Skip to main content
NCBI home page
Experimental Neurobiology logoLink to Experimental Neurobiology
. 2025 Nov 27;35(1):17–28. doi: 10.5607/en25028

Identification of Cortices with Characteristics of Rhythmic Entrainment and Its Periodicity

Youmin Shin 1,2,†, Jii Kwon 3,†, June Sic Kim 4, Chun Kee Chung 5,*
PMCID: PMC12977227  PMID: 41298224

Abstract

Listening to rhythmic patterns leads to neural entrainment to beat and meter periodicities. The debate over whether entrainment is a mere reflection of external stimuli, or an inherent intrinsic response persists. The objective of this study was to ascertain whether there are cerebral cortices, which satisfy 3 distinct features of intrinsic entrainment; first, the ability to sustain neural oscillations even in random beat omission; second, a requisite latency period for the build-up before initiating a response to rhythmic stimuli; and third, the persistence of these neural oscillations gradually recedes following the cessation of the stimulus. In 27 patients with medically intractable epilepsy, electrocorticography data were obtained with 2- or 3-beat sound stimulations with random omissions. We found that there are cortices which satisfy all three requirements of intrinsic entrainment. The cortices synchronized with beat were in Brodmann areas (BA) 21, and 22, whereas the cortices synchronized with meter corresponded to BA3, 6, 9, 22, 40, and 44. We showed that entrainment is an intrinsic response, with distinct neural processing for beat and meter. These insights advance our understanding of neural entrainment to beat and meter periodicities.

Keywords: Entrainment, Beat, Meter, Electrocorticography, Cortex

INTRODUCTION

In adults, listening to rhythmic patterns induces neural entrainment to both beat [1, 2] and meter [3-6] periodicities. The beat, serving as the basic unit of time in music, forms the backbone of rhythm through a repeating sequence of stressed (strong) and unstressed (weak) beats [7]. This regular pulse forms the foundation upon which more complex rhythmic structures are built. In meter refers to the hierarchical grouping of beats into regular patterns marked by recurring accents, such as duple or triple groupings, which give rise to familiar rhythmic frameworks like marches or waltzes [8].

Entrainment is a distinct dynamic process by which endogenous brain oscillations align temporally with external rhythmic stimuli. This active, reciprocal coupling enables not only the perception of rhythm but also its anticipation and generation in time [9, 10]. Importantly, beat and meter together constitute a multilayered temporal framework through which the brain encodes and interprets rhythm. While the beat supports immediate sensory tracking of temporal regularity, the meter enables more abstract, higher-order temporal organization.

Understanding how neural activity entrains to these nested rhythmic levels provides crucial insights into the brain’s mechanisms for temporal prediction, motor coordination, and auditory cognition—processes fundamental not only in music perception but also in speech, movement, and other time-dependent behaviors [11-13].

The primary indication of beat entrainment is the formation of neural oscillations synchronized with an external stimulus [14]. However, simple synchronization is not conclusive evidence of beat entrainment. Neural responses, such as auditory evoked potential (AEP), can be time-locked and synchronized with any sensory stimulus, not exclusively rhythmic ones. Although several studies have reported that mismatch negativity (MMN) can occur in the context of rhythmic stimulation and reflect sensitivity to stimulus regularity, it is not a reliable marker of entrainment [15-19]. This is because MMN can also be elicited by deviations in non-temporal acoustic features, such as pitch and timbre, not solely by disruptions in temporal patterns [20-22]. Therefore, MMN should be clearly distinguished from true beat entrainment.

A recent electroencephalography (EEG) study shed light on the concept of subjective rhythmizing [23]. Notable is the ‘build-up latency,’ which refers to the time period the brain requires to synchronize its neural oscillations with an external rhythm. While ‘build-up latency’ shares certain temporal characteristics with memory-based predictions, they differ notably in their duration of influence. Entrainment effects typically persist for over a second after stimulus offset, reflecting sustained internal engagement, whereas memory-based predictions are short-lived, often dissipating within one second [24-26].

This prolonged influence of beat entrainment, in combination with build-up latency, supports a top-down processing mechanism in which the brain actively organizes and modulates rhythmic input over time. Rather than passively mirroring external stimuli, the brain engages in temporally extended, internally driven modulation of rhythmic input. Such processing supports adaptive coordination and facilitates more efficient neural responses to structured auditory stimuli [27-31].

The objective of this study is to ascertain whether there are cortices, which satisfy all three requirements of entrainment as follows. 1) Entrainment should be maintained even if the stimulation is omitted randomly. This distinguishes entrainment from AEP, which vanish in the absence of stimuli, and from MMN, which emerges only in response to such omissions (see Supplementary Fig. 1); 2) entrainment should require a temporal delay to emerge, reflecting a non-instantaneous process of neural synchronization; and 3) once established, entrained oscillations should not terminate abruptly with the offset of stimulation, but rather recede progressively over time, reflecting a temporally sustained internal process that distinguishes entrainment from memory-based prediction. This persistence is generally defined as lasting for about one second or longer, distinguishing entrainment from short-lived, memory-based predictions. While our design does not allow direct confirmation of the 1-second threshold—since each epoch spans a full meter cycle of 2.1 seconds—oscillatory activity persisting for at least two cycles (i.e., >4.2 seconds) would represent even stronger evidence for sustained entrainment. This criterion thus serves as a conceptual distinction between entrained and transient rhythmic responses.

MATERIALS AND METHODS

Participants

Twenty-seven medically intractable epilepsy patients (10 males and 17 females; age (mean±standard deviation (SD)): 32.6±10.3 years) with intracranial electrodes implanted for localization of the epileptogenic zone participated in this study. Detailed information of participants including their diagnosis are provided in Supplementary Table 1. Prior to experiments, all patients were provided with sufficient information about the experiment. This study was carried out in accordance with recommendations of the Institutional Review Board of Seoul National University Hospital (H-1605-078-761). All subjects provided written informed consent in accordance with the Declaration of Helsinki.

Each participant completed one paradigm type (either 2-beat or 3-beat) depending on the recruitment period. Specifically, patients enrolled before February 2020 participated in the 2-beat paradigm, whereas those enrolled after February 2020 participated in the 3-beat paradigm. Thus, the two paradigms were assigned at the patient level rather than randomized within subjects, as summarized in Supplementary Table 1.

