Weighting of neural prediction error by rhythmic complexity: A predictive coding account using mismatch negativity.

The human brain's ability to extract and encode temporal regularities and to predict the timing of upcoming events is critical for music and speech perception. This work addresses how these mechanisms deal with different levels of temporal complexity, here the number of distinct durations in rhythmic patterns. We use electroencephalography (EEG) to relate the mismatch negativity (MMN), a proxy of neural prediction error, to a measure of information content of rhythmic sequences, the Shannon entropy. Within each of three conditions, participants listened to repeatedly presented standard rhythms of five tones (four inter-onset intervals) and of a given level of entropy: zero (isochronous), medium entropy (two distinct interval durations), or high entropy (four distinct interval durations). Occasionally, the fourth tone was moved forward in time that is it occurred 100 ms (small deviation) or 300 ms early (large deviation). According to the predictive coding framework, high-entropy stimuli are more difficult to model for the brain, resulting in less confident predictions and yielding smaller prediction errors for deviant sounds. Our results support this hypothesis, showing a gradual decrease in MMN amplitude as a function of entropy, but only for small timing deviants. For large timing deviants, in contrast, a modulation of activity in the opposite direction was observed for the earlier N1 component, known to also be sensitive to sudden changes in directed attention. Our results suggest the existence of a fine-grained neural mechanism that weights neural prediction error to the complexity of rhythms and that mostly manifests in the absence of directed attention.

Core auditory processing deficits in primary progressive aphasia.

The extent to which non-linguistic auditory processing deficits may contribute to the phenomenology of primary progressive aphasia is not established. Using non-linguistic stimuli devoid of meaning we assessed three key domains of auditory processing (pitch, timing and timbre) in a consecutive series of 18 patients with primary progressive aphasia (eight with semantic variant, six with non-fluent/agrammatic variant, and four with logopenic variant), as well as 28 age-matched healthy controls. We further examined whether performance on the psychoacoustic tasks in the three domains related to the patients' speech and language and neuropsychological profile. At the group level, patients were significantly impaired in the three domains. Patients had the most marked deficits within the rhythm domain for the processing of short sequences of up to seven tones. Patients with the non-fluent variant showed the most pronounced deficits at the group and the individual level. A subset of patients with the semantic variant were also impaired, though less severely. The patients with the logopenic variant did not show any significant impairments. Significant deficits in the non-fluent and the semantic variant remained after partialling out effects of executive dysfunction. Performance on a subset of the psychoacoustic tests correlated with conventional verbal repetition tests. In sum, a core central auditory impairment exists in primary progressive aphasia for non-linguistic stimuli. While the non-fluent variant is clinically characterized by a motor speech deficit (output problem), perceptual processing of tone sequences is clearly deficient. This may indicate the co-occurrence in the non-fluent variant of a deficit in working memory for auditory objects. Parsimoniously we propose that auditory timing pathways are altered, which are used in common for processing acoustic sequence structure in both speech output and acoustic input.

Dissociation of duration-based and beat-based auditory timing in cerebellar degeneration.

This work tests the hypothesis that the cerebellum is critical to the perception of the timing of sensory events. Auditory tasks were used to assess two types of timing in a group of patients with a stereotyped specific degeneration of the cerebellum: the analysis of single time intervals requiring absolute measurements of time, and the holistic analysis of rhythmic patterns based on relative measures of time using an underlying regular beat. The data support a specific role for the cerebellum only in the absolute timing of single subsecond intervals but not in the relative timing of rhythmic sequences with a regular beat. The findings support the existence of a stopwatch-like cerebellar timing mechanism for absolute intervals that is distinct from mechanisms for entrainment with a regular beat.

Distinct neural substrates of duration-based and beat-based auditory timing.

Research on interval timing strongly implicates the cerebellum and the basal ganglia as part of the timing network of the brain. Here we tested the hypothesis that the brain uses differential timing mechanisms and networks--specifically, that the cerebellum subserves the perception of the absolute duration of time intervals, whereas the basal ganglia mediate perception of time intervals relative to a regular beat. In a functional magnetic resonance imaging experiment, we asked human subjects to judge the difference in duration of two successive time intervals as a function of the preceding context of an irregular sequence of clicks (where the task relies on encoding the absolute duration of time intervals) or a regular sequence of clicks (where the regular beat provides an extra cue for relative timing). We found significant activations in an olivocerebellar network comprising the inferior olive, vermis, and deep cerebellar nuclei including the dentate nucleus during absolute, duration-based timing and a striato-thalamo-cortical network comprising the putamen, caudate nucleus, thalamus, supplementary motor area, premotor cortex, and dorsolateral prefrontal cortex during relative, beat-based timing. Our results support two distinct timing mechanisms and underlying subsystems: first, a network comprising the inferior olive and the cerebellum that acts as a precision clock to mediate absolute, duration-based timing, and second, a distinct network for relative, beat-based timing incorporating a striato-thalamo-cortical network.

