Skip to main content
NCBI home page
Human Brain Mapping logoLink to Human Brain Mapping
. 2024 Jul 22;45(11):e26793. doi: 10.1002/hbm.26793

Electrophysiological correlation of auditory selective spatial attention in the “cocktail party” situation

Hongxing Liu 1,2, Yanru Bai 1,2,3, Qi Zheng 1,2, Jihan Liu 1,2, Jianing Zhu 1,2, Guangjian Ni 1,2,3,4,✉
PMCID: PMC11261592  PMID: 39037186

Abstract

The auditory system can selectively attend to the target source in complex environments, the phenomenon known as the “cocktail party” effect. However, the spatiotemporal dynamics of electrophysiological activity associated with auditory selective spatial attention (ASSA) remain largely unexplored. In this study, single‐source and multiple‐source paradigms were designed to simulate different auditory environments, and microstate analysis was introduced to reveal the electrophysiological correlates of ASSA. Furthermore, cortical source analysis was employed to reveal the neural activity regions of these microstates. The results showed that five microstates could explain the spatiotemporal dynamics of ASSA, ranging from MS1 to MS5. Notably, MS2 and MS3 showed significantly lower partial properties in multiple‐source situations than in single‐source situations, whereas MS4 had shorter durations and MS5 longer durations in multiple‐source situations than in single‐source situations. MS1 had insignificant differences between the two situations. Cortical source analysis showed that the activation regions of these microstates initially transferred from the right temporal cortex to the temporal–parietal cortex, and subsequently to the dorsofrontal cortex. Moreover, the neural activity of the single‐source situations was greater than that of the multiple‐source situations in MS2 and MS3, correlating with the N1 and P2 components, with the greatest differences observed in the superior temporal gyrus and inferior parietal lobule. These findings suggest that these specific microstates and their associated activation regions may serve as promising substrates for decoding ASSA in complex environments.

Keywords: auditory selective spatial attention, cortical source analysis, electrophysiological correlation, ERP microstate analysis

Event‐related potential microstate analysis is introduced to reveal the electrophysiological correlation of auditory selective spatial attention in the “cocktail party” situation. Research has shown that some specific microstates and their associated activation regions hold promise as potential substrates for decoding auditory selective spatial attention in complex environments like “cocktail parties.”

graphic file with name HBM-45-e26793-g001.jpg

Practitioner Points.

  • The spatiotemporal sequence of auditory selective spatial attention is revealed through the ERP microstate analysis.

  • The superior temporal gyrus and inferior parietal lobule play critical modulation roles in auditory selective spatial attention.

  • Specific microstates and their associated activation regions are promising substrates for decoding auditory selective spatial attention.

  • This study offers a novel insight into electrophysiological correlates of auditory selective spatial attention in the “cocktail party” situation.

1. INTRODUCTION

Auditory spatial perception is one of the most important senses for humans. The human auditory system can selectively attend to target sources while ignoring irrelevant sources in complex auditory environments, and this phenomenon is known as the “cocktail party” effect (Cherry, 1953). This process involves not only the detection and discrimination of target sources, but also the dynamic switching and accessing of different levels of information within diverse stimuli, and so on (Grothe et al., 2010; Lewald et al., 2018). Especially in complex auditory environments with multiple distractor sources in space, the ability of auditory selective spatial attention (ASSA) plays a pivotal role in this process. Although this ability is critical for many activities, such as verbal communication and traffic driving, the underlying electrophysiological activity of this phenomenon remains insufficiently understood.

Recent research into the neural basis of auditory selective attention has extensively employed the dichotic listening paradigm, which is the psychophysical method that presents two auditory stimuli to the listener simultaneously. Specifically, studies have elucidated the cortical representations and encoding patterns of auditory objects in scenarios involving competing speakers (Ding & Simon, 2012b; Mesgarani & Chang, 2012). Moreover, concurrent auditory objects have been established as the fundamental representational units for top‐down attentional modulation and bottom‐up neural adaptation (Ding & Simon, 2012a). Research has demonstrated that selective attention is primarily mediated by certain brain regions associated with auditory control, attentional regulation, and speech processing (Kaufman & Golumbic, 2023). In addition, selective and enhanced auditory performance has been confirmed in higher auditory regions, such as the superior temporal gyrus (STG) (Hamilton et al., 2021). Furthermore, Hillyard et al. (1973) identified neural correlates of auditory selective attention using event‐related potential (ERP), noting increased ERP amplitude, particularly in the N1 component, when auditory stimuli were attended. However, these studies primarily concentrated on auditory selective attention mechanisms using the dichotic listening paradigm, without considering the complexities of the environment with multiple distractor sources in space.

In complex auditory environments, such as those with multiple distractor sources or dynamic sources, selectively attending to target sources demands heightened attentional resources from the auditory system. In recent years, numerous studies have employed various neuroscience techniques to investigate the electrophysiological activity of ASSA in complex auditory environments. In particular, investigations employing functional magnetic resonance imaging (fMRI) have shown greater activity in the posterior STG, anterior insula, supplementary motor area, and frontoparietal network under such conditions (Zündorf et al., 2013). Additionally, voxel‐based lesion behavior mapping (VLBM) in patients with stroke has demonstrated the pivotal role of the right temporal plane and left inferior frontal gyrus in ASSA (Zündorf et al., 2014). Moreover, researchers using magnetoencephalography (MEG) have revealed that specific activation is observed in the left precentral sulcus during auditory attention (Bharadwaj et al., 2014). Taken together, these findings underscore the distinctions in neural activity associated with ASSA in complex auditory environments.

Electroencephalography (EEG) has emerged as a valuable tool for investigating the electrophysiological activity of ASSA in complex auditory environments owing to its high temporal resolution, portability, and cost‐effectiveness. Studies employing EEG have shown that complex auditory scenes can evoke the stronger P1 and N2 components, whereas the N1 components exhibit reduced amplitude (Lewald & Getzmann, 2015). In addition, some studies have confirmed that attending to target sources has been associated with negativity in the N2 latency range at anterior contralateral electrodes, providing insights into selective spatial attention in complex auditory environments (Lewald et al., 2016; Gamble & Luck, 2011). Furthermore, spatial oddball paradigms have demonstrated the existence of deviant‐minus‐standard difference potentials in complex auditory environments, which reflect mismatched negative, P3a, and P3b components (Lewald et al., 2018). These findings suggest potential correlations between spatial attention allocation and ERP components in complex auditory environments. However, the above research did not consider the rich spatial information in the EEG signals and instead relied on the choice of reference electrodes.