Experimental design

The paradigm of this experiment extended that of our previous study [19], with a detailed design illustrated in Fig. 1. Sound stimuli were presented through an insert earphone (Tip-300, Nicolet, Madison, WI, USA) under the control of STIM2 Gentask (Compumedics Neuroscan, Charlotte, NC, USA). We had to distinguish between two beats and three beats by sound intensity alone. Therefore, we assigned a 7 dB difference for down-beat (69 dB) and up/middle-beat (62 dB), which is easily distinguished by the human ear and generally comfortable to hear [32-35].

Fig. 1.

Fig. 1

Open in a new tab

Experimental design. (A, B) Illustrate the rhythmic stimulation paradigms for the 2-beat and 3-beat conditions, respectively. In the 2-beat paradigm, auditory tones were presented every 500 ms (beat frequency=2 Hz), forming a full meter cycle every 1,000 ms (meter frequency=1 Hz). In the 3-beat paradigm, tones occurred every 700 ms (beat frequency≈1.42 Hz), completing one meter cycle every 2,100 ms (meter frequency≈0.47 Hz). Solid lines represent the amplitude envelope of auditory tones (stimulus onsets), while dashed lines indicate temporal positions corresponding to inter-stimulus intervals or omitted beats. In the 2-beat condition, the down-beat was presented at 69 dB and the up-beat at 62 dB. In the 3-beat condition, the down-beat was presented at 69 dB, while the middle- and up-beats were presented at 62 dB, forming a hierarchical metrical accent structure (“down,” “middle,” and “up” beats corresponding to the first, second, and third positions within a cycle). (C, D) Show the distributions of randomly omitted beats in the 2-beat and 3-beat paradigms, respectively, used to evaluate sustained entrainment during omission epochs.

Auditory stimuli were pure tones of 262 Hz with a duration of 100 ms, with a 5 ms fade-in and fade-out. Each subject underwent eight sessions. Each session consisted of three sections: a pre-stimulation section, a stimulation section, and a post-stimulation section. In all sections, the subject was asked to look at a fixation cross on the monitor and to minimize movements. In a 2-beat experiment, we presented the beat sound in alternating strong and weak amplitude stimuli with an onset-to-onset interval of 500 ms (2 Hz). The interval between a down-beat and the next down-beat was 1,000 ms (1 Hz) (Fig. 1A). In a 3-beat experiment, we presented the beat sound in alternating strong, weak and weak amplitude stimuli with an onset-to-onset interval of 700 ms (1.42 Hz) (Fig. 1B). The interval between a strong stimulus and the next strong was 2,100 ms (0.47 Hz). In a 2-beat experiment, down-beat up-beat were each omitted 19 and 20 times per session, respectively (Fig. 1C). In a 3-beat experiment, down-beat, middle-beat, and up-beat were each omitted 15 times per session (Fig. 1D). To ensure that entrainment was firmly established before omission, stimuli were consistently presented for the first minute without any omissions [36]. The beat and meter frequencies described above (2 Hz and 1 Hz for the 2-beat, 1.42 Hz and 0.47 Hz for the 3-beat) were later used as region-of-interest (ROI) frequencies in further analyses.

ECoG recording

Subjects were implanted with subdural Electrocorticography (ECoG) depth electrodes (Ad-Tech Medical, Racine, WI, USA and PMT, Chanhassen, MN, USA) or high-density ECoG (PMT, Chanhassen, MN, USA). Subdural electrodes were of 3 mm in diameter with a 10 mm inter-electrode distance, while high-density ECoG were of 2 mm in diameter with a 5 mm inter-electrode distance. The ECoG signal was recorded with a 128-channel amplifier system (Compumedics Neuroscan, Charlotte, NC, USA). Signals were digitized at a sampling frequency of 1,000 Hz for a 2-beat experiment and 2,000 Hz for a 3-beat experiment. Preoperative magnetic resonance (MR) images were acquired using a Magnetom Trio Tim 3T scanner (Siemens, Erlangen, Germany) or Signa 1.5-T scanner (GE, Boston, MA, USA). Postoperative Computed tomography (CT) images were acquired using a Somatom sensation device (64 eco; Siemens München, Germany). For localization and co-registration of electrodes, preoperative MR data were co-registered to postoperative CT images using CURRY software (versions 7.0 and 8.0; Compumedics Neuroscan) to localize electrode locations of individual subjects. Through Talairach client (UTHSCSA, Texas, USA), we identified the Brodmann area to which each electrode belonged [37, 38]. We converted the Talairach coordinates to Montreal Neurological Institute (MNI) coordinates using a MATLAB script called ‘tal2mni.m’, which is part of the GingerALE software version 3.0 (www.brainmap.org) [39, 40]. The electrode locations were then mapped to the nearest point on the cortical surface using on Euclidean distance. To visualize the results, the team created a surface map showing the electrode locations within Brodmann areas defined in MNI coordinates using a python library, MNE [41]. Cortical distribution of the 650 subdural electrodes used for recording the ECoG data is illustrated in Supplementary Fig. 2.

Preprocessing

All data processing procedures were performed using MATLAB software version 2019a (MathWorks, Natick, MA, USA). For preprocessing, we removed bad electrodes which showed epileptiform activities or could not measure the signal due to technical problems including high impedance of electrode. Before analysis, common average reference was used to remove global background activity [42]. To remove 60 Hz line noise and synchronized harmonics, signals were notch-filtered.

DFT to identify ROI frequency

To identify the formation of a neural oscillation and determine the ROI frequency for given stimulus, we employed the discrete Fourier transform (DFT). Due to the differing lengths of the 2-beat and 3-beat stimuli, we set the analysis window to 5 seconds for the 2-beat and 10.5 seconds for the 3-beat, with an overlap of 50% between windows. For statistical analysis, we first assessed the normality of the data distribution. Depending on the result, either a paired t-test was performed for normally distributed data, or a Wilcoxon rank-sum test was applied for non-parametric comparisons.

AEP and MMN

To measure AEP and MMN data were epoched from -100 ms to 500 ms relative to the onset of the beat (0 ms). Baseline correction was performed by subtracting the mean signal within the -100 to 0 ms pre-stimulus interval from each trial. MMN was calculated by subtracting the averaged response to standard stimuli from that of deviant stimuli, after normalizing each waveform individually.

