- Email: [email protected]
The visual mismatch negativity (vMMN): Toward the optimal paradigm
International Journal of Psychophysiology 93 (2014) 311–315
Contents lists available at ScienceDirect
International Journal of Psychophysiology journal homepage: www.elsevier.com/locate/ijpsycho
Short communication
The visual mismatch negativity (vMMN): Toward the optimal paradigm Xiaosen Qian a, Yi Liu a,⁎, Bing Xiao a, Li Gao a, Songlin Li a, Lijie Dang b, Cuiping Si c, Lun Zhao b,⁎⁎ a b c
Sleep Disorders Center, Civil Aviation General Hospital, Beijing, China School of Psychological Research, Beijing Yiran Sunny Technology Co. Ltd., Beijing, China Department of Neurology, Jining No. 1 People's Hospital, Jining, China
a r t i c l e
i n f o
Article history: Received 5 February 2014 Received in revised form 4 June 2014 Accepted 6 June 2014 Available online 14 June 2014 Keywords: Visual mismatch negativity Visual discrimination Change detection Optimal paradigm
a b s t r a c t In the present article, we tested an optimal vMMN paradigm allowing one to obtain vMMNs for several visual attributes in a short time. vMMN responses to changes in color, duration, orientation, shape, and size were compared between the traditional ‘oddball’ paradigm (a single type of visual change in each sequence) and the optimal paradigm in which all the 5 types of changes appeared within the same sequence. The vMMNs obtained in the optimal paradigm were equal or larger in amplitude to those in the traditional vMMN paradigm. The optimal paradigm can provide 5 different vMMNs in the same time in which usually only one MMN is obtained. This short objective measure could putatively be used as an index for visual cognition function especially in clinical research. © 2014 Elsevier B.V. All rights reserved.
1. Introduction As an important cognitive function for human survival, the automatic change detection occurs at a very early stage of information processing. To date, ample evidence indicated that the mismatch negativity (MMN) of event-related potential (ERP) components is a reliable indicator for exploring automatic change detection, reflecting the brain's detection of an unintentional disruption in the regularity of temporal events (e.g., Näätänen et al., 2007). Generally, MMN can be elicited by infrequent deviant stimuli inserted randomly in a sequence of frequent standard stimuli presented outside of the focus of attention. Although the MMN component has been widely investigated in auditory modality, analog of auditory MMN was also found in response to visual deviants such as color (Czigler et al., 2002), orientation (e.g., Kimura et al., 2009), size (Kimura et al., 2009), shape (Grimm et al., 2009), duration (Qiu et al., 2011) as well as facial expressions (e.g., Zhao and Li, 2006). Although visual MMN (vMMN) is elicited by task-irrelevant events, recent studies have shown that vMMN is influenced by attentional manipulations (e.g., Czigler, 2007; Czigler and Sulykos, 2010; for a review, Kimura, 2012). Therefore, instead of being an index of fully automatic processes, it was proposed that vMMN reflects the existence of unintentional temporal-context-based prediction (Kimura, 2012). As an inexpensive and non-invasive method, it is not surprising that the vMMN has recently received considerable attention as a tool of
⁎ Correspondence to: Y. Liu, Sleep Disorders Center, Civil Aviation General Hospital, Beijing 100123, China. ⁎⁎ Correspondence to: L. Zhao, School of Psychological Research, Beijing Yiran Sunny Technology Co. Ltd, Beijing 100088, China. E-mail addresses: [email protected] (Y. Liu), [email protected] (L. Zhao).
http://dx.doi.org/10.1016/j.ijpsycho.2014.06.004 0167-8760/© 2014 Elsevier B.V. All rights reserved.
clinical research. For instance, vMMN in response to the deviant duration was significantly reduced in depression patients (Qiu et al., 2011) as well as in Alzheimer's disease (Tales and Butler, 2006). It should be noted that different kinds of deviant features elicit distinctive vMMNs, (maybe) with different cortical origins. For instance, the orientation MMN was generated at the medial prefrontal areas as well as at the visual extrastriate area (Kimura et al., 2009), whereas the neural generator of vMMN elicited by directions of motion included the occipital visual extrastriate areas (including motion response areas) (PazoAlvarez et al., 2004). Previous vMMN clinical studies tested only a single deviant type and hence, it is not clear whether vMMNs elicited by other deviant types are affected by the disease. Therefore, it is necessary for developing a multi-feature vMMN paradigm with different deviant types integrated in one complex visual context. Indeed, the practical problem, especially in clinical application, is that traditional vMMN paradigms are usually time-consuming when several types of deviants are presented in different stimulus blocks. Näätänen et al. (2004) developed a new multi-feature paradigm in which five types of acoustic changes (frequency, intensity, duration, perceived sound-source location, and gap) were presented in the same sound sequence. Indeed, in this paradigm different auditory attributes can elicit different MMNs, and most importantly, the memory trace of the standard stimuli can enhance with respect to the stimulus attributes they have in common. Although the multi-feature paradigm reduced the time of the experiment, the MMNs using the new paradigm did not differ from those using the traditional oddball-paradigm. To date, only few studies investigated vMMN using the multi-deviant paradigm. For example, Grimm et al. (2009) found that the vMMN depended on the changed feature, which might reflect either the differences in saliency between the feature changes or the natural hierarchy
312
X. Qian et al. / International Journal of Psychophysiology 93 (2014) 311–315
in processing of various visual stimulus dimensions. Kreegipuu et al. (2013) compared vMMNs elicited by angry and happy schematic faces in a traditional oddball design (deviant stimuli, 12.5%) and in an optimal multi-feature paradigm with several deviant stimuli (altogether 37.5%) and found that the expressional MMN (EMMN) was equally elicited in both paradigms. They considered that in order to save time and other experimental resources the use of the optimum (multi-feature) design for EMMN should be encouraged (Kreegipuu et al., 2013). However, in their study, the presentation probability for a single deviant in the oddball paradigm was about twice as high as in the optimum paradigm, that is, the refractory state of same event-related potential components was different between two paradigms (Kreegipuu et al., 2013). On the basis of the auditory MMN optimal paradigm by Näätänen et al (2004), we recently proposed a time-saving and multi-feature visual MMN paradigm, in which 5 different vMMNs could be recorded with lowprobability deviant stimuli (color, duration, orientation, shape, and size) inserted randomly in a sequence of frequent standard stimuli (Shi et al., 2013). However, in order to successfully address the reliability of multi-feature visual MMN paradigm, it is necessary to show that in fact the vMMNs recorded from the multi-feature deviant paradigm are identical or not smaller than to those recorded in the same subjects using traditional single-feature deviants. This possibility will be investigated in the present study.