Beyond traditional ERP analysis, microstate analysis offers a reference‐free approach that uses multiple electrodes to capture the rich spatial information of the EEG signals. This method employs the electric field distribution of multichannel signals to characterize the overall electrophysiological state. It includes cognitive processing information extracted from the EEG signals and spatial information about the scalp voltage distribution (Murray et al., 2008). Recent applications of microstate analysis have explored its utility in understanding neural mechanisms across various disorders, including schizophrenia and dementia (Lehmann et al., 2005; Nishida et al., 2013). Studies on resting‐state EEG microstates have identified four distinct categories, microstates A, B, C, and D, which have shown high similarity across different studies (Tarailis et al., 2023). Nevertheless, the parameters and topography of these microstates were associated with various factors, including neurological and psychiatric disorders, age, and cognitive processes (Khanna et al., 2015; Metzger et al., 2023). Additionally, research has explored the correlation between resting‐state EEG microstates and resting‐state fMRI (Britz et al., 2010). However, the microstate analysis of resting‐state EEG may not be applicable to task states, as the EEG signals during tasks exhibit fluctuations and irregularities at different spatiotemporal scales.

In contrast, ERP microstate analysis has been proposed as a means of investigating task states. The topology of ERP microstates can reflect the instantaneous state of the overall activity of the functional brain network, and changes in the topological structure can represent shifts in collaborative brain activity. A multitude of studies have validated the reliability and objectivity of this method for task states, as different voltage topographies correspond to distinct neural sources (Murray et al., 2008). For instance, Dering and Donaldson (2016) identified distinct ERP microstate categories during the P1 and N1 time windows in response to different facial stimuli. Kong et al. (2018) investigated the dynamic ERP microstates during mental rotation tasks and found that the topologies of the N140, P200, and P300 corresponded to distinct microstate patterns. Overall, ERP microstate analysis provides a more comprehensive understanding of the spatiotemporal dynamics involved, thereby facilitating the uncovering of electrophysiological correlations.

This study aimed to introduce ERP microstate analysis to investigate the electrophysiological correlates of ASSA in situations akin to the “cocktail party” effect. First, single‐source and multiple‐source paradigms were designed to simulate different auditory environments. Second, the temporal correlation of ASSA processing was determined by ERP analysis. Furthermore, the spatiotemporal sequence of different situations was constructed by ERP microstate analysis and quantitatively measured by global explainable variance (GEV), global field power (GFP), duration, and coverage. Finally, the cortical source activity regions of each microstate and the differential activity regions of specific microstates were revealed using standardized low‐resolution brain electromagnetic tomography (sLORETA). The overall research framework is depicted in Figure 1.

FIGURE 1.

FIGURE 1

Open in a new tab

The overall research framework. (a) EEG data recording, (b) EEG data preprocessing, (c) ERP analysis, (d) ERP microstate analysis, (e) Microstate properties quantitatively measured, and (f) Cortical source analysis.

2. METHODS

2.1. Participants

Twenty‐six healthy right‐handed participants (16 males and 10 females) with a mean age of 24.5 years (SD 3.4) were recruited from Tianjin University for this study. All participants underwent pure‐tone audiometry at 0.5, 1, 2, and 4 kHz in both ears, with hearing thresholds at each frequency within the normal range (≤20 dB). None of the participants had a history of hearing impairment, neurological or psychiatric disorders, or drug addiction. Informed consent was obtained from all participants prior to the experiment, and the experimental protocol was approved by the Ethics Committee of Tianjin University.

2.2. Experimental paradigm design

The single‐source and multiple‐source paradigms were designed to simulate different auditory environments. Specifically, the experimental paradigm included four sound source directions at 90° and 45° on either side of the participant's midplane. In the single‐source experiment, one target sound was played randomly from one of the four loudspeakers, as shown in Figure 2a. In the multiple‐source experiment, four different sounds were played simultaneously and randomly from the four separate loudspeakers, with three serving as distractors, as shown in Figure 2b. The sound field environment was created by a semicircular loudspeaker array with a radius of 1 m. Four animal sounds (dog, duck, chicken, and sheep) were used as auditory stimuli to assist participants in focusing and eliminating semantic factors. The initial sound files were selected from an online sound library and adjusted to a constant duration of 600 ms using Adobe Audition 2022 software. The sounds were digitized at a 48 kHz sampling rate and 16‐bit resolution, then converted to analog form using a virtual sound card (DVS‐TK‐001, Dante).

FIGURE 2.

FIGURE 2

Open in a new tab

Schematic diagram of the experimental paradigm. (a) Single‐source and (b) Multiple‐source situations.

Prior to the experiment, the target sound was played from a front loudspeaker at 0° and displayed on the screen. Subsequently, the sound stimuli were played and participants identified the direction of the target sound by pressing different keys after the stimulus ended 1000 ms. The 1000 ms delay was implemented to eliminate the action initiation potential component (Di Russo et al., 2017; Yu et al., 2014). Following a 500 ms interval of silence, a new set of sound stimuli was presented. A total of 600 trials were performed in each group, divided into three blocks of approximately 12 min each. The ASSA tasks were executed using MATLAB 2021b and Psychtoolbox‐3 (Brainard, 1997). In addition, it should be emphasized that participants received brief training prior to the experiment to familiarize themselves with the tasks, and the experiment was conducted in a semi‐anechoic chamber.

2.3. Data recording and preprocessing

EEG data were recorded from 64 electrodes using the Neuracle device (NeuSen W, Neuracle Technology Co., Ltd., China), with electrode placement following the international 10–20 system. The sampling rate was 1000 Hz, with an online 0.1–150 Hz bandpass and a 50 Hz notch filter. When recording EEG data with the Neural software, the REF electrode served as the physical reference electrode, while the GND electrode served as the ground electrode, with all electrode impedances below 10 kΩ. In addition, behavioral data were recorded using the MATLAB software.

The raw EEG data were preprocessed using the open‐source EEGLAB toolbox of the MATLAB software. Initially, the data were digitally filtered with a 0.1–30 Hz bandpass filter to reduce low‐frequency drift and high‐frequency noise. Subsequently, the data were down‐sampled to 256 Hz for further analysis, and the artifacts of electromyography and electrooculography were removed using independent component analysis. Next, the data were divided into 700 ms periods, consisting of 100 ms before and 600 ms after each stimulus, and baseline correction was performed using the first 100 ms. In addition, periods where the maximum to minimum values exceeded 100 μV were manually removed. Finally, the reference was reset to the average of all the electrodes to improve the signal‐to‐noise ratio.

2.4. Event‐related potential analysis

The ERP was obtained by averaging single‐trial epochs for single‐source and multiple‐source situations, resulting in single‐subject and grand average epochs. In this study, the N1 and P2 components of the ERP signals were specifically extracted and analyzed. Specifically, the amplitude of the N1 and P2 components was measured by averaging 10 sample points within a 20 ms time window centered on the peak within the fixed time windows (N1: 70–170 ms, P2: 160–260 ms) (Si et al., 2019). The latency of the N1 and P2 components was defined as the time interval between the onset of the stimulus and the peak of the corresponding component.