Auditory sequence analysis and phonological skill.

This work tests the relationship between auditory and phonological skill in a non-selected cohort of 238 school students (age 11) with the specific hypothesis that sound-sequence analysis would be more relevant to phonological skill than the analysis of basic, single sounds. Auditory processing was assessed across the domains of pitch, time and timbre; a combination of six standard tests of literacy and language ability was used to assess phonological skill. A significant correlation between general auditory and phonological skill was demonstrated, plus a significant, specific correlation between measures of phonological skill and the auditory analysis of short sequences in pitch and time. The data support a limited but significant link between auditory and phonological ability with a specific role for sound-sequence analysis, and provide a possible new focus for auditory training strategies to aid language development in early adolescence.

Temporal predictions based on a gradual change in tempo.

Previous studies investigating sensitivity to step changes in tempo and prediction of tone onset time have generally utilized isochronous sequences. This study investigates subjects' ability to detect deviations from a gradual change in the tempo of a tone sequence (experiment 1) and their judgment of the perceptually optimal timing of this tone (experiment 2). In experiment 1, inter-onset-intervals within pairs of eight-tone sequences followed a geometric progression to create a gradual tempo change. In one sequence, the final tone was presented either earlier or later than specified by the progression. Subjects performed well at detecting deviations that exaggerated the tempo progression but poorly when it was counteracted. Experiment 2 used similar pairs except that the final tone was always presented earlier in one sequence than the other. Final interval length was adaptively adjusted to subjects' judgments; it was adjudged in best agreement with the progression when its length was roughly half way between the mathematically correct value and the length of the penultimate interval. The data support "multiple-look" and entrainment models of tempo sensitivity and suggest that temporal prediction is based less on the tempo contour of a whole sequence than on the duration of the preceding interval.

Left Frontal White Matter Links to Rhythm Processing Relevant to Speech Production in Apraxia of Speech.

Recent mechanistic models argue for a key role of rhythm processing in both speech production and speech perception. Patients with the non-fluent variant (NFV) of primary progressive aphasia (PPA) with apraxia of speech (AOS) represent a specific study population in which this link can be examined. Previously, we observed impaired rhythm processing in NFV with AOS. We hypothesized that a shared neurocomputational mechanism structures auditory input (sound and speech) and output (speech production) in time, a "temporal scaffolding" mechanism. Since considerable white matter damage is observed in NFV, we test here whether white matter changes are related to impaired rhythm processing. Forty-seven participants performed a psychoacoustic test battery: 12 patients with NFV and AOS, 11 patients with the semantic variant of PPA, and 24 cognitively intact age- and education-matched controls. Deformation-based morphometry was used to test whether white matter volume correlated to rhythmic abilities. In 34 participants, we also obtained tract-based metrics of the left Aslant tract, which is typically damaged in patients with NFV. Nine out of 12 patients with NFV displayed impaired rhythmic processing. Left frontal white matter atrophy adjacent to the supplementary motor area (SMA) correlated with poorer rhythmic abilities. The structural integrity of the left Aslant tract also correlated with rhythmic abilities. A colocalized and perhaps shared white matter substrate adjacent to the SMA is associated with impaired rhythmic processing and motor speech impairment. Our results support the existence of a temporal scaffolding mechanism structuring perceptual input and speech output.

Auditory beat perception is related to speech output fluency in post-stroke aphasia.

Aphasia affects at least one third of stroke survivors, and there is increasing awareness that more fundamental deficits in auditory processing might contribute to impaired language performance in such individuals. We performed a comprehensive battery of psychoacoustic tasks assessing the perception of tone pairs and sequences across the domains of pitch, rhythm and timbre in 17 individuals with post-stroke aphasia and 17 controls. At the level of individual differences we demonstrated a correlation between metrical pattern (beat) perception and speech output fluency with strong effect (Spearman's rho = 0.72). This dissociated from more basic auditory timing perception, which did not correlate with output fluency. This was also specific in terms of the language and cognitive measures, amongst which phonological, semantic and executive function did not correlate with beat detection. We interpret the data in terms of a requirement for the analysis of the metrical structure of sound to construct fluent output, with both being a function of higher-order "temporal scaffolding". The beat perception task herein allows measurement of timing analysis without any need to account for motor output deficit, and could be a potential clinical tool to examine this. This work suggests strategies to improve fluency after stroke by training in metrical pattern perception.

An information theoretic characterisation of auditory encoding.