2. Method 2.1. Participants Fourteen undergraduate volunteers (8 females, all right-handed) with normal or corrected-to-normal visual acuity served as participants (the average age = 22.3 years, age range = 19–25 years). All participants received payments for their contribution. This research was approved by the Ethical Committee of Civil Aviation General Hospital in
accordance with the Declaration of Helsinki and all participants gave their written informed consents before the experiment. 2.2. Stimuli and procedure The subject sat on a comfortable chair in a darkened, sound attenuated and electrically shielded room. In the center of screen, a black cross was displayed throughout the stimulus blocks. From time to time, the cross became bigger or smaller unpredictably. Subjects were instructed to ignore the peripheral stimuli and press the left or the right button as quickly and accurately as possible when the size of the cross became bigger or smaller. The number of targets was 18–22 (mean 20) in each sequence. In the peripheral sides of the field, two identical visual stimuli were simultaneously presented from a distance of 1 m, with the stimulus onset asynchrony (SOA) of 600 ms. The concurrently presented peripheral stimuli occurred on the left and right of the cross (4.5° distance from center of cross to center of the peripheral stimulus). The solid red rectangles (30 mm in length and 10 mm in width) served as standard with duration of 50 ms. As presented in Fig. 1, five types of deviant stimuli were included: two solid blue rectangles (color deviant), two shifted 90° solid red rectangles to the standard stimuli (orientation deviant), two red rectangles with 100 ms duration (duration deviant), two solid red ellipse (shape deviant) and two wide red rectangles (size deviant). There were 2 different conditions in the present study (Fig. 1): the traditional oddball paradigm (a single type of deviant in a sequence) and another condition (denoted as ‘Optimum’), in which all 5 types of deviants occurred within the same sequence (Näätänen et al., 2004). In line with the method proposed by Näätänen et al (2004), in the Optimum condition, all 5 deviants (p = 10% for each; total 50%) were presented in the same sequence so that every other visual stimulus was a standard stimulus (p = 50%). The rationale of this paradigm was that the other deviants can strengthen the memory trace of the standard with respect to those levels of stimulus attributes they had in
Fig. 1. Top: Examples for stimulus configurations: a) standard display, b) color deviant display, c) orientation deviant display, d) duration deviant (duration: 100 ms) display, e) shape deviant display, f) size deviant display. Down: Schematic illustration of the 2 stimulus conditions used: traditional Oddball and Optimum conditions. S denotes standard stimulus and DX stimuli of different deviant types (D1 — color deviant, D2 — orientation deviant, D3 — duration deviant, D4 — shape deviant, D5 — size deviant). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
X. Qian et al. / International Journal of Psychophysiology 93 (2014) 311–315
common. The deviants were presented in an array of 5 deviants (Fig. 1). The stimuli were presented in five 3 min sequences (1500 stimuli in total; 150 for each deviant stimulus and 750 for standard stimulus), with the total recording time for the 5 types of deviants thus being 15 min. The stimulus sequences of the traditional Oddball condition were created using those of Optimum condition by replacing with standards all deviants except for those of one category. In this way, the times of the occurrence of a certain type of deviant stimulus in the sequence were identical to those of Optimum condition. Stimuli were presented in five 3 min sequences for each deviant type, resulting in a total recording time of 75 min. In sum, Optimum condition was composed of 5 stimulus sequences, whereas in Oddball condition, there were 25 sequences, 5 sequences for each deviant type. The order of conditions as well as the order of sequences in different conditions was balanced across subjects. 2.3. EEG recording The electroencephalogram (EEG) was continuously recorded with ANT asaLab Amplifier (www.ant-neuro.com), using an electrode cap with 32-channel Ag/AgCl electrodes according to the extended international 10–20 system. The reference electrode was placed on the nose tip. Vertical EOG was recorded with two electrodes above and below the right eye. Horizontal EOG was recorded with two electrodes at the right and left outer canthi of the eyes. The impedances of the electrodes were kept below 5 kΩ throughout the experiment. EEG and EOG signals were amplified with a band pass of 0.1–100 Hz at a sampling rate of 500 Hz. After EOG artifact correction (Semlitsch, et al., 1986), the EEG was segmented into the epoch from 100 ms pre-stimulus (as the baseline) to 500 ms post-stimulus. The trails contaminated with artifacts greater than ± 100 μV were rejected before averaging. The trails with targets when participants' responses occurred and where participants' responses did not occur (i.e., missed targets) as well as the trials without targets
