Cascade and no-repetition rules are comparable controls for the auditory frequency mismatch negativity in oddball tasks

Two Prediction Error Systems in the Nonlemniscal Inferior Colliculus: "Spectral" and "Nonspectral"

According to the predictive processing framework, perception emerges from the reciprocal exchange of predictions and prediction errors (PEs) between hierarchically organized neural circuits. The nonlemniscal division of the inferior colliculus (IC) is the earliest source of auditory PE signals, but their neuronal generators, properties, and functional relevance have remained mostly undefined. We recorded single-unit mismatch responses to auditory oddball stimulation at different intensities, together with activity evoked by two sequences of alternating tones to control frequency-specific effects. Our results reveal a differential treatment of the unpredictable "many-standards" control and the predictable "cascade" control by lemniscal and nonlemniscal IC neurons that is not present in the auditory thalamus or cortex. Furthermore, we found that frequency response areas of nonlemniscal IC neurons reflect their role in subcortical predictive processing, distinguishing three hierarchical levels: (1) nonlemniscal neurons with sharply tuned receptive fields exhibit mild repetition suppression without signaling PEs, thereby constituting the input level of the local predictive processing circuitry. (2) Neurons with broadly tuned receptive fields form the main, "spectral" PE signaling system, which provides dynamic gain compensation to near-threshold unexpected sounds. This early enhancement of saliency reliant on spectral features was not observed in the auditory thalamus or cortex. (3) Untuned neurons form an accessory, "nonspectral" PE signaling system, which reports all surprising auditory deviances in a robust and consistent manner, resembling nonlemniscal neurons in the auditory cortex. These nonlemniscal IC neurons show unstructured and unstable receptive fields that could result from inhibitory input controlled by corticofugal projections conveying top-down predictions.

The sound of silence: Predictive error responses to unexpected sound omission in adults

The human auditory system excels at detecting patterns needed for processing speech and music. According to predictive coding, the brain predicts incoming sounds, compares predictions to sensory input, and generates a prediction error whenever a mismatch between the prediction and sensory input occurs. Predictive coding can be indexed in EEG with the mismatch negativity (MMN) and P3a components, two ERP components that are elicited by infrequent deviant sounds (e.g., differing in pitch, duration, loudness) in a stream of frequent sounds. If these components reflect prediction error, they should also be elicited by omitting an expected sound, but few studies have examined this. We compared ERPs elicited by infrequent randomly occurring omissions (unexpected silences) in tone sequences presented at 2 tones/sec to ERPs elicited by frequent, regularly occurring omissions (expected silences) within a sequence of tones and resting state EEG (a constant silence). We found that unexpected silences elicited significant MMN and P3a, although the magnitude of these components was quite small and variable. These results provide evidence for hierarchical predictive coding, indicating that the brain predicts silences as well as sounds.

Can intensity modulation of the auditory response explain intensity-decrement mismatch negativity?

Mismatch negativity (MMN) elicited by decrements in sound pressure level has been asserted as evidence for its dependence upon general deviance detection, while refuting the proposition that it is simply caused by modulating the intrinsic sensory response with different physical properties of sound. However, reports of intensity-decrement MMN are sparse compared with MMN to stimulus frequency or duration changes, and verifying the mechanisms that shape difference waveform morphology is essential for their responsible use as clinical biomarkers. In the present study, open-access EEG data from 40 healthy young adults recorded during an intensity-decrement oddball paradigm was analyzed to establish the effects of transitions between different level stimuli on the auditory evoked response. Standard stimuli were 80 dB and deviant stimuli were 70 dB. Event-related potentials were extracted from standards after standards (sS), deviants after standards (sD), and standards after deviants (dS). Mean amplitude across a recommended measurement window for MMN (125 to 225 ms) was calculated for each ERP waveform. This revealed statistically significant negative amplitude shift elicited by lower-intensity deviant stimuli, as expected, and an opposite direction, positive amplitude shift elicited by higher-intensity standard stimuli that followed lower-intensity deviants, relative to standard stimuli presented consecutively. These findings indicate that intensity-modulation of the auditory response influences cortical activity measured during the latency range of MMN. To what extent the hypothesized deviance detection mechanisms may also contribute is uncertain and remains to be elucidated.