The entropy metric derived from information theory provides a means to quantify the amount of information transmitted in acoustic streams like speech or music. By systematically varying the entropy of pitch sequences, we sought brain areas where neural activity and energetic demands increase as a function of entropy. Such a relationship is predicted to occur in an efficient encoding mechanism that uses less computational resource when less information is present in the signal: we specifically tested the hypothesis that such a relationship is present in the planum temporale (PT). In two convergent functional MRI studies, we demonstrated this relationship in PT for encoding, while furthermore showing that a distributed fronto-parietal network for retrieval of acoustic information is independent of entropy. The results establish PT as an efficient neural engine that demands less computational resource to encode redundant signals than those with high information content.

Metricality-enhanced temporal encoding and the subjective perception of rhythmic sequences.

Feeling the beat of a musical piece is easier for some pieces than others, depending on the underlying metrical structure. The present study sought to determine whether increasing metricality, meaning the amount of information supporting an intended meter, would elicit a corresponding increase in the precision of the temporal encoding of rhythmic sequences. Metricality was varied i) by using the Povel and Essens (1985) model of temporal accent induction to create a strong or weak sense of meter and ii) by including metrically plausible (compact) or implausible (open) endings. Precision of temporal encoding as a function of degree of metricality was assessed in an adaptively controlled change detection task. The change to be detected was a perturbation of relative interval timing that affected sequences as a whole rather than at specific points only. Change detection thresholds were significantly lower for sequences featuring a strong compared to a weak meter, and a compact compared to an open ending. Subjective ratings of rhythmicality of sequences also yielded main effects of strength of meter and ending. The data support an increase in the precision of temporal pattern encoding for sequences with a higher-order metrical time framework.

Auditory, Phonological, and Semantic Factors in the Recovery From Wernicke's Aphasia Poststroke: Predictive Value and Implications for Rehabilitation.

Background. Understanding the factors that influence language recovery in aphasia is important for improving prognosis and treatment. Chronic comprehension impairments in Wernicke's aphasia (WA) are associated with impairments in auditory and phonological processing, compounded by semantic and executive difficulties. This study investigated whether the recovery of auditory, phonological, semantic, or executive factors underpins the recovery from WA comprehension impairments by charting changes in the neuropsychological profile from the subacute to the chronic phase. Method. This study used a prospective, longitudinal observational design. Twelve WA participants with superior temporal lobe lesions were recruited 2 months post-stroke onset (2 MPO). Language comprehension was measured alongside a neuropsychological profile of auditory, phonological, and semantic processing and phonological short-term memory and nonverbal reasoning at 3 poststroke time points: 2.5, 5, and 9 MPO. Results. Language comprehension displayed a strong and consistent recovery between 2.5 and 9 MPO. Improvements were also seen for slow auditory temporal processing, phonological short-term memory, and semantic processing but not for rapid auditory temporal, spectrotemporal, and phonological processing. Despite their lack of improvement, rapid auditory temporal processing at 2.5 MPO and phonological processing at 5 MPO predicated comprehension outcomes at 9 MPO. Conclusions. These results indicate that recovery of language comprehension in WA can be predicted from fixed auditory processing in the subacute stage. This suggests that speech comprehension recovery in WA results from reorganization of the remaining language comprehension network to enable the residual speech signal to be processed more efficiently, rather than partial recovery of underlying auditory, phonological, or semantic processing abilities.

Deficits in Auditory Rhythm Perception in Children With Auditory Processing Disorder Are Unrelated to Attention.

Auditory processing disorder (APD) is defined as a specific deficit in the processing of auditory information along the central auditory nervous system, including bottom-up and top-down neural connectivity. Even though music comprises a big part of audition, testing music perception in APD population has not yet gained wide attention in research. This work tests the hypothesis that deficits in rhythm perception occur in a group of subjects with APD. The primary focus of this study is to measure perception of a simple auditory rhythm, i.e., short isochronous sequences of beats, in APD children and to compare their performance to age-matched normal controls. The secondary question is to study the relationship between cognition and auditory processing of rhythm perception. We tested 39 APD children and 25 control children aged between 6 and 12 years via (a) clinical APD tests, including a monaural speech in noise test, (b) isochrony task, a test measuring the detection of small deviations from perfect isochrony in a isochronous beats sequence, and (c) two cognitive tests (auditory memory and auditory attention). APD children scored worse in isochrony task compared to the age-matched control group. In the APD group, neither measure of cognition (attention nor memory) correlated with performance in isochrony task. Left (but not right) speech in noise performance correlated with performance in isochrony task. In the control group a large correlation (r = -0.701, p = 0.001) was observed between isochrony task and attention, but not with memory. The results demonstrate a deficit in the perception of regularly timed sequences in APD that is relevant to the perception of speech in noise, a ubiquitous complaint in this condition. Our results suggest (a) the existence of a non-attention related rhythm perception deficit in APD children and (b) differential effects of attention on task performance in normal vs. APD children. The potential beneficial use of music/rhythm training for rehabilitation purposes in APD children would need to be explored.

Autistic Traits and Enhanced Perceptual Representation of Pitch and Time.