313
where participants' false positive responses occurred were rejected from averaging. The EEG segments were averaged separately for standard and deviant (95.5 ± 8.6, 101.3 ± 10.0, 98.7 ± 9.9, 97.7 ± 11.2, and 99.9 ± 10.1 for color, duration, orientation, size, and shape deviants, respectively) stimuli in different conditions. MMN was obtained by subtracting ERPs to standard stimuli from ERPs to deviant stimuli for each visual feature, respectively. Since in many participants the vMMN peaks were not easily discernible at all sites and in each condition (see Figs. 2 and 3), the mean amplitudes of vMMN were measured for the 150–250 ms time windows post-stimulus onset at the occipital–temporal electrode sites, left occipital–temporal electrode sites (LOT) including T5 and O1, right occipital–temporal electrode sites (ROT) including T6 and O2, and midline occipital site (Oz), respectively. One-tailed t tests were conducted to determine whether the vMMN mean amplitudes significantly differed from zero. Three-way analyses of variance (ANOVA) with repeated measures were conducted to test the effects of Condition (2 levels: Optimum and Oddball), Deviant type (color, duration, orientation, shape, and size) and Site (LOT, Oz, ROT). Greenhouse–Geisser corrections were made when appropriate. Newman–Keuls tests were carried out as post hoc analyses. 3. Results Responses for detecting cross changes in the center of the screen were scored as hit if the correct button was pressed within 150 to 1000 ms after targets onset. Mean accuracy rate and RT were 96.9 ± 4.3%, 368 ± 72 ms and 96.2 ± 5.5%, 360 ± 88 ms, for Oddball (mean value for 5 deviant types) and Optimum conditions (F b 1), respectively, showing that subjects were paying attention to the cross changes. As presented in Figs. 2 and 3, in both conditions, deviants elicited MMNs that peaked around 200 ms during the time window between 100 ms and 250 ms from stimulus onset, with an occipital–posterior scalp distribution. As shown in Table 1, the MMN mean amplitudes
Fig. 2. The raw ERPs and vMMNs elicited in Oddball and Optimum conditions, respectively, at LOT (left occipito-temporal sites), Oz and ROT (right occipito-temporal sites) as well as the 2D mapping of mean vMMN amplitudes, within the 100–250 ms time-window. Note that the presented vMMN waveforms were averaged for 5 types of vMMNs.
314
X. Qian et al. / International Journal of Psychophysiology 93 (2014) 311–315
Fig. 3. Five types of vMMNs at ROT site in Oddball and Optimum conditions.
(150–250 ms) for the 5 types of deviants significantly differed from zero in both conditions (ps b 0.05). The three-way ANOVA analysis for vMMN mean amplitudes (150– 250 ms post-stimuli onset) showed a significant main effect of Deviant type, F(4, 13) = 4.52, p b 0.01, partial η2 (effect size) = 0.45. Importantly, this effect was qualified by a two-way interaction of Deviant type ∗ Condition, F(4, 13) = 3.91, p b 0.05, partial η2 = 0.18. Further analyses showed that although the vMMN mean amplitudes did not differ between two stimulus conditions (p N 0.05), the orientation MMN was larger significantly for Optimal than for Oddball conditions (p b 0.05, partial η2 = 0.25). Across conditions and deviant types, there was no significant hemisphere effect (p N 0.1).
4. Discussion In the present study, we used the optimal visual paradigm in which 5 types of visual stimuli were integrated in one sequence to examine the people's visual processing skills faster and more comprehensively. As predicted, in Optimum condition, 5 types of vMMNs were elicited although with variation between the deviant types, in line with our previous reports (Shi et al., 2013). Importantly, the 5 different visual changes presented in one short sequence (optimum) resulted in MMN amplitudes that were at least as large as those obtained in the traditional one-deviant oddball paradigm, indicating that the present Optimum condition allowed an even shorter measurement time and without compromising the MMN amplitude.
Table 1 Mean MMN amplitudes (μV) for 5 types of deviants in different conditions. Deviant
Shape Orientation Color Size Duration
Oddball
Optimum
LOT
Oz
ROT
LOT
Oz
ROT
−0.66⁎ −0.45⁎ −1.05⁎⁎ −0.55⁎ −0.62⁎
−0.48⁎ −0.38⁎ −1.02⁎⁎ −0.53⁎ −0.60⁎
−0.79⁎ −0.62⁎ −0.95⁎⁎ −0.54⁎ −0.68⁎
−0.77⁎ −1.19⁎⁎ −1.13⁎⁎ −0.53⁎ −0.66⁎
−0.41⁎ −0.90⁎ −0.95⁎⁎ −0.59⁎ −0.75⁎
−0.60⁎ −1.13⁎⁎ −1.24⁎⁎ −0.52⁎ −0.65⁎
Note: One-tailed t-tests were conducted to determine whether the MMN mean amplitudes significantly differed from zero. ⁎ p b 0.05. ⁎⁎ p b 0.01.
