Background
Auditory perception fundamentally depends on the analysis of temporal and spectral information. While spectral integration of frequency components supports timbre perception, accurate encoding and parsing of temporal structures are equally essential for interpreting dynamic acoustic events. However, most existing researches about multi-timescale temporal processing have relied heavily on the linguistic stimuli. Such approaches inherently limit research to humans with developed language abilities and depend largely on macroscopic recording methods such as EEG. Consequently, these studies offer limited insight into underlying neuronal computations. To advance our understanding of temporal integration across different timescales, it is necessary to adopt nonlinguistic stimuli and extend research across species using techniques capable of resolving mesoscopic and microscopic neural dynamics.

Research Progress
The researchers first employed a classical transitional click-train paradigm (Fig. 1A). This stimulus comprises two consecutive regular click trains with distinct inter-click intervals (ICIs). Electrocorticography (ECoG) recordings from macaque auditory cortex revealed pronounced transition-evoked change responses, whose magnitudes systematically decreased as the ICI increased. Fast Fourier Transform (FFT) analysis showed that synchronization to individual clicks (reflected by power spectral density peaks at corresponding repetition rates) strengthened with increasing ICI (Figs. 1B-E). Importantly, when ICIs fell within approximately 6–18 ms, both robust change responses and reliable click-level synchronization co-occurred, indicating simultaneous representation of fine-scale temporal cues and coarse-scale structural transitions (Figs. 1F-G). This co-representation was termed temporal integration during synchronization (TIDS), reflecting the auditory cortex’s capacity to integrate temporal information while phase-locking to rapid inputs.

To further verify the stability of TIDS, the researchers constructed an extended (18 s) click-train sequence by repeating short (0.2 s) transitional segments (Fig. 2A). Spectral analysis again confirmed synchronous tracking of both individual clicks and transitional click-train structures.

To probe higher-order temporal structure processing, the study introduced a click-train-based oddball paradigm (Fig. 3-1). This paradigm involved continuous click trains in two configurations: an initial 2-s consistent background click train (ICI_BG, gray
background in Fig. 3-1), followed by alternating click trains marked by ICI₁ (red background in Fig. 3-1) and ICI_BG every 300 ms. This sequence was repeated 9 times (for 5.4 s) followed by a deviant train with ICI₂ (blue background in Fig. 3-1). In the counterpart setup, the roles of the standard and deviant trains were reversed, while maintaining the background train constant.
This paradigm three hierarchical temporal levels:
• First level: individual clicks (tens of milliseconds)
• Second level: uniform click trains as auditory objects (hundreds of milliseconds)
• Third level: higher-order click trains requiring novelty detection (seconds)
Using the common SSA index (CSI), the study quantified third-level processing during single-unit recordings in A1 and MGB. Population analyses revealed that A1 robustly encoded temporal information across all three levels, whereas the MGB showed strong synchronization to individual clicks but comparatively limited capacity for integrating click-train objects or detecting higher-order deviations (Figs. 3-2, 3-3, 3-4). Thus, the primary auditory cortex demonstrated greater functional specialization for long-range temporal structure.

Given that deficits in temporal integration are associated with several psychiatric disorders, the researchers evaluated the paradigm’s potential clinical applicability using 64-channel EEG in 25 human participants. Human EEG data reproduced the multi-level synchronization patterns observed in macaques (Fig. 4). These results highlight the paradigm’s feasibility as a nonlinguistic assay of temporal processing, with promise for future clinical applications.

Future Prospects
This study introduces a novel nonlinguistic oddball paradigm based on click trains and demonstrates its effectiveness for probing hierarchical temporal processing across species. By bridging macroscopic EEG with mesoscopic ECoG and microscopic single-unit recordings, the work reveals that the auditory system can concurrently track multiple temporal scales and that A1 exhibits superior capacity relative to the MGB for processing complex temporal structures over longer timescales. Future studies should incorporate larger cohorts, employ high-resolution recording technologies, and systematically explore laminar specialization within auditory cortex. Additionally, validating the paradigm in rodent models will be essential for cross-species mechanistic comparison. These steps will deepen our understanding of the pathways and computations underlying temporal integration.

The complete study is accessible via DOI:10.34133/research.0960