We identified the electrodes exhibiting the clearest AEP and MMN responses across all subjects. It is well established that AEPs are most prominent in the auditory cortex and are typically characterized by the P100 component [12, 43]. Accordingly, the electrode showing the largest P100 amplitude within the 70~130 ms was designated as the clearest AEP electrode. MMN primarily occurs in the frontal and temporal lobes within 150~250 ms [44, 45]. Therefore, the electrode with the greatest MMN amplitude in this interval was selected as the clearest MMN electrode. For the purpose of comparing the responses of entrainment with AEP/MMN, we defined AEP/MMN-specific electrodes based on their relative responsiveness. An electrode was classified as an AEP or MMN electrode if its response amplitude exceeded 50% of the maximum AEP or MMN response, respectively (Supplementary Figs. 3, 4).

Confirmation of maintained response

Epoching was performed from one downbeat to the next. To facilitate analysis of pre-stimulation and post-stimulation sections, hypothetical time points were defined and the same epoching procedure was applied as in the stimulation section (Fig. 1). For both the 2-beat and 3-beat paradigms, hypothetical time points in the pre-stimulus section were defined by mirroring the temporal structure of the subsequent stimulation sequence. Pre-stimulus epochs were divided into equal temporal bins corresponding to the expected beat and meter onsets (every 500 ms for the 2-beat paradigm and every 700 ms for the 3-beat paradigm), without actual auditory presentation. This procedure enabled direct comparison of spectral power changes between pre-stimulus and stimulation periods under identical temporal references. The stimulation section included randomly omitted auditory stimuli. To evaluate whether neural entrainment was maintained in the absence of auditory input, DFT analysis was performed on each epoch. Cortical areas were identified as entrained if they showed no statistically significant difference in spectral power at the corresponding frequencies across stimulus-present and stimulus-omitted trials (one-way ANOVA, p>0.05).

Change time point detection

To clarify the relationships among Figs. 2~5, all analyses were conducted sequentially on the same electrode dataset. Electrodes were first identified as showing frequency-specific oscillations synchronized with beat (2 Hz or 1.42 Hz) or meter (1 Hz or 0.47 Hz) stimuli (Fig. 2). Among these, electrodes maintaining oscillatory power during omission epochs were considered sustained entrainment sites (Fig. 3). A subset of these electrodes further showed build-up and recede latencies, indicating intrinsic entrainment dynamics (Fig. 4). Finally, electrodes that sequentially satisfied all criteria were anatomically localized and visualized (Fig. 5). To determine the build-up and recede latencies of neural entrainment, we analyzed temporal fluctuations in the DFT magnitude at the ROI frequency. Change point detection was performed to identify time points at which a significant transition in DFT power occurred, corresponding to the onset (build-up) or offset (recede) of neural entrainment. Specifically, we employed the ruptures library in Python (v3.9.12), a well-established tool for change point detection in time series data. Using a standard deviation (SD)-based cost function, we detected abrupt changes in the mean or variance of the DFT magnitude over time.

Fig. 2.

Fig. 2

Open in a new tab

Formation of neural oscillation that synchronizes with either beat or meter in electrodes synchronized with the specific frequencies. Fig. 2 demonstrates the distinctive results of DFT analyses for 2-beat and 3-beat experiments. (A, B) Detail the DFT outcomes for 2-beat experiment, while (C, D) display the results for the 3-beat experiment. Electrodes were classified based on the frequency at which significant synchronization was observed: beat frequencies (2 Hz for the 2-beat paradigm, 1.42 Hz for the 3-beat paradigm) and meter frequencies (1 Hz for the 2-beat paradigm, 0.47 Hz for the 3-beat paradigm). In the beat-synchronized electrodes (A, C), the DFT spectra of representative electrodes show clear amplitude increases at beat frequencies (2 Hz or 1.42 Hz), but not at the corresponding meter frequencies, indicating selective entrainment to beat-level periodicity. Conversely, in the meter-synchronized electrodes (B, D), power increases are observed at both meter and beat frequencies (e.g., 1 Hz and 2 Hz for 2-beat; 0.47 Hz and 1.42 Hz for 3-beat), suggesting hierarchical encoding of rhythmic structure across multiple temporal levels in these specific electrodes. All of plots in Fig. 2 highlight the significant difference of DFT magnitudes between the stimulation section and the pre-stimulation section (*p-value <0.01). DFT magnitudes for all epochs were plotted by transparency 0.2. The average of all epochs was plotted in a bold line. Electrodes showing frequency-specific oscillations were identified as the initial dataset for subsequent analyses (Figs. 3~5).

Fig. 3.

Fig. 3

Open in a new tab

Maintained response when a beat is omitted. Electrodes maintaining spectral power at beat or meter frequencies (mean p=0.41±0.18 across electrodes; one-way ANOVA, FDR-corrected) were considered to exhibit maintained entrainment. This reflects an absence of significant reduction in rhythmic synchronization during omitted beats, rather than strict statistical equivalence between conditions. However, there is a significant difference from the pre-stimulation section (one-way ANOVA p-value <0.01, paired T-test p-value <0.01). Detailed p-values are provided in Supplementary Table 2. Analyses were performed on the same electrode subset identified in Fig. 2, focusing on those maintaining oscillatory power during omission epochs.

Fig. 4.

Fig. 4

Open in a new tab

Requisite of latency to emerge and vanish. Fig. 4 compares representative temporal changes of entrainment (A), memory-based prediction (B), and AEP (C). Red lines represent the beginning and termination of the stimulation section. Blue lines represent the change point of ROI frequency’s DFT magnitude. If red and blue lines are overlaid, only red is visible. Vertical dotted lines indicate stimulus onset and offset. The bold waveform represents the average DFT magnitude across all epochs, while lighter thin lines correspond to single-trial data. An increase preceding the onset (build-up latency) and a gradual decrease after the offset (recede latency) indicate sustained rhythmic entrainment beyond stimulus presentation. A subset of electrodes from Figs. 2 and 3 showed build-up and recede latencies, indicating intrinsic entrainment dynamics.

Fig. 5.

Fig. 5

Open in a new tab

Cortices synchronized with beat and meter, respectively. Fig. 5 summarizes electrodes that exhibited all characteristics of neural entrainment. Electrodes showing frequency-specific synchronization with the beat (2 Hz for the 2-beat and 1.42 Hz for the 3-beat paradigm; p<0.05) and those synchronized with the meter (1 Hz and 0.47 Hz; p<0.05) are depicted in panels A and B, respectively. Panels C and D display electrodes that maintained oscillatory power during randomly omitted beats, indicating sustained activity in the absence of auditory input. Panels E and F show electrodes that exhibited both build-up and recede latencies in the DFT magnitude change, reflecting the emergence and disappearance of entrainment beyond stimulus presentation. Only electrodes that sequentially satisfied all three criteria—frequency-specific synchronization (Fig. 2), maintenance of oscillatory power during omissions (Fig. 3), and latency for both emergence and recession (Fig. 4)—are visualized here, representing the final set of entrained cortical sites. Electrodes visualized here correspond to those that sequentially satisfied all criteria shown in Figs. 2~4.