2.5. Event‐related potential microstate analysis

Microstate analysis segments the EEG data based on spatial characteristics, forming stable states from segments with similar spatial distributions. The clustering process employs various strategies depending on the EEG signal's characteristics (Michel & Koenig, 2018; Mishra et al., 2020). Owing to fluctuations and irregularities in ERP signals, the hierarchical clustering method was employed for ERP microstate analysis. In particular, the topographic atomize and agglomerate hierarchical clustering algorithm was employed to cluster the brain topography, given that the polarity of the brain topography cannot be disregarded for ERP microstate analysis (Brunet et al., 2011).

ERP microstate analysis was conducted using the Cartool 3.61 software (Brunet et al., 2011). Initially, the total average ERP was derived by averaging the data of all subjects from the single‐source and multiple‐source situations and then segmented to extract optimal clustering templates. The optimal number of clusters was determined using the GEV, Krzanowski‐Lai, and MeanCriteria (Bréchet et al., 2019). Subsequently, the mean ERP data for each condition were fitted with the optimal clustering template and converted into a microstate sequence. Microstate segments that were shorter than 30 ms were reassigned to the next most probable microstate category through temporal smoothing. Finally, the ERP data for each participant were fitted with the clustering template, and microstates from different conditions were quantitatively measured using GEV, GFP, duration, and coverage. GEV represents the percentage of total variance explained by a given microstate, while GFP represents the standard deviation of all electrode channel voltages on the scalp at a given moment. Duration indicates the mean duration of microstate stability, while coverage denotes the percentage of time the microstates occupy within the time window (Khanna et al., 2015).

2.6. Cortical source analysis

The cortical source activity regions of specific microstates were evaluated using sLORETA, which performs inverse operations to determine the three‐dimensional (3D) brain space solution from the initial scalp EEG readings (Bradley et al., 2016; Pascual‐Marqui, 2002). This method assumes the brain as a head model composed of three tissue layers: brain, skull, and scalp. These layers were scanned and digitized using the MNI152 standard, which averages MRI images from 152 participants. The brain was calibrated into 6239 stereo pixels, each 5 mm in size, representing 3D XYZ coordinates and containing the current density value (Hofmann‐Shen et al., 2020). Specifically, sLORETA employed the non‐parametric statistical analysis (SnPM) based on log‐F‐ratio statistics for paired groups, with a significance threshold of p < .05 (Nichols & Holmes, 2002). A total of 5000 permutations were performed to determine the significance of each test and correct for multiple comparisons (Chen et al., 2023).

2.7. Statistical analysis

The paired t‐tests were employed for the statistical analysis of behavioral and ERP data between single‐source and multiple‐source situations. Repeated‐measures analysis of variance (ANOVA) was employed to analyze the behavioral data for different sound directions (90° and 45° of the left and right sides). Two‐factor repeated‐measures ANOVA was conducted for the four microstate parameters (GEV, GFP, duration, and coverage), with factors for condition (single or multiple sources) and microstate class. Results with p < .05 were considered statistically significant, with the p‐values corrected for multiple comparisons using the Bonferroni method. All procedures were conducted using SPSS Statistics 22 and GraphPad Prism 10 software.

3. RESULTS

3.1. Behavioral performance

The ability to localize sound sources in single‐source and multiple‐source situations was evaluated through behavioral data. Figure 3a demonstrates that the accuracy in localizing each sound direction was relatively high in single‐source situations, attributed to the simplicity of the task. Furthermore, the ANOVA revealed no main effect of direction on accuracy (F(2.167,54.18) = 0.845, p = .443). Figure 3b indicates that, in multiple‐source situations, the accuracy of localizing sounds at 45° was higher compared to localizing at 90°. Nevertheless, the ANOVA also showed no main effect of direction on accuracy (F(2.172,54.31) = 0.709, p = .508). Furthermore, the statistical analysis showed that the mean accuracy in single‐source situations (99.2%) was significantly higher than that in multiple‐source situations (87.6%) (t (25) = 7.113, p < .001), as shown in Figure 3c. It is noteworthy that trials with incorrect responses were excluded from all subsequent analyses to minimize the potential impact of differences in error rates.

FIGURE 3.

FIGURE 3

Open in a new tab

Behavioral results in single‐source and multiple‐source situations. (a) Single‐source, (b) Multiple‐source, and (c) Comparison of the mean in both situations. (*** p < .001). The negative and positive signs indicate the left and right directions, respectively. The thick dotted line indicates the second quartile, while the thin dotted lines indicate the first and third quartiles.

3.2. Event‐related potential

In both single‐source and multiple‐source situations, the onset of sound stimulation elicited a significant response at the FCz electrode, that is, the N1 component of the prefrontal lobe negative and the P2 component of the central positive, as shown in Figure 4a. The amplitude and latency of the N1 and P2 components were quantitatively analyzed, as shown in Figure 4b,c. Specifically, the absolute amplitude of the N1 component in single‐source situations (0.98 ± 0.49 μV) was significantly greater than that of multiple‐source situations (0.41 ± 0.28 μV) (t (25) = 4.163, p < .001). Similarly, the amplitude of the P2 component in single‐source situations (2.71 ± 0.87 μV) was significantly greater than that of multiple‐source situations (1.79 ± 0.66 μV) (t (25) = 4.586, p < .001). Although the latency of the N1 and P2 components in single‐source situations (140 ± 22.6 ms and 226 ± 27.0 ms) was longer than that of multiple‐source situations (132 ± 22.2 ms and 218 ± 20.8 ms), the differences were not significant (t (25) = 1.241, p = .226 and t (25) = 0.601, p = .553, respectively).

FIGURE 4.

FIGURE 4

Open in a new tab

ERP elicited by single‐source and multiple‐source situations. (a) ERP waveform at the FCz electrode and topology of the N1 and P2 for single‐source (top panel) and multiple‐source (bottom panel) situations, (b) The absolute amplitude of the N1 and P2 components, and (c) The latency of the N1 and P2 components. (*** p < .001).

3.3. Microstate sequences

The optimal clustering analysis of the microstate templates was conducted by averaging the total ERP data across all participants in both single‐source and multiple‐source situations. Consequently, this analysis yielded five optimal equivalent topographies. The optimal microstate templates, MS1, MS2, MS3, MS4, and MS5, are illustrated in Figure 5a. It can be seen that MS1 exhibited positive activation in the frontal lobe and negative activation in the parietal‐occipital lobes. MS2 displayed positive activation in the parietal‐occipital lobes. MS3 showed positive activation in the frontal lobe. MS4 demonstrated positive activation in the left and right temporal–parietal lobes. MS5 exhibited negative activation in the frontal lobe and positive activation in the parietal‐occipital lobe.

FIGURE 5.

FIGURE 5

Open in a new tab

Microstate templates and sequences of single‐source and multiple‐source situations. (a) The optimal clustering template. (b) Microstate sequences in single‐source situations. (c) Microstate sequences in multiple‐source situations.

