.2.2. Difference Potentials

Figure 3 shows the deviant minus standard difference potentials and the scalp distributions of the differences in the 90–130, 130–170, and 170–210 ms latency ranges, and Table 3 shows the amplitude values of the difference‐related potentials in this range.

FIGURE 3.

(A) Event‐related potentials to deviant and standard stimuli. (B) Difference potentials with 95% CI. (C) Surface distribution of difference potentials in the 90–130, 130–170 and 170–210 ms time ranges. In the difference potentials the 90–130, 130–170, 170–210 ms time ranges are indicated. Blue, standard events; Red, deviant events; Pink, difference‐related ERPs with 95% CI.

TABLE 3.

Amplitude (μV) values of the difference potentials in the 90–130, 130–170, 170–210 ms time ranges at Lower half‐field and Upper half‐field stimulation at the parieto‐occipital and occipital ROIs.

ROI	Lower half‐field				Upper half‐field
	PO		O		PO		O
Range	Standard	Deviant	Standard	Deviant	Standard	Deviant	Standard	Deviant
90–130	0.42 0.75 12.41	−0.71 0.71 12.86	−0.05 0.73 12.78	−1.13 0.74 13.34	2.48 0.46 10.13	3.04 0.46 7.81	2.79 0.37 8.73	3.14 0.36 7.86
130–170	3.87 0.65 12.79	3.25 0.57 11.42	4.45 0.63 11.86	3.86 0.58 11.88	−0.61 0.63 11.75	−0.36 0.59 9.22	0.26 0.58 12.52	0.44 0.52 9.43
170–210	7.27 0.77 15.27	7.07 0.72 13.18	7.69 0.77 14.12	7.37 0.73 12.89	1.41 0.59 8.70	1.44 0.50 6.66	2.14 0.44 6.64	2.22 0.38 6.06

Note: Bold: mean, middle row: S.E.M, bottom row: range.

2.2.2.2.1. Analyses in the Time Range of the C1 and C2 Exogenous Components

Here, we present only the results that showed significant effects of the Stimulus factor.

Lower half‐field stimulation: In the C1 range, the interaction was significant, F(1,20) = 6.06, p < 0.05, η _p ² = 0.23, but in the post hoc test, there were no significant differences on either ROI. For the C2 range, the main effect of Stimulus was significant, F(1,20) = 14.99, p < 0.001, η _p ² = 0.42, showing the negative difference to the deviant.

Upper half‐field stimulation: In the C1 range, there were no significant effects. For the C2 range, the interaction was significant, F(1,20) = 6.31, p < 0.05, η _p ² = 0.24, but according to the post hoc test, the difference was no longer significant on either ROI.

2.2.2.2.2. Analyses in the vMMR‐Related Three Time Windows

Lower half‐field stimulation: In the 90–130 ms time window, the main effect of Stimulus was significant, F(1,20) = 37.19, p < 0.001, η _p ² = 0.65, showing the negative difference to the deviant. For the 130–170 ms range, the main effect of Stimulus was also significant, F(1,20) = 7.30, p < 0.05, η _p ² = 0.26. For the 170–210 ms range, there were no significant effects.

Upper half‐field stimulation: In the 90–130 ms time window, the Stimulus × ROI interaction was significant F(1,20) = 9.51, p < 0.001, η _p ² = 0.32, but according to the post hoc test, the difference was no longer significant. For the 130–170 ms and 170–210 ms ranges, there were no significant effects.

2.3. Discussion

Event‐related potentials to checkerboard stimuli were fairly similar to those of previous studies: the C1 and C2 pattern‐specific components exhibited polarity reversal, being positive–negative for lower half‐field presentation and negative–positive for upper half‐field presentation (Jeffreys and Axford 1972; Di Russo et al. 2002; Czigler et al. 2004). However, these general findings can be further specified. We observed latency differences between upper and lower half‐field stimulation for both C1 and C2, suggesting that different brain areas may be involved in generating responses to half‐field stimulation.

The ERP signature of violated sequential regularity, known as the visual mismatch response (vMMR)—that is, the difference between the ERPs to deviant and standard stimuli of oddball sequences—was reliably observed at lower half‐field stimulation as a posterior component with negative polarity (visual mismatch negativity, vMMN) within the 90–170 ms latency range. At upper half‐field stimulation, even though the initial difference was positive, it no longer remained significant after the post hoc test. Most studies (Clifford et al. 2010; Lorenzo‐López et al. 2004; Müller et al. 2012) also reported a negative difference potential at upper half‐field stimulation. However, Czigler et al. (2004) only found an unreliable positivity, while Berti (2009) reported a negative difference, but only for upper half‐field stimulation. Thus, the present results—including the usual polarity changes of the C1 and C2 pattern‐specific components and a slight polarity reversal of difference potentials within a narrow latency range and area—may suggest the involvement of retinotopic structures in the deviant‐minus‐standard difference potentials and indicate that the polarity of exogenous components affects the emergence of the vMMR. Nevertheless, only lower half‐field stimulation elicited a significant vMMR. The explanation for this lower half‐field bias is not straightforward. Experiment 1 required simultaneous processing of two opposing regularities (with lower half‐field deviants being identical to upper half‐field standards, and vice versa). This increased cognitive load may have favored the dominance of processing one sequence over the other, in this case, the lower half‐field sequence.