Regarding the fact that the orientation changes elicited bigger MMN amplitude in the optimum paradigm in comparison to the traditional oddball paradigm, it is not accounted for easily. On the one hand, this difference could be related to the hierarchical dependence of deviance-related pre-attentive mechanisms, in line with earlier reports in auditory modality (Winkler and Czigler, 1998). Moreover, featurerelated visual ERP components were smaller or absent when another feature was processed more effectively (Smid et al., 1997). Therefore, the present pattern implicated that the detection of orientation changes could be more effective in Optimum condition with processing multifeatures than that in Oddball condition with processing one feature. This issue needs further investigation. On the other hand, it has been shown that less salient deviants elicited vMMN of lower amplitude/ delayed latency (Czigler et al., 2002) and hence, the enhanced orientation-MMN could be due to the orientation change in optimum paradigm more salient than that in oddball paradigm. Actually, the larger difference between standard and deviant stimuli could induce additional pop-out effect like parallel visual search and elicit lager visual MMN. Although the shapes of the deviant-minus-standard difference waves obtained in both paradigms and their prominently posterior location were comparable with previous vMMN studies (e.g., Kimura et al., 2009; Shi et al., 2013), our posterior vMMN also includes at least some N1 and refractory activity (Kimura, 2012). From this viewpoint, the posterior negativity observed in the present study could not be unambiguously regarded as representing the elicitation of genuine visual MMN. However, the observed negative differential posterior activity lasts to at least 250 ms and peaks about 200 ms post-stimulus onset, i.e., it was later than the pure early sensory activity indicated by N1 (see Fig. 2). Importantly, one recent vMMN study found that type of standard (physically the same or different from the deviant) did not matter for the generation of expression MMN, indicating that the contamination of the vMMN with the refractory reactions is non-fatal (Kreegipuu et al., 2013). To this end, to confirm the use of the current optimal paradigm in clinical populations, further investigation is necessary to eliminate the N1 refractory effect by use of the equiprobable vMMN paradigm (e.g., Kimura et al., 2009). In sum, we proposed an optimum paradigm in which every other visual feature is a standard and every other a deviant. This paradigm can be used to obtain 5 different visual MMNs in the same time in which usually only one visual MMN is obtained (the traditional one-deviant oddball paradigm). Particularly, this is of interest in view of clinical
X. Qian et al. / International Journal of Psychophysiology 93 (2014) 311–315
research work and praxis, where time effectiveness is of primary importance. References Czigler, I., 2007. Visual mismatch negativity — violation of nonattended environmental regularities. J. Psychophysiol. 21, 224–230. Czigler, I., Sulykos, I., 2010. Visual mismatch negativity to irrelevant changes is sensitive to task-relevant changes. Neuropsychologia 48, 1277–1282. Czigler, I., Balazs, L., Winkler, I., 2002. Memory-based detection of task-irrelevant visual changes. Psychophysiology 39, 869–873. Grimm, S., Bendixen, A., Deouell, L.Y., Schroger, E., 2009. Distraction in a visual multideviant paradigm: behavioral and event-related potential effects. Int. J. Psychophysiol. 72, 260–266. Kimura, M., 2012. Visual mismatch negativity and unintentional temporal-context-based prediction in vision. Int. J. Psychophysiol. 83, 144–155. Kimura, M., Katayama, J.i, Ohira, H., Schroeger, E., 2009. Visual mismatch negativity: new evidence from the equiprobable paradigm. Psychophysiology 46, 402–409. Kreegipuu, K., Kuldkepp, N., Sibolt, O., Toom, M., Allik, J., Näätänen, R., 2013. vMMN for schematic faces: automatic detection of change in emotional expression. Front. Hum. Neurosci. 7, 1–10. Näätänen, R., Pakarinen, S., Rinne, T., Takegata, R., 2004. The mismatch negativity (MMN): towards the optimal paradigm. Clin. Neurophysiol. 115, 140–144.
315
Näätänen, R., Paavilainen, P., Rinne, T., Alho, K., 2007. The mismatch negativity (MMN) in basic research of central auditory processing: A review. Clin. Neurophysiol. 118, 2544–2590. Pazo-Alvarez, P., Amenedo, E., Lorenzo-Lopez, L., Cadaveira, F., 2004. Effects of stimulus location on automatic detection of changes in motion direction in the human brain. Neurosci. Lett. 371, 111–116. Qiu, X., Yang, X., Qiao, Z., Wang, L., Ning, N., Shi, J., Zhao, L., Yang, Y., 2011. Impairment in processing visual information at the pre-attentive stage in patients with a major depressive disorder: a visual mismatch negativity study. Neurosci. Lett. 491, 53–57. Semlitsch, H.V., Anderer, P., Schuster, P., Presslich, O., 1986. A solution for reliable and valid reduction of ocular artifacts, applied to the P300 ERP. Psychophysiology 23 (6), 695–703. Shi, L., Wu, J., Sun, G., Dang, L., Zhao, L., 2013. Visual mismatch negativity in the “optimal” multi-feature paradigm. J. Integr. Neurosci. 12, 1–12. Smid, H., Jakob, A., Heinze, H.J., 1997. The organization of multidimensional selection on the basis of color and shape: an event-related brain potential study. Percept. Psychophys. 59, 693–713. Tales, A., Butler, S., 2006. Visual mismatch negativity highlights abnormal preattentive visual processing in Alzheimer's disease. Neuroreport 17, 887–890. Winkler, I., Czigler, I., 1998. Mismatch negativity: deviance detection or the maintenance of the ‘standard’. Neuroreport 17, 3809–3813. Zhao, L., Li, J., 2006. Visual mismatch negativity elicited by facial expressions under nonattentional condition. Neurosci. Lett. 410, 126–131.