The optimal clustering template was applied independently to single‐source and multiple‐source situations, as shown in Figure 5b,c, respectively. The findings indicated that the same microstate sequences, ranging from MS1 to MS5, were observed in both single‐source and multiple‐source situations. Nevertheless, it can be observed that distinctions exist in the properties of some microstates for single‐source and multiple‐source situations. These distinctions will be elucidated in greater detail in the subsequent section.

3.4. Microstate properties

The clustering template was applied independently to each participant in both the single‐source and multiple‐source situations. Each microstate parameter of the single‐source and multiple‐source situations was quantitatively analyzed, including GEV, GFP, duration, and coverage, as shown in Figure 6.

FIGURE 6.

FIGURE 6

Open in a new tab

Microstate properties for single‐source and multiple‐source situations. (a) GEV, (b) GFP, (c) Duration, and (d) Coverage. Error bars indicate standard deviation and only the significance between different situations was marked for simplicity. (*** p < .001, ** p < .01 and * p < .05).

Figure 6a shows the GEV of each microstate in both single‐source and multiple‐source situations. ANOVA revealed a main effect of condition on the GEV (F(1,25) = 7.153, p = .013). Bonferroni's post hoc analysis indicated that the GEV of MS3 was significantly higher in single‐source situations than in multiple‐source situations (t (100) = 2.972, p = .004). Additionally, ANOVA revealed a main effect of microstate class on the GEV (F(4,100) = 20.97, p < .001). Bonferroni's post hoc analysis showed that the GEV of MS3 and MS5 was significantly higher than that of MS1, MS2, and MS4 in both single‐source and multiple‐source situations (for all t (100) ≥ 3.838, p ≤ .002). Moreover, the GEV of MS5 was significantly higher than that of MS3 in multiple‐source situations (t (100) = 3.692, p = .004).

Figure 6b shows the GFP of each microstate in both single‐source and multiple‐source situations. ANOVA revealed a main effect of the condition on GFP (F(1,25) = 14.85, p < .001). Bonferroni's post hoc analysis indicated that the GFP of MS2 was significantly higher in single‐source situations than in multiple‐source situations (t (100) = 3.091, p = .003). Additionally, ANOVA revealed a main effect of microstate class on the GFP (F(4,100) = 31.13, p < .001). Bonferroni's post hoc analysis showed that the GFP of MS3 and MS5 was significantly higher than that of MS2 and MS4 in both single‐source and multiple‐source situations (for all t (100) ≥ 3.319, p ≤ .012). Additionally, the GFP of MS1 was significantly higher than that of MS2 in multiple‐source situations (t (100) = 4.577, p < .001).

Figure 6c shows the duration of each microstate in both single‐source and multiple‐source situations. ANOVA revealed a main effect of condition on duration (F(1,25) = 17.60, p < .001). Bonferroni's post hoc analysis indicated that the durations of MS2 and MS4 were significantly longer in single‐source situations than in multiple‐source situations (t (100) = 2.349, p = .021 and t (100) = 2.577, p = .011, respectively). In contrast, the duration of MS5 was significantly shorter in single‐source situations than in multiple‐source situations (t (100) = 4.350, p < .001). Additionally, ANOVA revealed a main effect of microstate class on duration (F(4,100) = 11.78, p < .001). Bonferroni's post hoc analysis showed that the duration of MS5 was significantly higher than that of the other classes in multiple‐source situations (for all t (100) ≥ 4.450, p < .001). Conversely, the duration of MS2 was significantly shorter than that of MS1, MS3, and MS4 in multiple‐source situations (for all t (100) ≥ 3.604, p ≤ .005).

Figure 6d shows the coverage of each microstate in both single‐source and multiple‐source situations. ANOVA revealed a main effect of condition on coverage (F(1,25) = 27.83, p < .001). Bonferroni's post hoc analysis indicated that the coverage of MS4 was significantly higher in single‐source situations than in multiple‐source situations (t (100) = 2.383, p = .019). Additionally, ANOVA revealed a main effect of microstate class on coverage (F(4,100) = 16.47, p < .001). Bonferroni's post hoc analysis showed that the coverage of MS3 and MS5 was significantly higher than that of MS2 and MS4 in both single‐source and multiple‐source situations (for all t (100) = 3.571, p ≤ .006). Additionally, the coverage of MS1 was significantly higher than that of MS2 in both single‐source and multiple‐source situations (t (100) = 3.561, p = .006 and t (100) = 3.544, p = .006, respectively).

3.5. Correlations between ERP and microstates

Figure 7a,b illustrate that the topology of the five microstates was comparable in both single‐source and multiple‐source situations. Particularly, MS1 and MS3 primarily activated the prefrontal and frontal cortices, similar to the topology of the P2 component. MS2 and MS5 mainly activated the parietal‐occipital cortex, which is similar to the topology of the N1 component, and MS4 activated the left and right temporal–parietal cortices. Pearson correlation coefficients between the five microstates and the N1 and P2 components further confirmed these correlations, as shown in Figure 7c. The positive correlation between MS2 and the N1 components was the highest in both single‐source (r = 0.90, p < .001) and multiple‐source (r = 0.86, p < .001) situations. Conversely, the positive correlation between MS3 and the P2 component was the highest in single‐source (r = .95, p < .001) and multiple‐source (r = 0.78, p < .001) situations.

FIGURE 7.

FIGURE 7

Open in a new tab

Correlations between ERP and microstates in both single‐source and multiple‐source situations. (a) Topology of five microstates for single‐source situations, (b) Topology of five microstates for multiple‐source situations, (c) Pearson correlation coefficients between each microstate and the N1 and P2 components. S‐N1 and S‐P2 refer to N1 and P2, respectively, in single‐source situations. M‐N1 and M‐P2 refer to N1 and P2, respectively, in multiple‐source situations.

3.6. Cortical source of microstates

The cortical source activity regions for each microstate were analyzed by comparing them to the 100 ms period before stimulation, as shown in Figure 8. The maximum activation region of MS1 was localized to the right temporal cortex, mainly including the STG (Brodmann 22), middle temporal gyrus (Brodmann 21), inferior temporal gyrus (Brodmann 20 and 37), and fusiform gyrus (Brodmann 37). The maximum activation region of MS2 shifted to the temporal–parietal cortex, mainly involving the inferior parietal lobule (IPL) (Brodmann 40), STG (Brodmann 22 and 42) and insula (Brodmann 13). The maximum activation region of MS3 moved to the limbic, mainly encompassing the cingulate gyrus (Brodmann 23 and 31), posterior cingulate (Brodmann 23), and insula (Brodmann 13). The maximum activation region of MS4 was localized in the frontal cortex, with the involvement of the superior frontal gyrus (SFG) (Brodmann 6 and 8), middle frontal gyrus (Brodmann 6), and medial frontal gyrus (Brodmann 8). The maximum activation region of MS5 was localized in the sub‐lobar and frontal cortex, primarily including the insula (Brodmann 13), SFG (Brodmann 8), and medial frontal gyrus (Brodmann 6 and 9).