In earlier work, we studied the processing of two interwoven oddball sequences. When using bilaterally presented geometric patterns (diamonds with their diagonals), we observed an interaction between the processing of the two deviant sequences (Csikós et al. 2023). However, with highly different stimuli, such as human faces and diamond patterns, the deviant‐related processes on each side were independent (Csikós et al. 2024). In the present study, both upper and lower half‐field deviants elicited vMMR‐like differences, with polarity corresponding to the C2 pattern‐specific component.

3. Experiment 2

In Experiment 2, we presented separate sequences for upper half‐field, lower half‐field, and whole‐field stimuli. For lower half‐field sequences, we expected the emergence of deviant‐related posterior negativity, similar to what was observed in Experiment 1. For upper half‐field stimulation, we aimed to determine whether the expected polarity reversal of the exogenous component (positive C2) was accompanied by similar positive deviant‐related activity. In the case of whole‐field stimulation, we investigated whether the deviant‐related activity resembled the combined effects of upper and lower half‐field stimulation or was dominated by the lower half‐field related negativity, which would be characteristic of a vMMN. Additionally, the use of a larger electrode array in this experiment allowed us to apply the sLORETA method to estimate the sources of ERP components and the difference potentials.

3.1. Methods

3.1.1. Participants

The participants were 22 paid healthy adults (16 female and 6 male, mean age: 22.00 years, SD = 3.47) with normal or corrected‐toto‐normal vision (measured with Snellen cards). Due to abundant alpha activity (as a consequence of the oscillation no exogenous components were detected) or an extreme number of artifacts, the results of 2 participants out of the original 24 were discarded. The study was conducted in accordance with the Declaration of Helsinki and approved by the United Ethical Review Committee for Research in Psychology, Hungary (EPKEB). Written informed consent was obtained from all participants prior to the experimental procedure.

To determine the sample size for Experiment 2, we used G*Power 3.9.1.6 (Faul et al. 2007). The calculations were based on the repeated‐measures ANOVA design with factors of Stimulus (standard, deviant) and ROI (PO, O) (for more details see the Methods section of Experiment 1). In the 90–130 ms range to detect the effect size of η _p ² = 0.65 for the main effect of Stimulus at 95% power and 0.05 alpha, a total sample of 5 participants is required. In the 130–170 ms range to detect the effect size of η _p ² = 0.26 for the main effect of Stimulus at 95% power and 0.05 alpha, a total sample of 12 participants is required.

3.1.2. Stimuli and Procedure

The details of this section correspond exactly to what was described in the same section for Experiment 1, except for the placement of vMMR‐related stimuli on the screen. While in Experiment 1, these task‐unrelated stimuli were presented at both the upper and lower half‐fields, in Experiment 2 there were three different versions. Here we presented the checkerboard pattern only to (1) the lower half‐field, (2) the upper half‐field, or (3) to the whole field. Because of that, there was only one sequence of events in each version, instead of having two independent, but interwoven sequences like in Experiment 1. In Experiment 2, each session included 9 blocks: 3 blocks for the 3 different types of half‐field conditions, presented in the following order: 3 blocks for the lower half‐field, 3 blocks for the upper half‐field, and 3 blocks for whole‐field stimulation. Each block contained 400 stimuli, resulting in 1200 stimuli per condition (600 for the “small square” and 600 for the “large square” sequences), of which 15% were deviants.

3.1.3. Measurement of Electrical Brain Activity

The details of this section also do not differ from that described in Experiment 1, except for the number of electrodes used. In this study, we recorded electrical brain activity from 58 locations according to the extended 10–20 system (FP1, FPZ, FP2, AF7, AF3, AF4, AF8, F7, F3, F1, FZ, F2, F4, F6, F8, FT7, FC5, FC3, FC1, FC2, FC4, FC6, FT8, T7, C5, C3, C1, CZ, C2, C4, C6, T8, TP7, CP5, CP3, CP1, CPZ, CP2, CP4, CP6, TP8, P7, P5, P3, P1, PZ, P2, P4, P6, P8, PO7, PO3, POZ, PO4, PO8, O1, OZ, O2) with BrainVision Recorder (1.21.0303), ActiChamp amplifier, Ag/AgCl active electrodes (EasyCap, Brain Products GmbH), sampling rate: 1000 Hz, DC‐70 Hz online filtering. The reference electrode was placed on the nose tip, and the ground electrode on the forehead (AFz). Horizontal and vertical electrooculograms (HEOG and VEOG) were recorded with bipolar configurations between two electrodes (placed lateral to the outer canthi of the two eyes and above and below the left eye, respectively). The process of averaging and calculating the latency and amplitude values for the exogenous components C1, C2, C3/P2, and vMMR was the same as before. Table 4 shows the mean number of averaged epochs for each event for each version of the task. To analyze the data, we, again, applied ANOVA analyses with the factors of Half‐field (lower, upper, whole) and ROI (PO, O) for exogenous components and Stimulus (standard, deviant) and ROI (PO, O) for the difference potentials.