Contents lists available at ScienceDirect
International Journal of Psychophysiology journal homepage: www.elsevier.com/locate/ijpsycho
Short communication
The visual mismatch negativity (vMMN): Toward the optimal paradigm Xiaosen Qian a, Yi Liu a,⁎, Bing Xiao a, Li Gao a, Songlin Li a, Lijie Dang b, Cuiping Si c, Lun Zhao b,⁎⁎ a b c
Sleep Disorders Center, Civil Aviation General Hospital, Beijing, China School of Psychological Research, Beijing Yiran Sunny Technology Co. Ltd., Beijing, China Department of Neurology, Jining No. 1 People's Hospital, Jining, China
a r t i c l e
i n f o
Article history: Received 5 February 2014 Received in revised form 4 June 2014 Accepted 6 June 2014 Available online 14 June 2014 Keywords: Visual mismatch negativity Visual discrimination Change detection Optimal paradigm
a b s t r a c t In the present article, we tested an optimal vMMN paradigm allowing one to obtain vMMNs for several visual attributes in a short time. vMMN responses to changes in color, duration, orientation, shape, and size were compared between the traditional ‘oddball’ paradigm (a single type of visual change in each sequence) and the optimal paradigm in which all the 5 types of changes appeared within the same sequence. The vMMNs obtained in the optimal paradigm were equal or larger in amplitude to those in the traditional vMMN paradigm. The optimal paradigm can provide 5 different vMMNs in the same time in which usually only one MMN is obtained. This short objective measure could putatively be used as an index for visual cognition function especially in clinical research. © 2014 Elsevier B.V. All rights reserved.
1. Introduction As an important cognitive function for human survival, the automatic change detection occurs at a very early stage of information processing. To date, ample evidence indicated that the mismatch negativity (MMN) of event-related potential (ERP) components is a reliable indicator for exploring automatic change detection, reflecting the brain's detection of an unintentional disruption in the regularity of temporal events (e.g., Näätänen et al., 2007). Generally, MMN can be elicited by infrequent deviant stimuli inserted randomly in a sequence of frequent standard stimuli presented outside of the focus of attention. Although the MMN component has been widely investigated in auditory modality, analog of auditory MMN was also found in response to visual deviants such as color (Czigler et al., 2002), orientation (e.g., Kimura et al., 2009), size (Kimura et al., 2009), shape (Grimm et al., 2009), duration (Qiu et al., 2011) as well as facial expressions (e.g., Zhao and Li, 2006). Although visual MMN (vMMN) is elicited by task-irrelevant events, recent studies have shown that vMMN is influenced by attentional manipulations (e.g., Czigler, 2007; Czigler and Sulykos, 2010; for a review, Kimura, 2012). Therefore, instead of being an index of fully automatic processes, it was proposed that vMMN reflects the existence of unintentional temporal-context-based prediction (Kimura, 2012). As an inexpensive and non-invasive method, it is not surprising that the vMMN has recently received considerable attention as a tool of
⁎ Correspondence to: Y. Liu, Sleep Disorders Center, Civil Aviation General Hospital, Beijing 100123, China. ⁎⁎ Correspondence to: L. Zhao, School of Psychological Research, Beijing Yiran Sunny Technology Co. Ltd, Beijing 100088, China. E-mail addresses: [email protected] (Y. Liu), [email protected] (L. Zhao).
http://dx.doi.org/10.1016/j.ijpsycho.2014.06.004 0167-8760/© 2014 Elsevier B.V. All rights reserved.
clinical research. For instance, vMMN in response to the deviant duration was significantly reduced in depression patients (Qiu et al., 2011) as well as in Alzheimer's disease (Tales and Butler, 2006). It should be noted that different kinds of deviant features elicit distinctive vMMNs, (maybe) with different cortical origins. For instance, the orientation MMN was generated at the medial prefrontal areas as well as at the visual extrastriate area (Kimura et al., 2009), whereas the neural generator of vMMN elicited by directions of motion included the occipital visual extrastriate areas (including motion response areas) (PazoAlvarez et al., 2004). Previous vMMN clinical studies tested only a single deviant type and hence, it is not clear whether vMMNs elicited by other deviant types are affected by the disease. Therefore, it is necessary for developing a multi-feature vMMN paradigm with different deviant types integrated in one complex visual context. Indeed, the practical problem, especially in clinical application, is that traditional vMMN paradigms are usually time-consuming when several types of deviants are presented in different stimulus blocks. Näätänen et al. (2004) developed a new multi-feature paradigm in which five types of acoustic changes (frequency, intensity, duration, perceived sound-source location, and gap) were presented in the same sound sequence. Indeed, in this paradigm different auditory attributes can elicit different MMNs, and most importantly, the memory trace of the standard stimuli can enhance with respect to the stimulus attributes they have in common. Although the multi-feature paradigm reduced the time of the experiment, the MMNs using the new paradigm did not differ from those using the traditional oddball-paradigm. To date, only few studies investigated vMMN using the multi-deviant paradigm. For example, Grimm et al. (2009) found that the vMMN depended on the changed feature, which might reflect either the differences in saliency between the feature changes or the natural hierarchy
312
X. Qian et al. / International Journal of Psychophysiology 93 (2014) 311–315
in processing of various visual stimulus dimensions. Kreegipuu et al. (2013) compared vMMNs elicited by angry and happy schematic faces in a traditional oddball design (deviant stimuli, 12.5%) and in an optimal multi-feature paradigm with several deviant stimuli (altogether 37.5%) and found that the expressional MMN (EMMN) was equally elicited in both paradigms. They considered that in order to save time and other experimental resources the use of the optimum (multi-feature) design for EMMN should be encouraged (Kreegipuu et al., 2013). However, in their study, the presentation probability for a single deviant in the oddball paradigm was about twice as high as in the optimum paradigm, that is, the refractory state of same event-related potential components was different between two paradigms (Kreegipuu et al., 2013). On the basis of the auditory MMN optimal paradigm by Näätänen et al (2004), we recently proposed a time-saving and multi-feature visual MMN paradigm, in which 5 different vMMNs could be recorded with lowprobability deviant stimuli (color, duration, orientation, shape, and size) inserted randomly in a sequence of frequent standard stimuli (Shi et al., 2013). However, in order to successfully address the reliability of multi-feature visual MMN paradigm, it is necessary to show that in fact the vMMNs recorded from the multi-feature deviant paradigm are identical or not smaller than to those recorded in the same subjects using traditional single-feature deviants. This possibility will be investigated in the present study.