FIGURE 8.

FIGURE 8

Open in a new tab

sLORETA images of the cortical source activity regions in five microstates. (a) MS1, (b) MS2, (c) MS3, (d) MS4, and (e) MS5. Color coding indicates the t‐value. R, L, A, and P indicate the right, left, anterior, and posterior sides of the brain, respectively.

Furthermore, the differences in the regions of cortical source activity for MS2 and MS3 between single‐source and multiple‐source situations were analyzed, as these microstates exhibited the strongest correlation with the N1 and P2 components. Specifically, the differences in cortical source activity were obtained by subtracting multiple‐source situations from single‐source situations, as shown in Figure 9. For MS2, greater activation was observed in the single‐source situations compared to the multiple‐source situations. Specifically, the region of maximum activation difference was localized in the temporal–parietal cortex, primarily including the STG (Brodmann 22 and 39) and IPL (Brodmann 40). For MS3, single‐source situations also exhibited greater activation than multiple‐source situations. Specifically, the region of maximum activation difference was localized in the right parietal cortex, mainly including the IPL (Brodmann 40) and supramarginal gyrus (Brodmann 40).

FIGURE 9.

FIGURE 9

Open in a new tab

sLORETA images of the cortical source activity difference regions for MS2 and MS3 between single‐source and multiple‐source situations. (a) MS2 and (b) MS3. Color coding indicates the t‐value, with warmer colors indicating greater activity in single‐source situations and cooler colors indicating greater activity in multiple‐source situations. R, L, A, P, S, and L indicate the right, left, anterior, posterior, superior, and inferior sides of the brain, respectively.

4. DISCUSSION

This study is the first to introduce ERP microstate analysis to reveal the electrophysiological correlation of ASSA in the “cocktail party” situation. We hypothesized that the “cocktail party” situation might influence the spatiotemporal dynamics of the electrophysiological activity associated with ASSA. To test this, we designed single‐source and multiple‐source paradigms to simulate different auditory environments and conducted ERP microstate analysis. First, the ERP signals elicited by the ASSA tasks were extracted and analyzed to investigate temporal correlation. Subsequently, the microstate sequence of the ERP signals was constructed and quantitatively measured to explore the spatiotemporal correlation. Finally, the cortical source activity regions of each microstate and the differential activity regions of specific microstates were revealed using sLORETA.

A multitude of studies has demonstrated that the N1 and P2 components are functional responses of distinct neural generators and are employed to assess the neural mechanisms of human cortical auditory spatial processing (Ross & Tremblay, 2009; Shahin et al., 2005). Therefore, we focused on investigating the N1 and P2 components of the ERP signals. The findings demonstrated that ASSA tasks could elicit significant N1 and P2 components in both single‐source and multiple‐source situations. A more detailed analysis revealed that the absolute amplitudes of the N1 and P2 components were lower in multiple‐source situations compared to single‐source situations, consistent with previous studies (Lewald & Getzmann, 2015; Liu et al., 2024). Inspired by previous knowledge and our findings, we speculated that this phenomenon may be caused by contralateral inhibitory effects in the auditory system (Ahlfors et al., 2015; Lewald et al., 2018). However, the ERP analysis can only observe significant N1 and P2 components and cannot fully reflect the spatiotemporal correlation of ASSA in the “cocktail party” situation.

Furthermore, ERP microstate analysis was introduced to investigate the spatiotemporal correlation of ASSA. The study found that five microstates can be used to explain the spatiotemporal dynamics of ASSA, which revealed more potential components than traditional ERP analysis. Specifically, MS1 exhibited strong activation in the frontal cortex, similar to microstate D during the resting state. MS2 and MS3 showed positive activation in the parietal‐occipital and frontal cortices, respectively, and were highly correlated with the N1 and P2 components. MS4 exhibited positive activation in the left and right temporal–parietal cortices, whereas MS5 showed negative activation in the frontal cortex and positive activation in the parietal‐occipital cortex. In addition, research has demonstrated that auditory cognitive processes encompass three distinct stages: the perception stage (110–140 ms), the recognition stage (260–320 ms), and the execution stage (500–700 ms) (Yu et al., 2014). Figure 5 illustrates that MS1 and MS2 were predominantly covered in the perception stage, MS3 and MS4 in the recognition stage, and MS5 in the execution stage.

Cortical source analysis revealed that the maximum activation regions of MS1 were mainly localized in the STG, middle temporal gyrus, inferior temporal gyrus, and fusiform gyrus. Some similar cortical regions have been demonstrated to play a crucial role in spatial analysis, as evidenced by fMRI studies (Zündorf et al., 2013, 2016). Additionally, studies have shown that the N1 period activation is primarily located in the temporal–parietal region, with the maximum difference found between tasks in the IPL (Lewald & Getzmann, 2011). The N1 period activation exhibited partial similarity to MS2 neural activation, with the maximum activation observed in the IPL, STG, and insula. Moreover, studies have shown that the P2 period activation is primarily located in the bilateral cingulate and left frontoparietal cortex, which partially aligns with MS3 neural activation, with the maximum activation in the cingulate gyrus, posterior cingulate, and insula (Lewald & Getzmann, 2015). MS4 and MS5 exhibited the greatest activation in the frontal lobe region, mainly including the SFG, middle frontal gyrus, and medial frontal gyrus. This finding is consistent with fMRI studies indicating the frontal lobe's essential role in spatial analysis (Arnott et al., 2004).

From the perspective of spatiotemporal sequences, it was found that microstate sequences ranged from MS1 to MS5 in both single‐source and multiple‐source situations. By combining the neural activation regions of each microstate, it was observed that activation regions initially transferred from the right temporal cortex to the temporal–parietal cortex, that is, IPL and STG, then to the cingulate cortex, and finally to the dorsofrontal cortex, including the SFG and middle frontal gyrus. Research has demonstrated that auditory spatial and non‐spatial information is processed through specialized pathways, known as the auditory dual‐pathway model (Brunetti et al., 2005; Rauschecker & Tian, 2000). Among these, the posterodorsal auditory stream was particularly adept at processing auditory spatial information, with the posterior parts of the auditory cortex, posterior STG, IPL, and superior frontal sulcus playing a pivotal role (Ahveninen et al., 2014; Zündorf et al., 2016). Beyond the aforementioned regions, the planum temporale was identified as a shared auditory region for sound localization and sound identification processes (Arnott et al., 2004). Interestingly, the spatiotemporal dynamics revealed by ERP microstate analysis roughly align with the posterodorsal auditory stream identified by fMRI and MEG.