TABLE 4.

Epochs averaged together for the lower, upper, and whole field events with mean, S.E.M., and the range values.

	Lower half‐field		Upper half‐field		Whole field
	Standard	Deviant	Standard	Deviant	Standard	Deviant
Mean	102.27	105.59	103.36	102.63	102.95	104.77
S.E.M.	1.82	1.34	1.66	1.54	1.44	1.23
Range	75–112	82–109	85–113	78–110	85–115	92–114

To localize the source of the exogenous components and the difference activity, the sLORETA distributed source localization analysis was applied. This analysis gives a solution for the EEG inverse problem by applying a weighted minimum norm estimation with spatial smoothing and standardization of the current density map. The forward model was generated on a realistic BEM head model (Gramfort et al. 2010) by applying a template MRI (ICBM152; 1 mm³ voxel resolution) with template electrode positions. The Q14 reconstructed dipoles (pA/m) were determined for every 15,002 sources in three orthogonal directions (unconstrained solution). Only brain regions were considered where the computed χ ²‐test value showed significance in at least 20 voxels and in at least one of the conditions. The sLORETA analysis was performed with Brainstorm (Tadel et al. 2016), which is documented and freely available for download online under the GNU general public license.

3.2. Results

3.2.1. Behavioral Results

The average hit rate of the change of fixation was 96.60% (SD = 3.49); and it was similar in the conditions: lower 97.25% (SD = 2.33), upper 97.31% (SD = 2.73), and whole field stimulation 95.24% (SD = 4.69), respectively. The average reaction time was 465.74 ms (SD = 72.59); and it was similar in the conditions: lower 463.04 ms (SD = 60.79), upper 467.34 ms (SD = 79.45), and whole field stimulation 466.85 ms (SD = 78.43), respectively. There were no significant differences between the conditions, measured ANOVA with the factor of Condition: F(1,21) = 3.05, p = 0.057, η _p ² = 0.12 for the average hit rate and F(1,21) = 0.79, p = 0.45, η _p ² = 0.03 for reaction time, respectively. Accordingly, we obtained no performance difference.

3.2.2. Event‐Related Potentials

3.2.2.1. Exogenous Components

Figure 4 shows the ERPs to the standard events of lower, upper, and whole field stimuli, and the surface distribution of these components in the peak latency ranges. Lower half‐field stimuli elicited a positive component (C1), followed by a negativity (C2), and then another positivity (C3). In upper half‐field stimulation, C1 was negative, C2 was positive, and these components were followed by a positive C3. When the whole field was stimulated, the latency and polarity of the exogenous components looked similar to those of lower half‐field stimulation. Table 5 shows the amplitude and latency values of the exogenous components.

FIGURE 4.

Exogenous components and the surface distribution of the exogenous components C1, C2, and C3. A: Exogenous components of Lower, Upper and Whole field stimuli presented on the PO3, O1, PO4, O2 electrodes with 95% CI. Red, standard lower half‐field stimuli; blue, standard upper half‐field stimuli. B: The surface distribution of exogenous components.

TABLE 5.

Amplitude (μV; upper part) and latency (ms; lower part) values of the exogenous components (C1, C2, C3/P2) at lower half‐field, upper half‐field, and whole field stimulation.

Stim	Lower half‐field				Upper half‐field				Whole field
	Standard		Deviant		Standard		Deviant		Standard		Deviant
ROI	PO	O	PO	O	PO	O	PO	O	PO	O	PO	O
Amplitude (μV)
C1	7.35 0.93 16.38	4.48 0.88 17.34	7.99 1.04 18.02	4.85 1.02 22.35	−1.87 0.62 11.49	−1.50 0.51 8.95	−1.92 0.78 12.90	−1.65 0.62 10.71	5.65 1.19 27.08	3.11 1.18 29.32	6.36 1.20 22.90	3.41 1.16 25.93
C2	1.36 0.74 14.21	0.48 0.85 14.43	0.34 0.82 11.94	−0.47 0.94 16.09	2.83 0.43 6.89	3.40 0.41 6.99	3.53 0.46 8.26	3.75 0.43 6.87	3.11 0.74 14.35	2.38 0.81 15.21	2.67 0.83 13.67	1.64 0.82 15.28
C3/P2	7.26 0.65 12.55	7.24 0.71 13.22	8.23 0.78 15.10	8.26 0.80 14.51	3.11 0.64 14.27	2.61 0.55 12.38	2.83 0.60 14.00	2.42 0.51 12.61	7.98 0.75 13.45	8.55 0.78 14.35	8.31 0.80 14.62	8.88 0.72 12.56
Latency (ms)
C1	71.45 1.38 32.00	67.68 3.36 84.00	72.36 1.00 16.00	69.27 2.91 72.00	61.63 3.15 62.00	65.50 3.36 78.00	58.18 4.25 89.00	63.18 3.81 69.00	81.45 3.06 55.00	85.86 5.43 89.00	78.00 2.42 56.00	83.68 6.55 98.00
C2	104.77 2.41 36.00	103.72 3.82 87.00	104.90 2.49 46.00	102.00 2.86 54.00	117.86 3.65 60.00	122.72 3.78 73.00	115.86 2.67 44.00	118.81 2.09 37.00	109.59 5.24 97.00	98.40 6.01 122.00	105.09 4.13 80.00	98.00 4.08 96.00
C3/P2	190.31 3.67 72.00	192.36 4.34 82.00	189.04 3.44 64.00	192.95 3.64 69.00	221.59 2.12 39.00	218.59 2.65 47.00	218.27 2.56 42.00	214.50 2.58 43.00	195.31 2.79 49.00	196.04 2.39 43.00	200.27 3.50 54.00	199.13 2.76 51.00

Note: Bold: mean, middle row: S.E.M, bottom row: range.