2. Method 2.1. Participants Fourteen undergraduate volunteers (8 females, all right-handed) with normal or corrected-to-normal visual acuity served as participants (the average age = 22.3 years, age range = 19–25 years). All participants received payments for their contribution. This research was approved by the Ethical Committee of Civil Aviation General Hospital in
accordance with the Declaration of Helsinki and all participants gave their written informed consents before the experiment. 2.2. Stimuli and procedure The subject sat on a comfortable chair in a darkened, sound attenuated and electrically shielded room. In the center of screen, a black cross was displayed throughout the stimulus blocks. From time to time, the cross became bigger or smaller unpredictably. Subjects were instructed to ignore the peripheral stimuli and press the left or the right button as quickly and accurately as possible when the size of the cross became bigger or smaller. The number of targets was 18–22 (mean 20) in each sequence. In the peripheral sides of the field, two identical visual stimuli were simultaneously presented from a distance of 1 m, with the stimulus onset asynchrony (SOA) of 600 ms. The concurrently presented peripheral stimuli occurred on the left and right of the cross (4.5° distance from center of cross to center of the peripheral stimulus). The solid red rectangles (30 mm in length and 10 mm in width) served as standard with duration of 50 ms. As presented in Fig. 1, five types of deviant stimuli were included: two solid blue rectangles (color deviant), two shifted 90° solid red rectangles to the standard stimuli (orientation deviant), two red rectangles with 100 ms duration (duration deviant), two solid red ellipse (shape deviant) and two wide red rectangles (size deviant). There were 2 different conditions in the present study (Fig. 1): the traditional oddball paradigm (a single type of deviant in a sequence) and another condition (denoted as ‘Optimum’), in which all 5 types of deviants occurred within the same sequence (Näätänen et al., 2004). In line with the method proposed by Näätänen et al (2004), in the Optimum condition, all 5 deviants (p = 10% for each; total 50%) were presented in the same sequence so that every other visual stimulus was a standard stimulus (p = 50%). The rationale of this paradigm was that the other deviants can strengthen the memory trace of the standard with respect to those levels of stimulus attributes they had in
Fig. 1. Top: Examples for stimulus configurations: a) standard display, b) color deviant display, c) orientation deviant display, d) duration deviant (duration: 100 ms) display, e) shape deviant display, f) size deviant display. Down: Schematic illustration of the 2 stimulus conditions used: traditional Oddball and Optimum conditions. S denotes standard stimulus and DX stimuli of different deviant types (D1 — color deviant, D2 — orientation deviant, D3 — duration deviant, D4 — shape deviant, D5 — size deviant). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
X. Qian et al. / International Journal of Psychophysiology 93 (2014) 311–315
common. The deviants were presented in an array of 5 deviants (Fig. 1). The stimuli were presented in five 3 min sequences (1500 stimuli in total; 150 for each deviant stimulus and 750 for standard stimulus), with the total recording time for the 5 types of deviants thus being 15 min. The stimulus sequences of the traditional Oddball condition were created using those of Optimum condition by replacing with standards all deviants except for those of one category. In this way, the times of the occurrence of a certain type of deviant stimulus in the sequence were identical to those of Optimum condition. Stimuli were presented in five 3 min sequences for each deviant type, resulting in a total recording time of 75 min. In sum, Optimum condition was composed of 5 stimulus sequences, whereas in Oddball condition, there were 25 sequences, 5 sequences for each deviant type. The order of conditions as well as the order of sequences in different conditions was balanced across subjects. 2.3. EEG recording The electroencephalogram (EEG) was continuously recorded with ANT asaLab Amplifier (www.ant-neuro.com), using an electrode cap with 32-channel Ag/AgCl electrodes according to the extended international 10–20 system. The reference electrode was placed on the nose tip. Vertical EOG was recorded with two electrodes above and below the right eye. Horizontal EOG was recorded with two electrodes at the right and left outer canthi of the eyes. The impedances of the electrodes were kept below 5 kΩ throughout the experiment. EEG and EOG signals were amplified with a band pass of 0.1–100 Hz at a sampling rate of 500 Hz. After EOG artifact correction (Semlitsch, et al., 1986), the EEG was segmented into the epoch from 100 ms pre-stimulus (as the baseline) to 500 ms post-stimulus. The trails contaminated with artifacts greater than ± 100 μV were rejected before averaging. The trails with targets when participants' responses occurred and where participants' responses did not occur (i.e., missed targets) as well as the trials without targets