The posterior auditory field is the chief generator of prediction error signals in the auditory cortex

The auditory cortex (AC) encompasses distinct fields subserving partly different aspects of sound processing. One essential function of the AC is the detection of unpredicted sounds, as revealed by differential neural activity to predictable and unpredictable sounds. According to the predictive coding framework, this effect can be explained by repetition suppression and/or prediction error signaling. The present study investigates functional specialization of the rat AC fields in repetition suppression and prediction error by combining a tone frequency oddball paradigm (involving high-probable standard and low-probable deviant tones) with two different control sequences (many-standards and cascade). Tones in the control sequences were comparable to deviant events with respect to neural adaptation but were not violating a regularity. Therefore, a difference in the neural activity between deviant and control tones indicates a prediction error effect, whereas a difference between control and standard tones indicates a repetition suppression effect. Single-unit recordings revealed by far the largest prediction error effects for the posterior auditory field, while the primary auditory cortex, the anterior auditory field, the ventral auditory field, and the suprarhinal auditory field were dominated by repetition suppression effects. Statistically significant repetition suppression effects occurred in all AC fields, whereas prediction error effects were less robust in the primary auditory cortex and the anterior auditory field. Results indicate that the non-lemniscal, posterior auditory field is more engaged in context-dependent processing underlying deviance-detection than the other AC fields, which are more sensitive to stimulus-dependent effects underlying differential degrees of neural adaptation.

Memorisation and implicit perceptual learning are enhanced for preferred musical intervals and chords

Is it true that we learn better what we like? Current neuroaesthetic and neurocomputational models of aesthetic appreciation postulate the existence of a correlation between aesthetic appreciation and learning. However, even though aesthetic appreciation has been associated with attentional enhancements, systematic evidence demonstrating its influence on learning processes is still lacking. Here, in two experiments, we investigated the relationship between aesthetic preferences for consonance versus dissonance and the memorisation of musical intervals and chords. In Experiment 1, 60 participants were first asked to memorise and evaluate arpeggiated triad chords (memorisation phase), then, following a distraction task, chords’ memorisation accuracy was measured (recognition phase). Memorisation resulted to be significantly enhanced for subjectively preferred as compared with non-preferred chords. To explore the possible neural mechanisms underlying these results, we performed an EEG study, directed to investigate implicit perceptual learning dynamics (Experiment 2). Through an auditory mismatch detection paradigm, electrophysiological responses to standard/deviant intervals were recorded, while participants were asked to evaluate the beauty of the intervals. We found a significant trial-by-trial correlation between subjective aesthetic judgements and single trial amplitude fluctuations of the ERP attention-related N1 component. Moreover, implicit perceptual learning, expressed by larger mismatch detection responses, was enhanced for more appreciated intervals. Altogether, our results showed the existence of a relationship between aesthetic appreciation and implicit learning dynamics as well as higher-order learning processes, such as memorisation. This finding might suggest possible future applications in different research domains such as teaching and rehabilitation of memory and attentional deficits.

Prediction error signaling explains neuronal mismatch responses in the medial prefrontal cortex