According to the ANOVAs on the latency values with factors of Half‐field and ROI: in C1, we obtained the significant main effect of Half‐field, F(1,21) = 11.01, p < 0.001, η _p ² = 0.34, showing the shorter latency for upper and lower half‐field stimulation compared to whole‐field stimulation. The post hoc Tukey HSD test showed that there were no significant differences between upper and lower half‐field stimulation. For C2, the main effect of Half‐field and the interaction were significant, F(1,21) = 5.45, p < 0.05, η _p ² = 0.20 and F(1,20) = 4.67, p < 0.05, η _p ² = 0.18, respectively. According to the post hoc Tukey HSD tests, the latency for upper half‐field stimulation was longer than for lower and whole‐field stimulation. Concerning the interaction, the latency for whole‐field stimulation was shorter at the O ROI. For the C3 component, we obtained a significant Half‐field main effect, F(1,21) = 27.51, p < 0.001, η _p ² = 0.56. According to the post hoc Tukey HSD tests, the latency for upper‐half field stimulation was shorter than for lower and whole‐field stimulation. Figure 5 shows the results of sLORETA calculations on the source of the exogenous components, and Table 6 shows the significant locations.

FIGURE 5.

sLORETA source localization for the C1, C2, C3 exogenous components in the lower, upper, and whole field versions. The darker pink color indicates the brain regions of increased deviant‐standard difference.

TABLE 6.

Brain regions in which the exogenous components, C1, C2, C3 were the largest in the lower, upper, and whole field versions. The digits show the number of significant voxels in the regions where χ ²‐test values are larger than mean plus two standard deviations of all χ ²‐test values. Average corrected p‐threshold: 3.33289 × 10⁻⁶ (Bonferroni, Ntests = 15,002). Only regions for which there were at least 20 voxels in at least one of the conditions are presented.

Half‐field	Lower						Upper						Whole
Ex.pot.	C1		C2		C3		C1		C2		C3		C1		C2		C3
Regions	L	R	L	R	L	R	L	R	L	R	L	R	L	R	L	R	L	R
Fusiform	20	106	26	26
Lat.occip.	94	102	110	84	77	25	130	30			112	43	192	98	66	24	97	67
Lingual	133	148	167	93	206	93	155	43	68	63	113	32	141	129	158	158	171	118
Perical.	48	78	97	85	109	91	103	37	52		85	54	64	82	85	84	92	107
Cuneus	24		32	37	65	54	31	27	32	37	24	46	28		70	71	67	58
Precun.		20	38		88				162	144		94		30	25	45
Inf.par.			22
Sup.par.					31				30	28			24		40	28	38	25
Isth.cing.									67	70	29					31

Abbreviations: cing., cingulate; inf., inferior; isth., isthmus; Lat., lateral; occip., occipital; par., parietal; perical., pericalcarine; precun., precuneus; sup., superior.

The sLORETA identified the activity of a large set of visual areas at the three pattern‐specific components, including the primary visual cortex (cuneus) and structures of the ventral stream (lateral occipital, lingual, and pericalcarine cortex). However, fusiform cortex activity was indicated only by lower half‐field stimulation. Structures of the dorsal stream (superior parietal cortex) and the isthmus of cingulate cortex appeared active only for upper half‐field and whole‐field stimulation.

3.2.3. Difference Potentials

Figure 6 shows the deviant minus standard difference potentials and the scalp distributions of the differences in the 90–130, 130–170, and 170–210 ms latency ranges. Table 7 shows the amplitude values of the standard and deviant stimuli in these ranges.

FIGURE 6.

(A) Event‐related potentials to deviant and standard stimuli. (B) Difference potentials. (C) Surface distribution of difference potentials in the 90–130, 130–170 and 170–210 ms time ranges. In the difference potentials the 90–130, 130–170, 170–210 ms time ranges are indicated. Blue, standard events; Red, deviant events; Pink, difference‐related ERPs with 95% CI.

TABLE 7.

Amplitude (μV) values of the ERPs in the 90–130, 130–170, 170–210 ms time ranges at lower, upper, and whole field stimulation at the parieto‐occipital and occipital ROIs.