Although the microstates in single‐source and multiple‐source situations follow the same spatiotemporal sequences, there were notable differences in the properties of some microstates. Specifically, the GFP and duration of MS2 in single‐source situations were greater than in multiple‐source situations, potentially related to the significant decrease in the N1 component owing to extensive sensory processing (Lewald & Getzmann, 2015). The GEV of MS3 was higher in single‐source situations than in multiple‐source situations, possibly owing to a significant decrease in the P2 component. MS4 exhibited a shorter duration and coverage in multiple‐source situations compared to single‐source situations. Based on prior knowledge and related studies, we speculated that this likely reflects increased cognitive load during the target recognition stage (Li et al., 2016; Yu et al., 2014). In contrast, MS5 showed a longer duration in multiple‐source situations compared to single‐source situations, suggesting frequent transitions among the initial four microstates in complex environments.

The differences between each microstate property in both single‐source and multiple‐source situations were further analyzed quantitatively, as different microstates represent different cognitive components (Murray et al., 2008). Significant differences were observed between some microstate properties in single‐source and multiple‐source situations, mainly reflected in MS3 and MS5 exhibited higher microstate properties, whereas MS2 and MS4 demonstrated lower microstate properties. Specifically, across both single‐source and multiple‐source situations, MS3 and MS5 showed significantly higher GEV, GFP, and coverage compared to MS2 and MS4. Additionally, MS5 had a longer duration than MS2 and MS4, whereas MS3's duration was significantly higher than that of MS2 only in multiple‐source situations. Based on prior knowledge and related research, we speculated that the higher value of MS3 may be related to the significant P2 component caused by attentional allocation (Potts, 2004), whereas the higher values of MS5 might be associated with slow‐wave components originating from the frontal lobe (Barry et al., 2020).

Furthermore, distinct activation regions of MS2 and MS3 were revealed between single‐source and multiple‐source situations, as these microstates exhibited the highest correlation with the N1 and P2 components. The cortical source activation levels of MS2 and MS3 tended to be lower in multiple‐source situations compared to single‐source situations, which is consistent with the results obtained at N1 and P2 moments from relevant studies (Lewald & Getzmann, 2015). This phenomenon potentially owes to heightened processing demands involving complex binaural spatial cues, that is, interaural time and level differences, at the perception and recognition stages. Specifically, MS2 exhibited the maximum activation differences in the temporal–parietal cortex, mainly including regions such as the STG and IPL. In contrast, MS3 showed maximum activation differences in the right parietal cortex, mainly encompassing the IPL and supramarginal gyrus. Previous human fMRI and MEG studies have consistently underscored the critical roles of these regions, particularly the STG and IPL, in selective auditory attention (Zündorf et al., 2013 and 2014; Hamilton et al., 2021). In summary, these findings may reflect neural processes specific in complex auditory environments and suggest that the STG and IPL play a pivotal role in modulating ASSA.

However, despite the interesting findings of the current study, some limitations should be considered. First, the identification of five optimal microstate topologies relies on prior knowledge and predefined criteria. Second, the EEG‐based cortical source analysis employed a universal MRI template, and this method was still insensitive to weak or deep sources. Additionally, other cognitive components, such as executive control, may also potentially influence ASSA. Future research should refine methods for determining optimal clustering templates, validate findings using individual MRI scans combined with fMRI or MEG, and explore ASSA's relationship with other cognitive components through functional brain networks. Further investigation into ASSA among individuals with hearing impairments may also provide insights into their electrophysiological correlates, providing a theoretical reference for improving hearing aids or cochlear implants.

5. CONCLUSION

This study explored the electrophysiological correlates of ASSA in the “cocktail party” situation using ERP microstate analysis. Under the premise that microstates clustered into five categories and microstate sequences ranged from MS1 to MS5 in both single‐source and multiple‐source situations. Notably, MS2 and MS3 corresponded to the N1 and P2 components, respectively, and the partial properties of these microstates were significantly lower in multiple‐source situations than in single‐source situations. Moreover, the durations of MS4 and MS5 were significantly shorter and longer, respectively, in multiple‐source situations than in single‐source situations. Cortical source analysis showed that the activation regions of these microstates initially transferred from the right temporal cortex to the temporal–parietal cortex, and then to the dorsofrontal cortex. Moreover, greater neural activation was observed in single‐source situations compared to multiple‐source situations in MS2 and MS3, with maximal differences noted in the STG and IPL. In summary, our study offers a novel insight into the electrophysiological correlates of ASSA in the “cocktail party” situation.

AUTHOR CONTRIBUTIONS

Hongxing Liu: Conceptualization, methodology, formal analysis, writing‐review & editing. Yanru Bai: Conceptualization, methodology, writing‐review & editing. Qi Zheng: Data curation. Jihan Liu: Formal analysis. Jianing Zhu: Data curation. Guangjian Ni: Conceptualization, methodology, writing‐review & editing.

FUNDING INFORMATION

This work was supported by projects from the National Key Research and Development Program of China (No. 2023YFF1203500), the National Natural Science Foundation of China (No. 81971698), and the Tianjin Research Innovation Project for Postgraduate Students (No. 2022BKY056).

CONFLICT OF INTEREST STATEMENT

The authors declare no competing interests.

ACKNOWLEDGMENTS

We would like to thank all the participants for their cooperation in this study.

Liu, H. , Bai, Y. , Zheng, Q. , Liu, J. , Zhu, J. , & Ni, G. (2024). Electrophysiological correlation of auditory selective spatial attention in the “cocktail party” situation. Human Brain Mapping, 45(11), e26793. 10.1002/hbm.26793

Hongxing Liu and Yanru Bai contributed equally to this work.

DATA AVAILABILITY STATEMENT

EEG data analyses are performed in MATLAB and the freely available toolbox EEGLAB. The software code and original data supporting this study's findings are available from the corresponding authors.