313
where participants' false positive responses occurred were rejected from averaging. The EEG segments were averaged separately for standard and deviant (95.5 ± 8.6, 101.3 ± 10.0, 98.7 ± 9.9, 97.7 ± 11.2, and 99.9 ± 10.1 for color, duration, orientation, size, and shape deviants, respectively) stimuli in different conditions. MMN was obtained by subtracting ERPs to standard stimuli from ERPs to deviant stimuli for each visual feature, respectively. Since in many participants the vMMN peaks were not easily discernible at all sites and in each condition (see Figs. 2 and 3), the mean amplitudes of vMMN were measured for the 150–250 ms time windows post-stimulus onset at the occipital–temporal electrode sites, left occipital–temporal electrode sites (LOT) including T5 and O1, right occipital–temporal electrode sites (ROT) including T6 and O2, and midline occipital site (Oz), respectively. One-tailed t tests were conducted to determine whether the vMMN mean amplitudes significantly differed from zero. Three-way analyses of variance (ANOVA) with repeated measures were conducted to test the effects of Condition (2 levels: Optimum and Oddball), Deviant type (color, duration, orientation, shape, and size) and Site (LOT, Oz, ROT). Greenhouse–Geisser corrections were made when appropriate. Newman–Keuls tests were carried out as post hoc analyses. 3. Results Responses for detecting cross changes in the center of the screen were scored as hit if the correct button was pressed within 150 to 1000 ms after targets onset. Mean accuracy rate and RT were 96.9 ± 4.3%, 368 ± 72 ms and 96.2 ± 5.5%, 360 ± 88 ms, for Oddball (mean value for 5 deviant types) and Optimum conditions (F b 1), respectively, showing that subjects were paying attention to the cross changes. As presented in Figs. 2 and 3, in both conditions, deviants elicited MMNs that peaked around 200 ms during the time window between 100 ms and 250 ms from stimulus onset, with an occipital–posterior scalp distribution. As shown in Table 1, the MMN mean amplitudes
Fig. 2. The raw ERPs and vMMNs elicited in Oddball and Optimum conditions, respectively, at LOT (left occipito-temporal sites), Oz and ROT (right occipito-temporal sites) as well as the 2D mapping of mean vMMN amplitudes, within the 100–250 ms time-window. Note that the presented vMMN waveforms were averaged for 5 types of vMMNs.
314
X. Qian et al. / International Journal of Psychophysiology 93 (2014) 311–315
Fig. 3. Five types of vMMNs at ROT site in Oddball and Optimum conditions.
(150–250 ms) for the 5 types of deviants significantly differed from zero in both conditions (ps b 0.05). The three-way ANOVA analysis for vMMN mean amplitudes (150– 250 ms post-stimuli onset) showed a significant main effect of Deviant type, F(4, 13) = 4.52, p b 0.01, partial η2 (effect size) = 0.45. Importantly, this effect was qualified by a two-way interaction of Deviant type ∗ Condition, F(4, 13) = 3.91, p b 0.05, partial η2 = 0.18. Further analyses showed that although the vMMN mean amplitudes did not differ between two stimulus conditions (p N 0.05), the orientation MMN was larger significantly for Optimal than for Oddball conditions (p b 0.05, partial η2 = 0.25). Across conditions and deviant types, there was no significant hemisphere effect (p N 0.1).
4. Discussion In the present study, we used the optimal visual paradigm in which 5 types of visual stimuli were integrated in one sequence to examine the people's visual processing skills faster and more comprehensively. As predicted, in Optimum condition, 5 types of vMMNs were elicited although with variation between the deviant types, in line with our previous reports (Shi et al., 2013). Importantly, the 5 different visual changes presented in one short sequence (optimum) resulted in MMN amplitudes that were at least as large as those obtained in the traditional one-deviant oddball paradigm, indicating that the present Optimum condition allowed an even shorter measurement time and without compromising the MMN amplitude.
Table 1 Mean MMN amplitudes (μV) for 5 types of deviants in different conditions. Deviant
Shape Orientation Color Size Duration
Oddball
Optimum
LOT
Oz
ROT
LOT
Oz
ROT
−0.66⁎ −0.45⁎ −1.05⁎⁎ −0.55⁎ −0.62⁎
−0.48⁎ −0.38⁎ −1.02⁎⁎ −0.53⁎ −0.60⁎
−0.79⁎ −0.62⁎ −0.95⁎⁎ −0.54⁎ −0.68⁎
−0.77⁎ −1.19⁎⁎ −1.13⁎⁎ −0.53⁎ −0.66⁎
−0.41⁎ −0.90⁎ −0.95⁎⁎ −0.59⁎ −0.75⁎
−0.60⁎ −1.13⁎⁎ −1.24⁎⁎ −0.52⁎ −0.65⁎
Note: One-tailed t-tests were conducted to determine whether the MMN mean amplitudes significantly differed from zero. ⁎ p b 0.05. ⁎⁎ p b 0.01.