ROI	Lower half‐field				Upper half‐field				Whole field
	PO		O		PO		O		PO		O
Range	Stan	Dev	Stan	Dev	Stan	Dev	Stan	Dev	Stan	Dev	Stan	Dev
90–130	1.78 0.52 9.07	0.54 0.64 9.51	1.59 0.58 11.68	0.50 0.73 12.31	2.04 0.34 6.37	2.55 0.38 8.06	2.36 0.34 5.72	2.55 0.35 7.02	3.26 0.45 9.44	2.61 0.59 9.12	3.29 0.57 11.07	2.39 0.63 10.80
130–170	3.86 0.69 10.72	3.46 0.78 12.93	4.80 0.61 10.93	4.42 0.71 11.93	0.02 0.60 10.56	−0.32 0.55 10.77	1.12 0.55 9.12	0.52 0.54 8.80	2.76 0.89 18.34	1.87 0.90 16.04	4.22 0.85 17.67	3.17 0.78 16.42
170–210	7.02 0.66 13.39	7.29 0.75 13.88	7.04 0.68 12.30	7.18 0.80 15.50	0.72 0.71 14.28	0.28 0.65 10.69	1.43 0.67 14.30	0.93 0.62 11.82	7.36 0.72 12.23	6.99 0.75 12.86	7.93 0.75 14.27	7.39 0.73 13.71

Note: Bold: mean, middle row: S.E.M, bottom row: range.

3.2.3.1. Analyses in the Time Range of the C1 and C2 Exogenous Components

Lower half‐field stimulation: In the C1 range the interaction was significant, F(1,21) = 8.45, p < 0.01, η _p ² = 0.28, but the post hoc test showed no significant differences on either ROI. For the C2 range the main effect of Stimulus was significant, F(1,21) = 21.70, p < 0.005, η _p ² = 0.32, showing the negative difference to the deviant.

Upper half‐field stimulation: In the C1 range there were no significant effects. For the C2 range the interaction was significant, F(1,21) = 6.62, p < 0.05, η _p ² = 0.23. According to the post hoc test, the difference was again no longer significant.

Whole field stimulation: In the C1 range, the interaction was significant, F(1,21) = 15.00, p < 0.001, η _p ² = 0.41, but not according to the post hoc test. For the C2 range, the interaction was significant, F(1,21) = 5.28, p < 0.05, η _p ² = 0.20, but again was no longer in the post hoc test.

3.2.3.2. Analyses in the vMMR‐Related Three Time Windows

Lower half‐field stimulation: In the 90–130 ms range, the main effect of Stimulus was significant, F(1,21) = 26.20, p < 0.001, η _p ² = 0.55, showing the negative difference to the deviant. For the 130–170 ms range, only the main effect of ROI was significant, F(1,21) = 5.29, p < 0.05, η _p ² = 0.20, showing the dominance of the O ROI. For the 170–210 ms range, there were no significant effects.

Regarding lower half‐field stimulation, we conducted an additional ANOVA for the 130–150 ms time range. As Figure 6 shows, the latency of the difference‐potential did not fit exactly to the a priori defined time ranges. We obtained significant main effects of Stimulus, F(1,21) = 5.397, p < 0.05, η _p ² = 0.22, and ROI, F(1,21) = 15.934, p < 0.001, η _p ² = 0.43. The ROI effect was due to the larger parieto‐occipital responses.

Upper half‐field stimulation: In the 90–130 ms range, the interaction of Stimulus × ROI was significant, F(1,21) = 10.39, p < 0.005, η _p ² = 0.33. According to the post hoc test, the difference was no longer significant. For the 130–170 ms range, the interaction was also significant, F(1,21) = 6.62, p < 0.05, η _p ² = 0.23, but again, there were no significant differences in the post hoc test. There were no significant effects in the 170–210 ms range.

Whole field stimulation: In the 90–130 ms range, the main effect of Stimulus was significant, F(1,21) = 6.76, p < 0.05, η _p ² = 0.24, showing the negative difference to the deviant. For the 130–170 ms range, the main effect of Stimulus was also significant, F(1,21) = 7.44, p < 0.05, η _p ² = 0.26. For the 170–210 ms range, there were no significant effects.

Figure 7 shows the source of difference potentials‐related activity according to the sLORETA calculations, and Table 8 shows the significant locations of the activity.

FIGURE 7.

sLORETA source localization for the difference‐related potentials in the upper, lower, and whole field versions in the 90–130 and 130–170 ms ranges. The darker pink color indicates the brain regions of increased deviant‐standard difference.

TABLE 8.

Activated brain regions of the difference potentials in the 90–130 and 130–170 ms ranges according to the sLORETA calculations at lower, upper, and whole field stimulation. The digits show the number of significant voxels in the regions where χ ²‐test values are larger than the mean plus two standard deviations of all χ ²‐test values. Average corrected p‐threshold: 3.33289 × 10–6 (Bonferroni, Ntests = 15,002). Only regions for which there were at least 20 voxels in the particular region are presented.

Half‐field	Lower		Upper				Whole
Time	90–130 ms		90–130 ms		130–170 ms		90–130 ms		130–170 ms
Regions	Left	Right	Left	Right	Left	Right	Left	Right	Left	Right
Cuneus	43	44					54	51	25	25
Inferior parietal	59	32	164		121
Latera loccipital	65						112	20	76	23
Lingual	24	34					52	20	59	26
Pericalcarine	91	72		22			89	57	72	20
Precuneus	38
Superior parietal	93	40	35		39	23	52		28
Caud.ant.cingulate									36
Superior frontal										36

Abbreviations: ant., anterior; caud., caudal.