RESULTS

Formation of neural oscillations synchronizing with beat and meter

We performed a DFT analysis to identify significant changes in frequency magnitudes between the stimulation and pre-stimulation sections. The analysis revealed increases in frequencies corresponding to either beat or meter stimuli, confirming the formation of neural oscillations synchronized with these rhythmic components. Specifically, beat-synchronized frequencies increased at 2 Hz for the 2-beat paradigm (Fig. 2A) and 1.42 Hz for the 3-beat paradigm (Fig. 2C). Similarly, meter-synchronized frequencies increased at 1 Hz for the 2-beat paradigm (Fig. 2B) and 0.47 Hz for the 3-beat paradigm (Fig. 2D). These figures present representative examples of electrodes that showed significant synchronization to either beat or meter frequencies. The statistical significance of these frequency increases was assessed. To assess the statistical significance of these increases while accounting for multiple comparisons across electrodes within each subject, we applied false discovery rate correction (Benjamini-Hochberg).

The analysis revealed increases in spectral magnitude at both meter- synchronized frequencies (1 Hz for the 2-beat; 0.47 Hz for the 3-beat) and their harmonic components corresponding to the beat (2 Hz; 1.42 Hz). Because beat frequencies are integer harmonics of the meter frequencies, spectral peaks naturally appear at both, even when the underlying oscillation is driven by meter periodicity. Notably, the relative amplitude of the meter- synchronized peak was smaller than that of the beat- synchronized harmonic. This is consistent with the power-law (1/f) characteristic of cortical power spectra: lower-frequency bands have higher baseline power, which makes their relative power changes appear smaller after normalization. Both beat and meter frequencies nevertheless showed statistically significant increases (FDR-corrected p<0.05), confirming distinct yet harmonically synchronized rhythmic entrainment.

Maintained neural oscillation

Fig. 3 illustrates one of the core criteria for identifying beat entrainment (first requirement for entrainment). In this analysis, we focused on cortical regions that exhibited synchronization with either the beat (Fig. 3B, D) or the meter (Fig. 3A, C) in the 2-beat and 3-beat conditions, respectively. We observed that the increased DFT magnitude at the corresponding frequencies was sustained even when auditory stimuli were randomly omitted. During the stimulation phase, significant spectral synchronization was found compared with the pre-stimulation period (one-way ANOVA, p<0.05; one-tailed paired t-test, p<0.01). During omission epochs, the spectral power at the corresponding frequencies showed no significant reduction relative to the stimulation condition (full p-values in Supplementary Table 2). This result indicates that the entrained oscillations were not significantly diminished in the absence of auditory input, suggesting a maintained oscillatory process rather than an immediate stimulus-locked response. We note that this finding reflects a lack of significant reduction rather than strict statistical equivalence between the two conditions (Supplementary Fig. 5).

Latency periods for build-up and receding

Fig. 4 demonstrates the second and third requirements for beat entrainment. Within the group of electrodes that exhibited a maintained response, several electrodes displayed a latency to build-up or recede (Fig. 4A, B), rather than showing an immediate change following stimulus onset or offset. The vertical dotted lines indicate the time points of stimulus onset and offset, corresponding to the emergence and recession of entrained oscillations. The bold waveform represents the average DFT magnitude across all epochs, whereas the lighter traces denote single-trial data. A gradual increase in DFT magnitude before stimulus onset (build-up latency) and a delayed decrease after stimulus offset (recede latency) indicate that entrained oscillations evolve dynamically over time, persisting beyond the presence of external rhythmic input. In contrast, an immediate change was observed in the clearest AEP electrode (Fig. 4C), confirming that the latency-dependent responses reflect intrinsic entrainment rather than evoked potentials.

Cortices synchronized with beat and meter

Fig. 5 depicts cortical locations of electrodes that exhibit characteristics indicative of an inherent response to entrainment. The first step involved identifying electrodes that showed a significant change in DFT magnitude in synchronization with either the beat or the meter (Fig. 5A, B). Secondly, pinpointed electrodes that maintained oscillation, despite the random omission of stimuli, as depicted in Fig. 5C and 5D. Lastly, the final step was to identify electrodes exhibiting a build-up and recede latency, shown in Fig. 5E and 5F, which retained all characteristics of the intrinsic response to entrainment.

Beat-synchronized cortex corresponded to Brodmann areas (BA) 21 and 22. On the other hand, the meter-synchronized cortex exhibited a more distributed pattern across the cortex, including BA3, 6, 9, 22, 40, and 44. Notably, BA 22 was unique in that it corresponded to both beat-synchronized and meter-synchronized cortex, with no other areas showing such overlap. The precise locations of electrodes associated uniquely with beat- and meter-synchronized cortices are detailed in Supplementary Table 3 and 4, respectively.

DISCUSSION

There are cortices, satisfying all three requirements of entrainment as an intrinsic response, revealing distinct processing for beat and meter. The cortices synchronized with beat were BA 21, 22, and 47, whereas the cortices synchronized with meter corresponded to BA3, 6, 7, 9, 18, 20, 22, 28, 40, and 44. This widespread occurrence of the meter synchronized cortex is a novel observation, not distinctly highlighted in previous research, which often did not differentiate between beat and meter entrainment.

We employed two distinct stimuli in our study: one with an inter-beat interval of 500 ms to represent a 2-beat pattern, and another with a 700 ms interval for a 3-beat pattern. Beats 2 and 3 are considered the most foundational in rhythm perception, forming the core structure upon which more complex rhythms are built. Entrainment to these beats is a basic human capability, instinctive and intrinsic, requiring no explicit learning or training [46].

Fig. 2 depict neural oscillation synchronized with beat and meters. Notably, there is a significant increase in the magnitude of oscillations corresponding to the frequencies of the beat and meter when compared to the pre-stimulation period. This indicates that the brain’s response to rhythmic stimuli involves a clear enhancement in neural activity at specific frequencies aligned with the rhythm. However, since such a response is not unique to entrainment and is also a characteristic of AEP, MMN, and memory-based predictions, it is essential to distinguish entrainment from these other phenomena.