REFERENCES

  1. Ahlfors, S. P. , Jones, S. R. , Ahveninen, J. , Hämäläinen, M. S. , Belliveau, J. W. , & Bar, M. (2015). Direction of magnetoencephalography sources associated with feedback and feedforward contributions in a visual object recognition task. Neuroscience Letters, 585, 149–154. 10.1016/j.neulet.2014.11.029 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Ahveninen, J. , Kopčo, N. , & Jääskeläinen, I. P. (2014). Psychophysics and neuronal bases of sound localization in humans. Hearing Research, 307, 86–97. 10.1016/j.heares.2013.07.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Arnott, S. R. , Binns, M. A. , Grady, C. L. , & Alain, C. (2004). Assessing the auditory dual‐pathway model in humans. NeuroImage, 22(1), 401–408. 10.1016/j.neuroimage.2004.01.014 [DOI] [PubMed] [Google Scholar]
  4. Barry, R. J. , Steiner, G. Z. , De Blasio, F. M. , Fogarty, J. S. , Karamacoska, D. , & MacDonald, B. (2020). Components in the P300: Don't forget the novelty P3! Psychophysiology, 57, e13371. 10.1111/psyp.13371 [DOI] [PubMed] [Google Scholar]
  5. Bharadwaj, H. M. , Lee, A. K. C. , & Shinn‐Cunningham, B. G. (2014). Measuring auditory selective attention using frequency tagging. Frontiers in Integrative Neuroscience, 8, 1–12. 10.3389/fnint.2014.00006 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bradley, A. , Yao, J. , Dewald, J. , & Richter, C. P. (2016). Evaluation of electroencephalography source localization algorithms with multiple cortical sources. Plos One, 11(1), e0147266. 10.1371/journal.pone.0147266 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Brainard, D. H. (1997). The psychophysics toolbox. Spatial Vision, 10, 433–436. 10.1163/156856897x00357 [DOI] [PubMed] [Google Scholar]
  8. Bréchet, L. , Brunet, D. , Birot, G. , Gruetter, R. , Michel, C. M. , & Jorge, J. (2019). Capturing the spatiotemporal dynamics of self‐generated, task‐initiated thoughts with EEG and fMRI. NeuroImage, 194, 82–92. 10.1016/j.neuroimage.2019.03.029 [DOI] [PubMed] [Google Scholar]
  9. Britz, J. , Van De Ville, D. , & Michel, C. M. (2010). BOLD correlates of EEG topography reveal rapid resting‐state network dynamics. NeuroImage, 52(4), 1162–1170. 10.1016/j.neuroimage.2010.02.052 [DOI] [PubMed] [Google Scholar]
  10. Brunet, D. , Murray, M. M. , & Michel, C. M. (2011). Spatiotemporal analysis of multichannel EEG: CARTOOL. Computational Intelligence and Neuroscience, 2011, 813870. 10.1155/2011/813870 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Brunetti, M. , Belardinelli, P. , Caulo, M. , del Gratta, C. , Della Penna, S. , Ferretti, A. , Lucci, G. , Moretti, A. , Pizzella, V. , Tartaro, A. , Torquati, K. , Olivetti Belardinelli, M. , & Romani, G. L. (2005). Human brain activation during passive listening to sounds from different locations: An fMRI and MEG study. Human Brain Mapping, 26(4), 251–261. 10.1002/hbm.20164 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Chen, J. , Ke, Y. , Ni, G. , Liu, S. , & Ming, D. (2023). Evidence for modulation of EEG microstates by mental workload levels and task types. Human Brain Mapping, 45(1), e26552. 10.1002/hbm.26552 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Cherry, E. C. (1953). Some experiments on the recognition of speech, with one and with two ears. The Journal of the Acoustical Society of America, 25(5), 975–979. 10.1121/1.1907229 [DOI] [Google Scholar]
  14. Dering, B. , & Donaldson, D. I. (2016). Dissociating attention effects from categorical perception with ERP functional microstates. Plos One, 11(9), e0163336. 10.1371/journal.pone.0163336 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Di Russo, F. , Berchicci, M. , Bozzacchi, C. , Perri, R. L. , Pitzalis, S. , & Spinelli, D. (2017). Beyond the “Bereitschaftspotential”: Action preparation behind cognitive functions. Neuroscience Biobehavioral Reviews, 78, 57–81. 10.1016/j.neubiorev.2017.04.019 [DOI] [PubMed] [Google Scholar]
  16. Ding, N. , & Simon, J. Z. (2012a). Emergence of neural encoding of auditory objects while listening to competing speakers. Proceedings of the National Academy of Sciences, 109(29), 11854–11859. 10.1073/pnas.1205381109 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Ding, N. , & Simon, J. Z. (2012b). Neural coding of continuous speech in auditory cortex during monaural and dichotic listening. Journal of Neurophysiology, 107, 78–89. 10.1152/jn.00297.2011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Gamble, M. L. , & Luck, S. J. (2011). N2ac: An ERP component associated with the focusing of attention within an auditory scene. Psychophysiology, 48(8), 1057–1068. 10.1111/j.1469-8986.2010.01172.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Grothe, B. , Pecka, M. , & McAlpine, D. (2010). Mechanisms of sound localization in mammals. Physiological Reviews, 90, 983–1012. 10.1152/physrev.00026.2009 [DOI] [PubMed] [Google Scholar]
  20. Hamilton, L. S. , Oganian, Y. , Hall, J. , & Chang, E. F. (2021). Parallel and distributed encoding of speech across human auditory cortex. Cell, 184(18), 4626–4639. 10.1016/j.cell.2021.07.019 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Hillyard, S. A. , Hink, R. F. , Schwent, V. L. , & Picton, T. W. (1973). Electrical signs of selective attention in the human brain. Science, 182(4108), 177–180. 10.1126/science.182.4108.177 [DOI] [PubMed] [Google Scholar]
  22. Hofmann‐Shen, C. , Vogel, B. O. , Kaffes, M. , Rudolph, A. , Brown, E. C. , Tas, C. , Brüne, M. , & Neuhaus, A. H. (2020). Mapping adaptation, deviance detection, and prediction error in auditory processing. NeuroImage, 207(116), 116432. 10.1016/j.neuroimage.2019.116432 [DOI] [PubMed] [Google Scholar]
  23. Kaufman, M. , & Golumbic, E. (2023). Listening to two speakers: Capacity and tradeoffs in neural speech tracking during selective and distributed attention. Neuroimage, 270, 119984. 10.1016/j.neuroimage.2023.119984 [DOI] [PubMed] [Google Scholar]
  24. Khanna, A. , Pascual‐Leone, A. , Michel, C. M. , & Farzan, F. (2015). Microstates in resting‐state EEG: Current status and future directions. Neuroscience Biobehavioral Reviews, 49, 105–113. 10.1016/j.neubiorev.2014.12.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Kong, W. , Wang, L. , Zhang, J. , Zhao, Q. , & Sun, J. (2018). The dynamic EEG microstates in mental rotation. Sensors, 18(9), 2920–2936. 10.3390/s18092920 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Lehmann, D. , Faber, P. L. , Galderisi, S. , Herrmann, W. M. , Kinoshita, T. , Koukkou, M. , Mucci, A. , Pascual‐Marqui, R. D. , Saito, N. , Wackermann, J. , Winterer, G. , & Koenig, T. (2005). EEG microstate duration and syntax in acute, medication‐naïve, first‐episode schizophrenia: A multi‐center study. Psychiatry Research: Neuroimaging, 138, 141–156. 10.1016/j.pscychresns.2004.05.007 [DOI] [PubMed] [Google Scholar]