According to the sLORETA results, in the 90–130 ms range, cuneus activity increased for lower half‐field deviancy, and for both ranges for whole‐field deviancy. Activities within the structures belonging to the ventral stream of visual areas increased similarly. Deviance‐related activity at upper half‐field stimulation, however, concentrated in the inferior parietal cortex, that is, in a structure belonging to the dorsal stream of visual areas. Increased activity of control structures (caudal anterior cortex and superior frontal cortex) appeared only at whole‐field stimulation.

As an extra set of statistical analyses, we added together ERPs from the lower and upper half‐field sequences and conducted ANOVAs with factors of Stimulus (deviant, standard) and ROI (PO, O). In the 90–130 ms range the main effect of Stimulus was significant F(1,21) = 7.87, p < 0.01, η _p ² = 0.27, showing the negative difference to the deviant. For the 130–170 ms range both the main effects of Stimulus and ROI were significant, F(1,21) = 6.06, p < 0.05, η _p ² = 0.22 and F(1,21) = 5.24, p < 0.05, η _p ² = 0.19, respectively, the latter showing a dominance for the O ROI. For the 170–210 ms range there were no significant effects. Then we compared the difference potentials from the whole‐field and the aforementioned combined (lower plus upper half‐field) activities in ANOVAs with factors of Condition (whole, combined) and ROI (PO, O). Based on these ANOVAs, there were no significant effects in any of the 90–130, 130–170, 170–210 ms time ranges.

3.3. Discussion

In Experiment 2, we presented stimuli to the lower and upper half‐fields in separate sequences and also applied whole‐field stimulation. The ERPs to the checkerboard stimuli were similar to those observed in Experiment 1, with the C1 and C2 pattern‐specific components reversing polarity: positive–negative for lower half‐field presentation and negative–positive for upper half‐field presentation. The contribution of the ERPs from the separate half‐field stimuli is evident in the ERPs from whole‐field stimulation (see Figure 6A). However, the absence of fusiform activity during whole‐field stimulation in the sLORETA analysis indicates a limit to the additivity of half‐field activities.

We found latency differences between the pattern‐specific components of lower and upper half‐field stimulation; however, these differences did not match those observed in Experiment 1. We do not provide an explanation for these discrepancies.

The vMMR was reliable at lower half‐field stimulation in the 90–130 ms range, but in the later range it was significant only at a 20 ms interval, between 130 and 150 ms. In contrast, for upper half‐field stimulation, the difference was positive in the earlier (90–130 ms) range and negative in the later (130–170 ms) range, but these differences were not reliable after the post hoc tests. For whole‐field stimulation, the vMMR results were negative in both latency ranges. Additional analyses indicated that combining ERPs from separate upper and lower half‐field stimulations did not produce reliable differences from the whole‐field responses.

Regarding the relationship between exogenous components and the difference potentials, when comparing the amplitudes of the exogenous components, at lower half‐field stimulation the C2 amplitude was larger for the deviant. Such a result does not rule out SSA‐related deviant‐standard differences. However, at lower half‐field stimulation, the difference was negative also in a later latency range (130–150 ms), that is, later than the peak C2 region, and at whole field stimulation, it was negative in the 130–170 ms range, that is, similarly later than the C2 peak. At upper half‐field stimulation, we obtained no significant differences; the larger, positive C2 peak amplitude to the deviant was insignificant, similarly to the negativity in the 130–170 ms range. Accordingly, rareness had effects on the amplitudes of the C2 pattern‐specific components, but following the amplitude maximum, the negativity persisted at lower‐ and whole‐field stimulation.

In summary, both hemifields contributed to the ERP activities when the whole visual field was stimulated. However, when calculating difference potentials (i.e., vMMR), the results were predominantly influenced by the lower half‐field activity.

The sLORETA results on the difference potentials indicated that, compared to the average activity of the whole cortex, lower and upper half‐field stimulations elicited different activity patterns. The most obvious difference was the involvement of the cuneus in lower half‐field stimulation and the distinct contributions of the ventral and dorsal structures. In the case of whole‐field stimulation, the sources in the 90–130 ms range were similar to those observed for the lower half‐field, while in the 130–170 ms range, the set of structures also included anterior areas. It is important to emphasize that the level of whole‐brain activity influences the sLORETA results, making these findings less suitable for discussing the additivity of the lower and upper half‐field activities. A notable aspect of the sLORETA results is that the deviant vs. standard difference involves a broad set of visual structures.

4. General Discussion

In this study, we investigated ERPs in response to checkerboard patterns presented to the lower and upper halves of the visual field, as well as to the whole field. Two methods were applied: in Experiment 1, lower and upper half‐field stimuli were delivered within the same sequence, while in Experiment 2, they were presented in separate sequences. Additionally, whole‐field stimuli were only presented in separate sequences in Experiment 2.