To distinguish entrainment from AEP and MMN responses, we introduced randomly omitted beats (Fig. 3). The sustained neural activity during omission epochs—statistically indistinguishable from responses to full stimulation, yet significantly different from pre-stimulation baselines—suggests an entrainment-like mechanism rather than mere evoked responses [47-49].

In addition, Fig. 4 shows that the AEP exhibited an immediate change upon stimulus onset, in contrast to the gradual dynamics observed in entrainment and memory-based prediction. Specifically, entrainment demonstrated both a noticeable build-up and recede latency, while memory-based prediction showed only a build-up latency without a recede component [25]. In the 3-beat paradigm, however, each epoch spanned 2.1 seconds, which limits the ability to clearly observe whether memory-based prediction is sustained beyond the initial response. Nevertheless, the consistent stability of neural responses—despite variations in omission rates that could influence attention or task complexity—strongly suggests that the observed effects arise from rhythmic entrainment rather than attentional modulation [50].

In Fig. 5, meter-synchronized cortical activity appears more spatially distributed than beat-synchronized activity. Beat entrainment is primarily localized to the secondary auditory cortex (BA 21~22), whereas meter—reflecting the hierarchical grouping of beats—engages additional regions involved in sensorimotor processing (BA 3, 6), music and language perception (BA 22, 40, 44), and executive functions (BA 9), indicating its broader cognitive demands [51]. This widespread engagement aligns with recent findings that adjacent cortical areas can encode distinct acoustic features depending on cortical layer [52], and our results similarly indicate that beat and meter are represented in distinct, spatially specific cortical sites.

The observed overlap between beat- and meter-synchronized regions, particularly within BA22, provides additional insight into the hierarchical organization of rhythm processing. This overlap may indicate that the superior temporal gyrus (STG), including BA22, serves as an integrative hub that encodes both periodic temporal prediction (beat) and broader temporal grouping (meter). Such dual involvement aligns with prior evidence suggesting that the STG mediates the transition from sensory driven to abstract temporal representations.

Importantly, while BA22 is often treated as a unitary auditory area, the observed overlap suggests functional heterogeneity within this region, where distinct or partially overlapping neuronal populations may encode rhythmic regularities at different hierarchical timescales. This finding also resonates with the concept of functional multiplicity, wherein anatomically defined regions exhibit continuity and overlap in function [53]. Within this framework, BA22 may support differentiated yet integrated temporal processing through subregional specialization.

Future studies incorporating effective connectivity analyses, high-density functional mapping, or laminar-specific recordings may help elucidate how such functional heterogeneity within BA22 contributes to multilevel rhythmic entrainment.

Areas involved in entrainment have been extensively studied using various neuroimaging techniques, including functional MRI (fMRI), EEG, and magnetoencephalography (MEG) [47, 54-57]. fMRI provides high spatial resolution for detailed brain mapping but lacks the temporal resolution needed to track rapid rhythmic dynamics. EEG offers excellent temporal resolution for fast neural responses but limited spatial precision. ECoG overcomes these limitations by providing both high spatial and temporal resolution, along with superior signal fidelity and resistance to noise [58].

However, several caveats must be acknowledged. First, this study was limited to cortical analyses, whereas subcortical structures—particularly the thalamus—are known to play a critical role in entrainment [59, 60]. Second, although ECoG offers high spatial and temporal resolution, it has inherent limitations. Electrode coverage varies across individuals, and factors such as patient pathology and surgical necessity influence electrode placement. Our analyses did not fully control for these inter-individual differences. Lastly, we did not conduct phase-based analyses. The low frequencies associated with beat and meter rhythms result in slow oscillatory cycles, making phase measures less sensitive to subtle or transient changes. Under such conditions, detecting omissions or gradual temporal dynamics becomes difficult, reducing the utility of phase metrics in distinguishing entrainment from evoked responses such as AEP, MMN, or memory-based predictions.

While our study operationalizes entrainment through frequency-specific spectral power, we acknowledge that this amplitude-based approach does not directly assess phase alignment, which is central to several contemporary entrainment models [61-63].

However, the temporal dynamics observed in our data—namely the gradual build-up and decay of oscillatory power around stimulus transitions—suggest underlying alignment processes consistent with phase-based entrainment mechanisms. In this sense, our results extend to previous models by showing that amplitude-level neural tracking can exhibit signatures similar to those described in phase-based frameworks. Future work combining spectral amplitudes and single-trial phase measures may help to bridge these perspectives and clarify the neural mechanisms underlying rhythmic tracking.

To address the distinction between entrainment and memory-based prediction, we evaluated not only frequency-specific synchronization during stimulation, but also the temporal persistence of rhythmic activity following stimulus offset. While the fixed epoch duration (2.1 seconds) corresponds to a single meter cycle in the 3-beat condition, our analysis included post-stimulation epochs to assess whether entrained responses receded gradually rather than terminating abruptly. For example, Fig. 4A illustrates a clear case of rhythmic activity persisting after the offset, showing both a build-up and a recede latency pattern. Such temporal profiles support the interpretation of entrainment as an internally maintained process, rather than a short-term memory-based prediction.

Although our design does not permit a definitive test of the commonly proposed 1-second threshold for sustained entrainment, we observed oscillatory persistence beyond 2.1 seconds in several electrodes, providing stronger evidence for entrained dynamics. This limitation and its implications have been clarified in the revised manuscript.

Nevertheless, we identified cortical regions that satisfy all three criteria of entrainment as an intrinsic neural response, with distinct processing mechanisms for beat and meter. Moreover, the cortices synchronized with beat and those synchronized with meter were spatially dissociable, suggesting functional segregation in rhythmic encoding. These findings not only deepen our understanding of the spatiotemporal dynamics underlying rhythm perception but also provide a neurophysiological framework for distinguishing entrainment from other evoked responses such as AEP and MMN. Importantly, this framework may inform future applications in clinical and cognitive neuroscience, including the development of neurodiagnostic tools for rhythm-related disorders, brain–computer interfaces, and rhythm-based neuromodulation therapies.

Supplemental Materials

en-35-1-17-supple.pdf (34MB, pdf)

ACKNOWLEDGEMENTS

The results of this paper were partially mentioned in the Master’s degree thesis of the author which applied behind closed doors for posting journals, and this paper was written by developing the contents of it. This research was supported by Samsung Research Funding & Incubation Center of Samsung Electronics under Project Number SRFC-IT1902-08, the National Research Foundation of Korea (NRF) grants funded by the Korean government (MSIT) (Nos. NRF-2021M3E5D2A01019538), and the Technology Innovation Program (Alchemist Project, 20012355) funded By the Ministry of Trade (MOTIE), South Korea.