Regarding the fact that the orientation changes elicited bigger MMN amplitude in the optimum paradigm in comparison to the traditional oddball paradigm, it is not accounted for easily. On the one hand, this difference could be related to the hierarchical dependence of deviance-related pre-attentive mechanisms, in line with earlier reports in auditory modality (Winkler and Czigler, 1998). Moreover, featurerelated visual ERP components were smaller or absent when another feature was processed more effectively (Smid et al., 1997). Therefore, the present pattern implicated that the detection of orientation changes could be more effective in Optimum condition with processing multifeatures than that in Oddball condition with processing one feature. This issue needs further investigation. On the other hand, it has been shown that less salient deviants elicited vMMN of lower amplitude/ delayed latency (Czigler et al., 2002) and hence, the enhanced orientation-MMN could be due to the orientation change in optimum paradigm more salient than that in oddball paradigm. Actually, the larger difference between standard and deviant stimuli could induce additional pop-out effect like parallel visual search and elicit lager visual MMN. Although the shapes of the deviant-minus-standard difference waves obtained in both paradigms and their prominently posterior location were comparable with previous vMMN studies (e.g., Kimura et al., 2009; Shi et al., 2013), our posterior vMMN also includes at least some N1 and refractory activity (Kimura, 2012). From this viewpoint, the posterior negativity observed in the present study could not be unambiguously regarded as representing the elicitation of genuine visual MMN. However, the observed negative differential posterior activity lasts to at least 250 ms and peaks about 200 ms post-stimulus onset, i.e., it was later than the pure early sensory activity indicated by N1 (see Fig. 2). Importantly, one recent vMMN study found that type of standard (physically the same or different from the deviant) did not matter for the generation of expression MMN, indicating that the contamination of the vMMN with the refractory reactions is non-fatal (Kreegipuu et al., 2013). To this end, to confirm the use of the current optimal paradigm in clinical populations, further investigation is necessary to eliminate the N1 refractory effect by use of the equiprobable vMMN paradigm (e.g., Kimura et al., 2009). In sum, we proposed an optimum paradigm in which every other visual feature is a standard and every other a deviant. This paradigm can be used to obtain 5 different visual MMNs in the same time in which usually only one visual MMN is obtained (the traditional one-deviant oddball paradigm). Particularly, this is of interest in view of clinical
X. Qian et al. / International Journal of Psychophysiology 93 (2014) 311–315
research work and praxis, where time effectiveness is of primary importance. References Czigler, I., 2007. Visual mismatch negativity — violation of nonattended environmental regularities. J. Psychophysiol. 21, 224–230. Czigler, I., Sulykos, I., 2010. Visual mismatch negativity to irrelevant changes is sensitive to task-relevant changes. Neuropsychologia 48, 1277–1282. Czigler, I., Balazs, L., Winkler, I., 2002. Memory-based detection of task-irrelevant visual changes. Psychophysiology 39, 869–873. Grimm, S., Bendixen, A., Deouell, L.Y., Schroger, E., 2009. Distraction in a visual multideviant paradigm: behavioral and event-related potential effects. Int. J. Psychophysiol. 72, 260–266. Kimura, M., 2012. Visual mismatch negativity and unintentional temporal-context-based prediction in vision. Int. J. Psychophysiol. 83, 144–155. Kimura, M., Katayama, J.i, Ohira, H., Schroeger, E., 2009. Visual mismatch negativity: new evidence from the equiprobable paradigm. Psychophysiology 46, 402–409. Kreegipuu, K., Kuldkepp, N., Sibolt, O., Toom, M., Allik, J., Näätänen, R., 2013. vMMN for schematic faces: automatic detection of change in emotional expression. Front. Hum. Neurosci. 7, 1–10. Näätänen, R., Pakarinen, S., Rinne, T., Takegata, R., 2004. The mismatch negativity (MMN): towards the optimal paradigm. Clin. Neurophysiol. 115, 140–144.
315
Näätänen, R., Paavilainen, P., Rinne, T., Alho, K., 2007. The mismatch negativity (MMN) in basic research of central auditory processing: A review. Clin. Neurophysiol. 118, 2544–2590. Pazo-Alvarez, P., Amenedo, E., Lorenzo-Lopez, L., Cadaveira, F., 2004. Effects of stimulus location on automatic detection of changes in motion direction in the human brain. Neurosci. Lett. 371, 111–116. Qiu, X., Yang, X., Qiao, Z., Wang, L., Ning, N., Shi, J., Zhao, L., Yang, Y., 2011. Impairment in processing visual information at the pre-attentive stage in patients with a major depressive disorder: a visual mismatch negativity study. Neurosci. Lett. 491, 53–57. Semlitsch, H.V., Anderer, P., Schuster, P., Presslich, O., 1986. A solution for reliable and valid reduction of ocular artifacts, applied to the P300 ERP. Psychophysiology 23 (6), 695–703. Shi, L., Wu, J., Sun, G., Dang, L., Zhao, L., 2013. Visual mismatch negativity in the “optimal” multi-feature paradigm. J. Integr. Neurosci. 12, 1–12. Smid, H., Jakob, A., Heinze, H.J., 1997. The organization of multidimensional selection on the basis of color and shape: an event-related brain potential study. Percept. Psychophys. 59, 693–713. Tales, A., Butler, S., 2006. Visual mismatch negativity highlights abnormal preattentive visual processing in Alzheimer's disease. Neuroreport 17, 887–890. Winkler, I., Czigler, I., 1998. Mismatch negativity: deviance detection or the maintenance of the ‘standard’. Neuroreport 17, 3809–3813. Zhao, L., Li, J., 2006. Visual mismatch negativity elicited by facial expressions under nonattentional condition. Neurosci. Lett. 410, 126–131.