The results show that both the lower and upper half‐field stimulation elicited the pattern‐specific C1 and C2 components, similarly to findings in studies by Aine et al. (1996), Clark et al. (1994), Fortune et al. (2009). The third pattern‐specific component (C3) was inseparable from the P2 exogenous ERP component. As expected, C1 and C2 reversed polarity: at lower half‐field stimulation, we recorded positive–negative components, whereas at upper half‐field stimulation, we recorded positive–negative components. We also observed peak latency differences as a function of the stimulated half‐fields, but these results were ambiguous. Whole‐field stimulation was similar to the combined activity of ERPs to the lower and upper half‐fields. However, in the C1 and C2 range, the whole‐field ERP was dominated by the larger lower half‐field activity.

One of the main purposes of the study was to examine the possible causal relationship between the above‐described pattern‐specific components and the difference potential between standard and deviant stimuli in oddball sequences. The stimuli were task‐irrelevant, so the expected difference was identified as the signature of automatic detection of violated sequential regularity, that is, the visual mismatch response (vMMR). The specific issue was whether vMMR to half‐field stimulation followed the polarity change of the pattern‐specific ERPs. Table 9 shows the amplitude differences between the standards and deviants of the C1 and C2 pattern‐specific components and the differences in the three latency ranges for the putative emergence of the vMMR component. The results are not “silver bullets” against the possibility of the involvement of the SSA of pattern‐specific components in the deviant‐standard difference. This is because we obtained larger negative peak amplitudes of the deviant‐related C2 in both experiments at lower half‐field stimulation. However, in Experiment 1 at lower half‐field stimulation negativity emerged also in the 130–170 ms range, that is, later than the C2 peak. This result was supported by the negativity in the 130–150 ms range in Experiment 2, and we also obtained negativity in the 130–170 ms range at whole field stimulation. Concerning upper half‐field stimulation, we obtained no reliable differences, neither in the pattern‐specific components nor in the investigated latency ranges. These later latency ranges for the difference potentials were unrelated to the ranges of pattern‐specific components.

TABLE 9.

Amplitude (μV) values of the difference‐related vMMR component in both Experiments in the time ranges of the exogenous components C1 and C2, and in the 90–130, 130–170, 170–210 ms time ranges at lower, upper, and whole field stimulation at the parieto‐occipital and occipital ROIs.

	Condition	Lower half‐field		Upper half‐field		Whole field
	ROI	PO	O	PO	O	PO	O
Experiment 1	C1	0.50	0.30	−0.30	−0.33
	C2	−0.99	−0.99	0.35	0.17
	90–130	−1.13	−1.07	0.56	0.35
	130–170	−0.61	−0.59	0.24	0.18
	170–210	−0.20	−0.31	0.02	0.05
Experiment 2	C1	0.64	0.36	−0.04	−0.14	0.70	0.30
	C2	−1.02	−0.95	0.69	0.35	−0.43	−0.74
	90–130	−1.19	−1.08	0.50	0.19	−0.65	−0.89
	130–170	−0.39	−0.38	−0.35	−0.60	−0.88	−1.05
	130–150	−0.73	−0.67
	170–210	0.26	0.14	−0.43	−0.50	−0.37	−0.54

Note: Significant differences (p < 0.05) are presented by bold numbers.

A central issue in vMMR research is whether the amplitude difference between the deviant and standard stimuli is due to additional mismatch‐related activity or it can be explained as the larger repetition‐related stimulus‐specific adaptation (SSA) on the standard stimuli (e.g., May and Tiitinen 2010). Apparently, in the visual modality, this view is supported by the results of studies applying equal probability control (File et al. 2017; Male et al. 2020; Jacobsen and Schröger 2001). In this control procedure, the ERP to oddball deviant is compared to the ERP to the physically same stimuli from sequences of various stimuli. The probability of each stimulus type is equal to the probability of oddball deviant. In case of no difference, the deviant minus standard oddball difference is attributed to the effect on the standard. Importantly, this control procedure cannot provide direct data on diminished repetition‐related amplitude decrement. In a recent study, Kővári et al. (2024) investigated the possibility of amplitude change following oddball deviants but obtained no amplitude decrement of exogenous components, that is, SSA. To explain the amplitude difference of the exogenous components, it is possible to speculate on the opposite possibility, that is, increased amplitude of the deviant stimuli. In fact, at the field of visual (e.g., Anllo‐Vento and Hillyard 1996) and auditory attention (e.g., Woldorff and Hillyard 1991), such modulation is a well‐documented finding. Similar modulation is possible also at task‐irrelevant events.

The other aspect of the putative deviant‐related activity is the deviant minus standard difference outside the C2 range. As Table 9 indicates, at lower half‐field stimulation in both Experiments and at whole‐field stimulation in Experiment 2, significant negativity emerged later then this exogenous component. C2 peak latencies were similar to the standard and to the deviant. Therefore, C2 latency shift does not explain this effect, that is, these results are at least a “shooting with slingshot” weight of argument in the direction of the emergence of a “genuine” mismatch‐related activity. However, as another possibility to consider, this difference might have emerged due to a “silent” exogenous activity, masked by the C3/P2 component.