Footnotes

AUTHOR CONTRIBUTIONS

YS contributes to formal analysis, data acquisition, draft the work and data interpretation. JK contributes to data acquisition, visualization and draft & revise the work. JSK contribute to conceptualization and quality control. CKC contributes to funding acquisition, conceptualization, supervision, review & editing. All authors had full access to the study design information and all data and approved final version to be published. Agreement to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

CONFLICT OF INTEREST

The authors declare that there is no conflict of interest regarding the publication of this paper.

SIGNIFICANCE STATEMENT

This study provides the first direct evidence of distinct cortical regions involved in beat and meter entrainment using high-resolution electrocorticography. We identified specific brain areas that demonstrate three key characteristics of true neural entrainment: maintenance during stimulus omission, build-up latency, and sustained neural oscillations that gradually diminish after stimulus offset. The findings reveal that beat processing is localized primarily in auditory areas (BA 21, 22), while meter processing engages a broader network including sensorimotor and cognitive regions (BA 3, 6, 9, 22, 40, 44). This dissociation between beat and meter processing networks advances our understanding of how the human brain processes rhythm and may have implications for treating rhythm-related neurological disorders.

DATA AVAILABILITY

Data is available from the corresponding author upon reasonable request.

References

  • 1.Fujioka T, Trainor LJ, Large EW, Ross B (2012) Internalized timing of isochronous sounds is represented in neuromagnetic β oscillations. J Neurosci 32:1791-1802. 10.1523/JNEUROSCI.4107-11.2012 [DOI] [PMC free article] [PubMed]
  • 2.Chang LJ, Jolly E, Cheong JH, Rapuano KM, Greenstein N, Chen PA, Manning JR (2021) Endogenous variation in ventromedial prefrontal cortex state dynamics during naturalistic viewing reflects affective experience. Sci Adv 7:eabf7129. 10.1126/sciadv.abf7129 [DOI] [PMC free article] [PubMed]
  • 3.Edalati M, Wallois F, Safaie J, Ghostine G, Kongolo G, Trainor LJ, Moghimi S (2023) Rhythm in the premature neonate brain: very early processing of auditory beat and meter. J Neurosci 43:2794-2802. 10.1523/JNEUROSCI.1100-22.2023 [DOI] [PMC free article] [PubMed]
  • 4.Nozaradan S, Peretz I, Mouraux A (2012) Steady-state evoked potentials as an index of multisensory temporal binding. Neuroimage 60:21-28. 10.1016/j.neuroimage.2011.11.065 [DOI] [PubMed]
  • 5.Lenc T, Keller PE, Varlet M, Nozaradan S (2020) Neural and behavioral evidence for frequency-selective context effects in rhythm processing in humans. Cereb Cortex Commun 1:tgaa037. 10.1093/texcom/tgaa037 [DOI] [PMC free article] [PubMed]
  • 6.Nozaradan S, Peretz I, Missal M, Mouraux A (2011) Tagging the neuronal entrainment to beat and meter. J Neurosci 31:10234-10240. 10.1523/JNEUROSCI.0411-11.2011 [DOI] [PMC free article] [PubMed]
  • 7.Berry W (1987) Structural functions in music. Dover Publications, Inc., New York, NY.
  • 8.Cooper G, Meyer LB (1963) The rhythmic structure of music. University of Chicago press, Chicago, IL.
  • 9.Large EW, Snyder JS (2009) Pulse and meter as neural resonance. Ann N Y Acad Sci 1169:46-57. 10.1111/j.1749-6632.2009.04550.x [DOI] [PubMed]
  • 10.Patel AD, Iversen JR (2014) The evolutionary neuroscience of musical beat perception: the Action Simulation for Auditory Prediction (ASAP) hypothesis. Front Syst Neurosci 8:57. 10.3389/fnsys.2014.00057 [DOI] [PMC free article] [PubMed]
  • 11.Jin X, Wang B, Lv Y, Lu Y, Chen J, Zhou C (2019) Does dance training influence beat sensorimotor synchronization? Differences in finger-tapping sensorimotor synchronization between competitive ballroom dancers and nondancers. Exp Brain Res 237:743-753. 10.1007/s00221-018-5410-4 [DOI] [PubMed]
  • 12.Durante AS, Wieselberg MB, Carvalho S, Costa N, Pucci B, Gudayol N, Almeida Kd (2014) Cortical auditory evoked potential: evaluation of speech detection in adult hearing aid users. Codas 26:367-373. 10.1590/2317-1782/20142013085 [DOI] [PubMed]
  • 13.Novembre G, Iannetti GD (2018) Tagging the musical beat: neural entrainment or event-related potentials? Proc Natl Acad Sci U S A 115:E11002-E11003. 10.1073/pnas.1815311115 [DOI] [PMC free article] [PubMed]
  • 14.Fiveash A, Schön D, Canette LH, Morillon B, Bedoin N, Tillmann B (2020) A stimulus-brain coupling analysis of regular and irregular rhythms in adults with dyslexia and controls. Brain Cogn 140:105531. 10.1016/j.bandc.2020.105531 [DOI] [PubMed]
  • 15.Ladinig O, Honing H, Háden G, Winkler I (2009) Probing attentive and preattentive emergent meter in adult listeners without extensive music training. Music Percept 26:377-386. 10.1525/mp.2009.26.4.377 [DOI]
  • 16.Geiser E, Ziegler E, Jancke L, Meyer M (2009) Early electrophysiological correlates of meter and rhythm processing in music perception. Cortex 45:93-102. 10.1016/j.cortex.2007.09.010 [DOI] [PubMed]
  • 17.Bouwer FL, Werner CM, Knetemann M, Honing H (2016) Disentangling beat perception from sequential learning and examining the influence of attention and musical abilities on ERP responses to rhythm. Neuropsychologia 85:80-90. 10.1016/j.neuropsychologia.2016.02.018 [DOI] [PubMed]