  27. Lewald, J. , & Getzmann, S. (2011). When and where of auditory spatial processing in cortex: A novel approach using electrotomography. PLos One, 6(9), e25146. 10.1371/journal.pone.0025146 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Lewald, J. , & Getzmann, S. (2015). Electrophysiological correlates of cocktail‐party listening. Behavioural Brain Research, 292, 157–166. 10.1016/j.bbr.2015.06.025 [DOI] [PubMed] [Google Scholar]
  29. Lewald, J. , Hanenberg, C. , & Getzmann, S. (2016). Brain correlates of the orientation of auditory spatial attention onto speaker location in a “cocktail‐party” situation. Psychophysiology, 53(10), 1484–1495. 10.1111/psyp.12692 [DOI] [PubMed] [Google Scholar]
  30. Lewald, J. , Schlüter, M. C. , & Getzmann, S. (2018). Cortical processing of location changes in a “cocktail‐party” situation: Spatial oddball effects on electrophysiological correlates of auditory selective attention. Hearing Research, 365, 49–61. 10.1016/j.heares.2018.04.009 [DOI] [PubMed] [Google Scholar]
  31. Li, F. , Chen, B. , Li, H. , Zhang, T. , Wang, F. , Jiang, Y. , Li, P. , Ma, T. , Zhang, R. , Tian, Y. , Liu, T. , Guo, D. , Yao, D. , & Xu, P. (2016). The time‐varying networks in P300: A task‐evoked EEG study. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 24(7), 725–733. 10.1109/tnsre.2016.2523678 [DOI] [PubMed] [Google Scholar]
  32. Liu, H. X. , Bai, Y. R. , Xu, Z. H. , Liu, J. H. , Ni, G. J. , & Ming, D. (2024). The scalp time‐varying network of auditory spatial attention in “cocktail‐party” situations. Hearing Research, 442, 108946. 10.1016/j.heares.2023.108946 [DOI] [PubMed] [Google Scholar]
  33. Mesgarani, N. , & Chang, E. F. (2012). Selective cortical representation of attended speaker in multi‐talker speech perception. Nature, 485(7397), 233–236. 10.1038/nature11020 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Metzger, M. , Dukic, S. , McMackin, R. , Giglia, E. , Mitchell, M. , Bista, S. , Costello, E. , Peelo, C. , Tadjine, Y. , Sirenko, V. , Plaitano, S. , Coffey, A. , McManus, L. , Farnell Sharp, A. , Mehra, P. , Heverin, M. , Bede, P. , Muthuraman, M. , Pender, N. , … Nasseroleslami, B. (2023). Functional network dynamics revealed by EEG microstates reflect cognitive decline in amyotrophic lateral sclerosis. Human Brain Mapping, 45, e26536. 10.1002/hbm.26536 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Michel, C. M. , & Koenig, T. (2018). EEG microstates as a tool for studying the temporal dynamics of whole‐brain neuronal networks: A review. NeuroImage, 180, 577–593. 10.1016/j.neuroimage.2017.11.062 [DOI] [PubMed] [Google Scholar]
  36. Mishra, A. , Englitz, B. , & Cohen, M. X. (2020). EEG microstates as a continuous phenomenon. NeuroImage, 208(116), 454. 10.1016/j.neuroimage.2019.116454 [DOI] [PubMed] [Google Scholar]
  37. Murray, M. M. , Brunet, D. , & Michel, C. M. (2008). Topographic ERP analyses: A step‐by‐step tutorial review. Brain Topography, 20(4), 249–264. 10.1007/s10548-008-0054-5 [DOI] [PubMed] [Google Scholar]
  38. Nichols, T. E. , & Holmes, A. P. (2002). Nonparametric permutation tests for functional neuroimaging: A primer with examples. Human Brain Mapping, 15, 1–25. 10.1002/hbm.1058 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Nishida, K. , Morishima, Y. , Yoshimura, M. , Isotani, T. , Irisawa, S. , Jann, K. , Dierks, T. , Strik, W. , Kinoshita, T. , & Koenig, T. (2013). EEG microstates associated with salience and frontoparietal networks in frontotemporal dementia, schizophrenia and Alzheimer's disease. Clinical Neurophysiology, 124(6), 1106–1114. 10.1016/j.clinph.2013.01.005 [DOI] [PubMed] [Google Scholar]
  40. Pascual‐Marqui, R. D. (2002). Standardized low‐resolution brain electromagnetic tomography (sLORETA): Technical details. Methods and Findings in Experimental and Clinical Pharmacology, 24D, 5–12. [PubMed] [Google Scholar]
  41. Potts, G. F. (2004). An ERP index of task relevance evaluation of visual stimuli. Brain and Cognition, 56, 5–13. 10.1016/j.bandc.2004.03.006 [DOI] [PubMed] [Google Scholar]
  42. Rauschecker, J. P. , & Tian, B. (2000). Mechanisms and streams for processing of “what” and “where” in auditory cortex. Proceedings of the National Academy of Sciences, 97(22), 11800–11806. 10.1073/pnas.97.22.11800 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Ross, B. , & Tremblay, K. (2009). Stimulus experience modifies auditory neuromagnetic responses in young and older listeners. Hearing Research, 248, 48–59. 10.1016/j.heares.2008.11.012 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Shahin, A. , Roberts, L. E. , Pantev, C. , Trainor, L. J. , & Ross, B. (2005). Modulation of P2 auditory‐evoked responses by the spectral complexity of musical sounds. NeuroReport, 16(16), 1781–1785. 10.1097/01.wnr.0000185017.29316.63 [DOI] [PubMed] [Google Scholar]
  45. Si, Y. , Wu, X. , Li, F. , Zhang, L. , Duan, K. , Li, P. , & Xu, P. (2019). Different decision‐making responses occupy different brain networks for information processing: A study based on EEG and TMS. Cerebral Cortex, 29(10), 4119–4129. 10.1093/cercor/bhy294 [DOI] [PubMed] [Google Scholar]
  46. Tarailis, P. , Koenig, T. , Michel, C. M. , & Griškova‐Bulanova, I. (2023). The functional aspects of resting EEG microstates: A systematic review. Brain Topography, 37(2), 181–217. 10.1007/s10548-023-00958-9 [DOI] [PubMed] [Google Scholar]
  47. Yu, B. , Li, H. F. , Wang, X. D. , Ma, L. , & Wang, X. D. (2014). Automatic brain cognitive control detection method. International Conference on Frontiers of Internet of Things 2014. 10.1049/cp.2014.1571 [DOI] [Google Scholar]
  48. Zündorf, I. C. , Karnath, H. O. , & Lewald, J. (2014). The effect of brain lesions on sound localization in complex acoustic environments. Brain, 137(5), 1410–1418. 10.1093/brain/awu044 [DOI] [PubMed] [Google Scholar]
  49. Zündorf, I. C. , Lewald, J. , & Karnath, H. O. (2013). Neural correlates of sound localization in complex acoustic environments. PLos One, 8(5), e64259. 10.1371/journal.pone.0064259 [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Zündorf, I. C. , Lewald, J. , & Karnath, H. O. (2016). Testing the dual‐pathway model for auditory processing in human cortex. NeuroImage, 124, 672–681. 10.1016/j.neuroimage.2015.09.026 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

EEG data analyses are performed in MATLAB and the freely available toolbox EEGLAB. The software code and original data supporting this study's findings are available from the corresponding authors.

Articles from Human Brain Mapping are provided here courtesy of Wiley

RESOURCES