The mismatch negativity (MMN) is a key biomarker of automatic deviance detection thought to emerge from 2 cortical sources. First, the auditory cortex (AC) encodes spectral regularities and reports frequency-specific deviances. Then, more abstract representations in the prefrontal cortex (PFC) allow to detect contextual changes of potential behavioral relevance. However, the precise location and time asynchronies between neuronal correlates underlying this frontotemporal network remain unclear and elusive. Our study presented auditory oddball paradigms along with "no-repetition" controls to record mismatch responses in neuronal spiking activity and local field potentials at the rat medial PFC. Whereas mismatch responses in the auditory system are mainly induced by stimulus-dependent effects, we found that auditory responsiveness in the PFC was driven by unpredictability, yielding context-dependent, comparatively delayed, more robust and longer-lasting mismatch responses mostly comprised of prediction error signaling activity. This characteristically different composition discarded that mismatch responses in the PFC could be simply inherited or amplified downstream from the auditory system. Conversely, it is more plausible for the PFC to exert top-down influences on the AC, since the PFC exhibited flexible and potent predictive processing, capable of suppressing redundant input more efficiently than the AC. Remarkably, the time course of the mismatch responses we observed in the spiking activity and local field potentials of the AC and the PFC combined coincided with the time course of the large-scale MMN-like signals reported in the rat brain, thereby linking the microscopic, mesoscopic, and macroscopic levels of automatic deviance detection.

The effect of NMDA-R antagonist, MK-801, on neuronal mismatch along the rat auditory thalamocortical pathway

Efficient sensory processing requires that the brain maximize its response to unexpected stimuli, while suppressing responsivity to expected events. Mismatch negativity (MMN) is an auditory event-related potential that occurs when a regular pattern is interrupted by an event that violates the expected properties of the pattern. According to the predictive coding framework there are two mechanisms underlying the MMN: repetition suppression and prediction error. MMN has been found to be reduced in individuals with schizophrenia, an effect believed to be underpinned by glutamate N-methyl-d-aspartate receptor (NMDA-R) dysfunction. In the current study, we aimed to test how the NMDA-R antagonist, MK-801 in the anaesthetized rat, affected repetition suppression and prediction error processes along the auditory thalamocortical pathway. We found that low-dose systemic administration of MK-801 differentially affect thalamocortical responses, namely, increasing thalamic repetition suppression and cortical prediction error. Results demonstrate an enhancement of neuronal mismatch, also confirmed by large scale-responses. Furthermore, MK-801 produces faster and stronger dynamics of adaptation along the thalamocortical hierarchy. Clearly more research is required to understand how NMDA-R antagonism and dosage affects processes contributing to MMN. Nonetheless, because a low dose of an NMDA-R antagonist increased neuronal mismatch, the outcome has implications for schizophrenia treatment.

Do rat auditory event related potentials exhibit human mismatch negativity attributes related to predictive coding?

Rodent models play a significant role in understanding disease mechanisms and the screening of new treatments. With regard to psychiatric disorders such as schizophrenia, however, it is difficult to replicate the human symptoms in rodents because these symptoms are often either 'uniquely human' or are only conveyed via self-report. There is a growing interest in rodent mismatch responses (MMRs) as a translatable 'biomarker' for disorders such as schizophrenia. In this review, we will summarize the attributes of human MMN, and discuss the scope of exploring the attributes of human MMN in rodents. Here, we examine how reliably MMRs that are measured in rats mimic human attributes, and present original data examining whether manipulations of stimulus conditions known to modulate human MMN, do the same for rat MMRs. Using surgically-implanted epidural electroencephalographic electrodes and wireless telemetry in freely-moving rats, we observed human-like modulations of MMRs, namely that larger MMRs were elicited to unexpected (deviant) stimuli that a) had a larger change in pitch compared to the expected (standard) stimulus, b) were less frequently presented (lower probability), and c) had no jitter (stable stimulus onset asynchrony) compared to high jitter. Overall, these findings contribute to the mounting evidence for rat MMRs as a good analogue of human MMN, bolstering the development of a novel approach in future to validate the preclinical models based on a translatable biomarker, MMN.