  • 18.Geiser E, Sandmann P, Jäncke L, Meyer M (2010) Refinement of metre perception--training increases hierarchical metre processing. Eur J Neurosci 32:1979-1985. 10.1111/j.1460-9568.2010.07462.x [DOI] [PubMed]
  • 19.Kim CH (2020) Expectation in music: electromagnetic studies on harmony, melody, and beat. Seoul National University Graduate School, Seoul.
  • 20.Winkler I (2007) Interpreting the mismatch negativity. J Psychophysiol 21:147-163. 10.1027/0269-8803.21.34.147 [DOI]
  • 21.Todd J, Heathcote A, Whitson LR, Mullens D, Provost A, Winkler I (2014) Mismatch negativity (MMN) to pitch change is susceptible to order-dependent bias. Front Neurosci 8:180. 10.3389/fnins.2014.00180 [DOI] [PMC free article] [PubMed]
  • 22.Timm L, Vuust P, Brattico E, Agrawal D, Debener S, Büchner A, Dengler R, Wittfoth M (2014) Residual neural processing of musical sound features in adult cochlear implant users. Front Hum Neurosci 8:181. 10.3389/fnhum.2014.00181 [DOI] [PMC free article] [PubMed]
  • 23.Nave KM, Hannon EE, Snyder JS (2022) Steady state-evoked potentials of subjective beat perception in musical rhythms. Psychophysiology 59:e13963. 10.1111/psyp.13963 [DOI] [PubMed]
  • 24.Breska A, Deouell LY (2017) Neural mechanisms of rhythm-based temporal prediction: delta phase-locking reflects temporal predictability but not rhythmic entrainment. PLoS Biol 15:e2001665. 10.1371/journal.pbio.2001665 [DOI] [PMC free article] [PubMed]
  • 25.Bouwer FL, Fahrenfort JJ, Millard SK, Slagter HA (2020) A silent disco: persistent entrainment of low-frequency neural oscillations underlies beat-based, but not memory-based temporal expectations. bioRxiv. doi: 10.1101/2020.01.08.899278. 10.1101/2020.01.08.899278 [DOI] [PubMed]
  • 26.Bouwer FL, Honing H, Slagter HA (2020) Beat-based and memory-based temporal expectations in rhythm: similar perceptual effects, different underlying mechanisms. J Cogn Neurosci 32:1221-1241. 10.1162/jocn_a_01529 [DOI] [PubMed]
  • 27.Guo Y, Bufacchi RJ, Novembre G, Kilintari M, Moayedi M, Hu L, Iannetti GD (2020) Ultralow-frequency neural entrainment to pain. PLoS Biol 18:e3000491. 10.1371/journal.pbio.3000491 [DOI] [PMC free article] [PubMed]
  • 28.Rosso M, Moens B, Leman M, Moumdjian L (2023) Neural entrainment underpins sensorimotor synchronization to dynamic rhythmic stimuli. Neuroimage 277:120226. 10.1016/j.neuroimage.2023.120226 [DOI] [PubMed]
  • 29.Morillon B, Schroeder CE, Wyart V, Arnal LH (2016) Temporal prediction in lieu of periodic stimulation. J Neurosci 36:2342-2347. 10.1523/JNEUROSCI.0836-15.2016 [DOI] [PMC free article] [PubMed]
  • 30.Jemmer P (2009) Getting in a (brain-wave) state through entrainment, meditation and hypnosis. Hypnother J 2:24-29.
  • 31.McPherson T, Berger D, Alagapan S, Fröhlich F (2018) Intrinsic rhythmicity predicts synchronization-continuation entrainment performance. Sci Rep 8:11782. 10.1038/s41598-018-29267-z [DOI] [PMC free article] [PubMed]
  • 32.Yamamoto K, Nakagawa S (2010) Evaluation of privacy protection techniques for speech signals. In: International conference on information processing and management of uncertainty in knowledge-based systems (Hüllermeier E, Kruse R, Hoffmann F, eds), pp 653-662. Springer, Berlin. 10.1007/978-3-642-14058-7_67 [DOI]
  • 33.Fink DJ (2017) What is a safe noise level for the public? Am J Public Health 107:44-45. 10.2105/AJPH.2016.303527 [DOI] [PMC free article] [PubMed]
  • 34.Hoppe U, Hocke T, Hast A, Iro H (2019) Maximum monosyllabic score as a predictor for cochlear implant outcome. HNO 67:199-206. 10.1007/s00106-018-0605-3 [DOI] [PubMed]
  • 35.Cook AM, Polgar JM (2014) Assistive technologies-e-book: principles and practice. Elsevier/Mosby, St. Louis, MO.
  • 36.Thaut MH, McIntosh GC, Hoemberg V (2015) Neurobiological foundations of neurologic music therapy: rhythmic entrainment and the motor system. Front Psychol 5:1185. 10.3389/fpsyg.2014.01185 [DOI] [PMC free article] [PubMed]
  • 37.Lancaster JL, Woldorff MG, Parsons LM, Liotti M, Freitas CS, Rainey L, Kochunov PV, Nickerson D, Mikiten SA, Fox PT (2000) Automated Talairach atlas labels for functional brain mapping. Hum Brain Mapp 10:120-131. 10.1002/1097-0193(200007)10:3<120::AID-HBM30>3.0.CO;2-8 [DOI] [PMC free article] [PubMed]
  • 38.Lancaster JL, Rainey LH, Summerlin JL, Freitas CS, Fox PT, Evans AC, Toga AW, Mazziotta JC (1997) Automated labeling of the human brain: a preliminary report on the development and evaluation of a forward-transform method. Hum Brain Mapp 5:238-242. 10.1002/(SICI)1097-0193(1997)5:4<238::AID-HBM6>3.0.CO;2-4 [DOI] [PMC free article] [PubMed]
  • 39.Lancaster JL, Tordesillas-Gutiérrez D, Martinez M, Salinas F, Evans A, Zilles K, Mazziotta JC, Fox PT (2007) Bias between MNI and Talairach coordinates analyzed using the ICBM-152 brain template. Hum Brain Mapp 28:1194-1205. 10.1002/hbm.20345 [DOI] [PMC free article] [PubMed]
  • 40.Laird AR, Robinson JL, McMillan KM, Tordesillas-Gutiérrez D, Moran ST, Gonzales SM, Ray KL, Franklin C, Glahn DC, Fox PT, Lancaster JL (2010) Comparison of the disparity between Talairach and MNI coordinates in functional neuroimaging data: validation of the Lancaster transform. Neuroimage 51:677-683. 10.1016/j.neuroimage.2010.02.048 [DOI] [PMC free article] [PubMed]