As mentioned in the Introduction, the present results are relevant to a critical issue in vMMR research: how the activity elicited by the lower and upper half‐fields contributes to the vMMR in whole‐field stimulation. Based on our current results, the answer is unequivocal: whole‐field vMMR is dominated by the effect of lower half‐field stimulation. One might have assumed that the simultaneous presentation of stimuli to both the lower and upper parts of the visual field would result in a deflated vMMR at whole‐field stimulation, as the putative reversed polarities of the deviant‐related ERPs would cancel each other out. However, this was not the case, suggesting that presenting vMMR‐related stimuli to the whole field is an adequate practice, but utilizing only the lower half‐field can also be sufficient.

Studies investigating lower levels of visual deviancy, such as color (Winkler et al. 2005), spatial frequency (Heslenfeld 2003), size of elements (Petro et al. 2021), movement direction (Lorenzo‐López et al. 2004), orientation (Kimura et al. 2010) usually apply stimuli with identical lower and upper parts. The response to the basic questions of these studies did not depend on the different contributions of the lower and upper half‐fields. However, the contribution of the lower and upper half‐fields is particularly interesting in the case of the most popular stimulus type of vMMR research: studies with emotions on the human face (e.g., Astikainen and Heitanen 2009; Kreegipuu et al. 2013; Li et al. 2012). In most studies, faces were presented at the center of the screen or two laterally presented faces on the horizontal midline. It remains an open question how the lower part (mouth) and the upper part (eyes) contribute to the difference potentials.

As the sLORETA results show, our findings contrast with the view developed by Jeffreys and Axford (1972) based on the interpretation of the polarity change of the pattern‐specific components, which was widely accepted later (e.g., Di Russo et al. 2002). According to this view, the subsequent components (C1, C2, and C3) are indices of a processing hierarchy within the visual system, indicating the subsequent activation of the striate and extrastriate cortices. However, the present results show that extrastriate areas were active as early as 70 ms, while the striate cortex (cuneus) maintained activity even in the range of the C3/P2 component (at least until 210 ms). This suggests a not strictly sequential order of processing visual stimuli. The long‐lasting activity of the primary visual areas can be considered as feedback loops within the visual areas (e.g., Budd 1998).

As expected, the dominant source of the pattern‐specific components is within the ventral stream of the visual system, that is, the occipital and occipito‐temporal cortex. Interestingly, we isolated the involvement of a parietal structure at upper half‐field stimulation. While it is possible that the lower half‐field dominance in ERPs is simply the consequence of the preferential availability of retinotopic structures processing lower half‐field stimuli, a functional interpretation is also possible. Previc (1990) attributed peripersonal processing to the lower and extrapersonal processing to the upper half‐field, but this dichotomy does not explain the ERP dominance of pattern‐specific components at lower half‐field presentation. Nevertheless, lower half‐field advantage appeared in various studies, such as direction discrimination, motion processing, temporal resolution, perception of illusory contours, and color matching (for review see Thomas and Elias 2011).

Comparing the sLORETA results of the pattern‐specific C2 component and the deviant‐related vMMR, we observed activity in both the primary visual cortex and areas belonging to the ventral stream of visual processing. This finding differs from the data reported by Kimura et al. (2010), where vMMR‐related activity was absent in the primary visual cortex, and they reported the involvement of frontal structures in vMMR. Instead, we observed superior frontal activity in vMMR during whole‐field stimulation. Interestingly, the superior parietal cortex, part of the ventral stream of the visual system, appeared active in both the C2 range and the difference potentials, indicating the complexity of processing the checkerboard patterns and detecting deviancy.

However, there are limitations to this study. We applied the onset of high‐contrast checkerboard patterns and check size differences as the basis for deviancy, and it remains an open question whether other visual features, such as color or orientation, would elicit similar results. Additionally, the sensitivity of the sLORETA method used in this study is less precise compared to other imaging techniques like fMRI for localizing brain activity.

In conclusion, we recorded the polarity reversal of the pattern‐specific ERP components. The emergence of the C1 component, followed by the C2, did not indicate the successive activation of visual structures; on the contrary, similar posterior structures were active in the latency ranges of both components. Deviant stimuli within the oddball sequences elicited larger activity than the standard stimuli (vMMR), and this difference cannot be attributed solely to the repetition‐related stimulus‐specific adaptation of any of the pattern‐specific components. However, the amplitude changes of these components (either SSA‐related amplitude decrease to the standard, or amplitude increase of the deviant) may contribute to the ERP difference between the rare and frequent stimuli of oddball sequences.

Author Contributions

Nóra Csikós: data curation, formal analysis, investigation, visualization, writing – review and editing. Bela Petro: methodology, software. Zsófia Anna Gaál: conceptualization, methodology, project administration, supervision, writing – review and editing. István Czigler: conceptualization, funding acquisition, methodology, supervision, writing – original draft, writing – review and editing.

Conflicts of Interest

The authors declare no conflicts of interest.

Csikós, N. , Petro B., Gaál Z. A., and Czigler I.. 2025. “Shaping the Visual Mismatch Response: The Influence of Exogenous Components and Stimulus Placement.” Psychophysiology 62, no. 7: e70111. 10.1111/psyp.70111.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.