Deviance detection in physiologically identified cell types in the rat auditory cortex

Auditory deviance detection is a function of the auditory system that allows reduction of the processing demand for repetitive stimuli while stressing unpredictable ones, which are potentially more informative. Deviance detection has been extensively studied in humans using the oddball paradigm, which evokes an event-related potential known as mismatch negativity (MMN). The same stimulation paradigms are used in animal studies that aim to elucidate the neuronal mechanisms underlying deviance detection. In order to understand the circuitry responsible for deviance detection in the auditory cortex (AC), it is necessary to determine the properties of excitatory and inhibitory neurons separately. Measuring the spike widths of neurons recorded extracellularly from the anaesthetized rat AC, we classified them as fast spiking or regular spiking units. These two neuron types are generally considered as putative inhibitory or excitatory, respectively. In response to an oddball paradigm, we found that both types of units showed similar amounts of deviance detection overall. When considering each AC field separately, we found that only in A1 fast spiking neurons showed higher deviance detection levels than regular spiking neurons, while in the rest of the fields there was no such distinction. Interpreting these responses in the context of the predictive coding framework, we found that the responses of both types of units reflect mainly prediction error signaling (i.e., genuine deviance detection) rather than repetition suppression.

Cortical Microcircuit Mechanisms of Mismatch Negativity and Its Underlying Subcomponents

In the neocortex, neuronal processing of sensory events is significantly influenced by context. For instance, responses in sensory cortices are suppressed to repetitive or redundant stimuli, a phenomenon termed “stimulus-specific adaptation” (SSA). However, in a context in which that same stimulus is novel, or deviates from expectations, neuronal responses are augmented. This augmentation is termed “deviance detection” (DD). This contextual modulation of neural responses is fundamental for how the brain efficiently processes the sensory world to guide immediate and future behaviors. Notably, context modulation is deficient in some neuropsychiatric disorders such as schizophrenia (SZ), as quantified by reduced “mismatch negativity” (MMN), an electroencephalography waveform reflecting a combination of SSA and DD in sensory cortex. Although the role of NMDA-receptor function and other neuromodulatory systems on MMN is established, the precise microcircuit mechanisms of MMN and its underlying components, SSA and DD, remain unknown. When coupled with animal models, the development of powerful precision neurotechnologies over the past decade carries significant promise for making new progress into understanding the neurobiology of MMN with previously unreachable spatial resolution. Currently, rodent models represent the best tool for mechanistic study due to the vast genetic tools available. While quantifying human-like MMN waveforms in rodents is not straightforward, the “oddball” paradigms used to study it in humans and its underlying subcomponents (SSA/DD) are highly translatable across species. Here we summarize efforts published so far, with a focus on cortically measured SSA and DD in animals to maintain relevance to the classically measured MMN, which has cortical origins. While mechanistic studies that measure and contrast both components are sparse, we synthesize a potential set of microcircuit mechanisms from the existing rodent, primate, and human literature. While MMN and its subcomponents likely reflect several mechanisms across multiple brain regions, understanding fundamental microcircuit mechanisms is an important step to understand MMN as a whole. We hypothesize that SSA reflects adaptations occurring at synapses along the sensory-thalamocortical pathways, while DD depends on both SSA inherited from afferent inputs and resulting disinhibition of non-adapted neurons arising from the distinct physiology and wiring properties of local interneuronal subpopulations and NMDA-receptor function.

Mismatch-negativity (MMN) in animal models: Homology of human MMN?

Mismatch negativity (MMN) has long been considered to be one of the deviance-detecting neural characteristics. Animal models exhibit similar neural activities, called MMN-like responses; however, there has been considerable debate on whether MMN-like responses are homologous to MMN in humans. Herein, we reviewed several studies that compared the electrophysiological, pharmacological, and functional properties of MMN-like responses and adaptation-exhibiting middle-latency responses (MLRs) in animals with those in humans. Accumulating evidence suggests that there are clear differences between MMN-like responses and MLRs, in particular that MMN-like responses can be distinguished from mere effects of adaptation, i.e., stimulus-specific adaptation. Finally, we discuss a new direction for research on MMN-like responses by introducing our recent work, which demonstrated that MMN-like responses represent empirical salience of deviant stimuli, suggesting a new functional role of MMN beyond simple deviance detection.