  • 41.Gramfort A, Luessi M, Larson E, Engemann DA, Strohmeier D, Brodbeck C, Goj R, Jas M, Brooks T, Parkkonen L, Hämäläinen M (2013) MEG and EEG data analysis with MNE-Python. Front Neurosci 7:267. 10.3389/fnins.2013.00267 [DOI] [PMC free article] [PubMed]
  • 42.Bertrand O, Perrin F, Pernier J (1985) A theoretical justification of the average reference in topographic evoked potential studies. Electroencephalogr Clin Neurophysiol 62:462-464. 10.1016/0168-5597(85)90058-9 [DOI] [PubMed]
  • 43.Knight RT, Scabini D, Woods DL, Clayworth C (1988) The effects of lesions of superior temporal gyrus and inferior parietal lobe on temporal and vertex components of the human AEP. Electroencephalogr Clin Neurophysiol 70:499-509. 10.1016/0013-4694(88)90148-4 [DOI] [PubMed]
  • 44.Woldorff MG, Hillyard SA, Gallen CC, Hampson SR, Bloom FE (1998) Magnetoencephalographic recordings demonstrate attentional modulation of mismatch-related neural activity in human auditory cortex. Psychophysiology 35:283-292. 10.1017/S0048577298961601 [DOI] [PubMed]
  • 45.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. 10.1016/j.clinph.2007.04.026 [DOI] [PubMed]
  • 46.Winkler I, Háden GP, Ladinig O, Sziller I, Honing H (2009) Newborn infants detect the beat in music. Proc Natl Acad Sci U S A 106:2468-2471. 10.1073/pnas.0809035106 [DOI] [PMC free article] [PubMed]
  • 47.Grahn JA, Brett M (2007) Rhythm and beat perception in motor areas of the brain. J Cogn Neurosci 19:893-906. 10.1162/jocn.2007.19.5.893 [DOI] [PubMed]
  • 48.Thut G, Schyns PG, Gross J (2011) Entrainment of perceptually relevant brain oscillations by non-invasive rhythmic stimulation of the human brain. Front Psychol 2:170. 10.3389/fpsyg.2011.00170 [DOI] [PMC free article] [PubMed]
  • 49.Thut G, Miniussi C, Gross J (2012) The functional importance of rhythmic activity in the brain. Curr Biol 22:R658-R663. 10.1016/j.cub.2012.06.061 [DOI] [PubMed]
  • 50.Esterman M, Noonan SK, Rosenberg M, Degutis J (2013) In the zone or zoning out? Tracking behavioral and neural fluctuations during sustained attention. Cereb Cortex 23:2712-2723. 10.1093/cercor/bhs261 [DOI] [PubMed]
  • 51.Thaut MH, Trimarchi PD, Parsons LM (2014) Human brain basis of musical rhythm perception: common and distinct neural substrates for meter, tempo, and pattern. Brain Sci 4:428-452. 10.3390/brainsci4020428 [DOI] [PMC free article] [PubMed]
  • 52.Leonard MK, Gwilliams L, Sellers KK, Chung JE, Xu D, Mischler G, Mesgarani N, Welkenhuysen M, Dutta B, Chang EF (2024) Large-scale single-neuron speech sound encoding across the depth of human cortex. Nature 626:593-602. 10.1038/s41586-023-06839-2 [DOI] [PMC free article] [PubMed]
  • 53.Haak KV, Beckmann CF (2020) Understanding brain organisation in the face of functional heterogeneity and functional multiplicity. Neuroimage 220:117061. 10.1016/j.neuroimage.2020.117061 [DOI] [PubMed]
  • 54.Vuust P, Roepstorff A, Wallentin M, Mouridsen K, Østergaard L (2006) It don't mean a thing…: keeping the rhythm during polyrhythmic tension, activates language areas (BA47). Neuroimage 31:832-841. 10.1016/j.neuroimage.2005.12.037 [DOI] [PubMed]
  • 55.Vuust P, Ostergaard L, Pallesen KJ, Bailey C, Roepstorff A (2009) Predictive coding of music--brain responses to rhythmic incongruity. Cortex 45:80-92. 10.1016/j.cortex.2008.05.014 [DOI] [PubMed]
  • 56.Kung SJ, Chen JL, Zatorre RJ, Penhune VB (2013) Interacting cortical and basal ganglia networks underlying finding and tapping to the musical beat. J Cogn Neurosci 25:401-420. 10.1162/jocn_a_00325 [DOI] [PubMed]
  • 57.Popescu M, Otsuka A, Ioannides AA (2004) Dynamics of brain activity in motor and frontal cortical areas during music listening: a magnetoencephalographic study. Neuroimage 21:1622-1638. 10.1016/j.neuroimage.2003.11.002 [DOI] [PubMed]
  • 58.Schalk G, Leuthardt EC (2011) Brain-computer interfaces using electrocorticographic signals. IEEE Rev Biomed Eng 4:140-154. 10.1109/RBME.2011.2172408 [DOI] [PubMed]
  • 59.Gauss R, Seifert R (2000) Pacemaker oscillations in heart and brain: a key role for hyperpolarization-activated cation channels. Chronobiol Int 17:453-469. 10.1081/CBI-100101057 [DOI] [PubMed]
  • 60.Lőrincz ML, Kékesi KA, Juhász G, Crunelli V, Hughes SW (2009) Temporal framing of thalamic relay-mode firing by phasic inhibition during the alpha rhythm. Neuron 63:683-696. 10.1016/j.neuron.2009.08.012 [DOI] [PMC free article] [PubMed]
  • 61.Lakatos P, Karmos G, Mehta AD, Ulbert I, Schroeder CE (2008) Entrainment of neuronal oscillations as a mechanism of attentional selection. Science 320:110-113. 10.1126/science.1154735 [DOI] [PubMed]
  • 62.Haegens S, Zion Golumbic E (2018) Rhythmic facilitation of sensory processing: a critical review. Neurosci Biobehav Rev 86:150-165. 10.1016/j.neubiorev.2017.12.002 [DOI] [PubMed]
  • 63.Obleser J, Kayser C (2019) Neural entrainment and attentional selection in the listening brain. Trends Cogn Sci 23:913-926. 10.1016/j.tics.2019.08.004 [DOI] [PubMed]

Associated Data

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

Supplementary Materials

en-35-1-17-supple.pdf (34MB, pdf)

Articles from Experimental Neurobiology are provided here courtesy of Korean Society for Brain and Neural Science

RESOURCES