Deviance detection in sound frequency in simple and complex sounds in urethane-anesthetized rats

Mismatch negativity (MMN), which is an electrophysiological response demonstrated in humans and animals, reflects memory-based deviance detection in a series of sounds. However, only a few studies on rodents have used control conditions that were sufficient in eliminating confounding factors that could also explain differential responses to deviant sounds. Furthermore, it is unclear if change detection occurs similarly for sinusoidal and complex sounds. In this study, we investigated frequency change detection in urethane-anesthetized rats by recording local-field potentials from the dura above the auditory cortex. We studied change detection in sinusoidal and complex sounds in a series of experiments, controlling for sound frequency, probability, and pattern in a series of sounds. For sinusoidal sounds, the MMN controlled for frequency, adaptation, and pattern, was elicited at approximately 200 ms onset latency. For complex sounds, the MMN controlled for frequency and adaptation, was elicited at 60 ms onset latency. Sound frequency affected the differential responses. MMN amplitude was larger for the sinusoidal sounds than for the complex sounds. These findings indicate the importance of controlling for sound frequency and stimulus probabilities, which have not been fully controlled for in most previous animal and human studies. Future studies should confirm the preference for sinusoidal sounds over complex sounds in rats.

Visual Perceptual Load Does Not Affect the Frequency Mismatch Negativity

The mismatch negativity (MMN) has been of particular interest in auditory perception because of its sensitivity to auditory change. It is typically measured in an oddball task and is computed as the difference of deviant minus standard tones. Previous studies suggest that the oddball MMN can be reduced by crossmodal attention to a concurrent, difficult visual task. However, more recent studies did not replicate this effect. Because previous findings seem to be biased, we preregistered the present study and used Bayesian hypothesis testing to measure the strength of evidence for or against an effect of visual task difficulty. We manipulated visual perceptual load (high and low load). In the task, the visual stimuli were identical for both loads to avoid confounding effects from physical differences of the visual stimuli. We also measured the corrected MMN because the oddball MMN may be confounded by physical differences between deviant and standard tones. The corrected MMN is obtained with a separate control condition in which the same tone as the deviant (critical tone) is equiprobable with other tones. The corrected MMN is computed as deviant minus critical tones. Furthermore, we assessed working memory capacity to examine its moderating role. In our large sample (N = 49), the evidential strength in support of no effect of visual load was moderate for the oddball MMN (9.09 > BF01 > 3.57) and anecdotal to moderate for the corrected MMN (4.55 > BF01 > 2.17). Also, working memory capacity did not correlate with the visual load effect on the oddball MMN and the corrected MMN. The present findings support the robustness of the auditory frequency MMN to manipulations of crossmodal, visual attention and suggest that this relationship is not moderated by working memory capacity.

Cascade and no-repetition rules are comparable controls for the auditory frequency mismatch negativity in oddball tasks

The mismatch negativity (MMN) has been widely studied with oddball tasks to index processing of unexpected auditory change. The MMN is computed as the difference of deviant minus standard and is used to capture the pattern violation by the deviant. However, this oddball MMN is confounded because the deviant differs physically from the standard and is presented less often. To improve measurement, the same tone as the deviant is presented in a separate condition. This control tone is equiprobable with other tones and is used to compute a corrected MMN (deviant minus control). Typically, the tones are in random order except that consecutive tones are not identical (no-repetition rule). In contrast, a recent study on frequency MMN presented tones in a regular up-and-down sequence (cascade rule). If the cascade rule is detected more easily than the no-repetition rule, there should be a lower risk of a confounding MMN within the cascade condition. However, in previous research, the cascade and no-repetition conditions differed not only in the regularity of the tone sequence but also in number of tones, frequency range, and proportion of tones. We controlled for these differences to isolate effects of regularity in the tone sequence. Results of our preregistered analyses provided moderate evidence (BF₀₁ >6) that the corrected MMN did not differ between cascade and no-repetition conditions. These findings imply that no-repetition and cascade rules are processed similarly and that the no-repetition condition provides an adequate control in frequency MMN.