Enhanced basic perceptual discrimination has been reported for pitch in individuals with autism spectrum conditions. We test whether there is a correlational pattern of enhancement across the broader autism phenotype and whether this correlation occurs for the discrimination of pitch, time and loudness. Scores on the Autism-Spectrum Quotient correlated significantly with the pitch discrimination (r = -0.51, p < 0.05) and the time-interval discrimination (r = -0.45, p < 0.05) task that were based on a fixed reference. No correlation was found for intensity discrimination based on a fixed reference, nor for a variable reference based time-interval discrimination. The correlations suggest a relationship between autistic traits and the ability to form an enhanced, stable and highly accurate representation of auditory events in the pitch and time dimensions.

Why Do Durations in Musical Rhythms Conform to Small Integer Ratios?

One curious aspect of human timing is the organization of rhythmic patterns in small integer ratios. Behavioral and neural research has shown that adjacent time intervals in rhythms tend to be perceived and reproduced as approximate fractions of small numbers (e.g., 3/2). Recent work on iterated learning and reproduction further supports this: given a randomly timed drum pattern to reproduce, participants subconsciously transform it toward small integer ratios. The mechanisms accounting for this "attractor" phenomenon are little understood, but might be explained by combining two theoretical frameworks from psychophysics. The scalar expectancy theory describes time interval perception and reproduction in terms of Weber's law: just detectable durational differences equal a constant fraction of the reference duration. The notion of categorical perception emphasizes the tendency to perceive time intervals in categories, i.e., "short" vs. "long." In this piece, we put forward the hypothesis that the integer-ratio bias in rhythm perception and production might arise from the interaction of the scalar property of timing with the categorical perception of time intervals, and that neurally it can plausibly be related to oscillatory activity. We support our integrative approach with mathematical derivations to formalize assumptions and provide testable predictions. We present equations to calculate durational ratios by: (i) parameterizing the relationship between durational categories, (ii) assuming a scalar timing constant, and (iii) specifying one (of K) category of ratios. Our derivations provide the basis for future computational, behavioral, and neurophysiological work to test our model.

Mismatch negativity on the cone of confusion.

Localization of sounds by the auditory system is based on the analysis of three sources of information: interaural level differences (ILD, caused by an attenuation of the sound as it travels to the more distant ear), interaural time differences (ITD, caused by the additional amount of time it takes for the sound to arrive at the more distant ear), and spectral cues (caused by direction-specific spectral filter properties of the pinnae). Although in a number of psychophysiological studies cortical processes of ITD and ILD analysis were investigated, there is hitherto no evidence on the cortical processing of spectral cues for sound localization. The objective of the present experiment was to test whether it is possible to observe electrophysiological correlates of sound localization based on spectral cues. In an auditory oddball experiment, 80 ms of broadband noise from varying free field locations were presented to inattentive participants. Mismatch negativities (MMNs) were observed for pairs of standards and location deviants located symmetrically with respect to the interaural axis. As interaural time and level differences are identical for such pairs of sounds, the observed MMNs most likely reflect cognitive processes of sound localization utilizing the spectral filter properties of the pinnae. MMN latencies suggest that sound localization based on spectral cues is slower than ITD- or ILD-based localization.

Artificial grammar learning in vascular and progressive non-fluent aphasias.

Patients with non-fluent aphasias display impairments of expressive and receptive grammar. This has been attributed to deficits in processing configurational and hierarchical sequencing relationships. This hypothesis had not been formally tested. It was also controversial whether impairments are specific to language, or reflect domain general deficits in processing structured auditory sequences. Here we used an artificial grammar learning paradigm to compare the abilities of controls to participants with agrammatic aphasia of two different aetiologies: stroke and frontotemporal dementia. Ten patients with non-fluent variant primary progressive aphasia (nfvPPA), 12 with non-fluent aphasia due to stroke, and 11 controls implicitly learned a novel mixed-complexity artificial grammar designed to assess processing of increasingly complex sequencing relationships. We compared response profiles for otherwise identical sequences of speech tokens (nonsense words) and tone sweeps. In all three groups the ability to detect grammatical violations varied with sequence complexity, with performance improving over time and being better for adjacent than non-adjacent relationships. Patients performed less well than controls overall, and this was related more strongly to aphasia severity than to aetiology. All groups improved with practice and performed well at a control task of detecting oddball nonwords. Crucially, group differences did not interact with sequence complexity, demonstrating that aphasic patients were not disproportionately impaired on complex structures. Hierarchical cluster analysis revealed that response patterns were very similar across all three groups, but very different between the nonsense word and tone tasks, despite identical artificial grammar structures. Overall, we demonstrate that agrammatic aphasics of two different aetiologies are not disproportionately impaired on complex sequencing relationships, and that the learning of phonological and non-linguistic sequences occurs independently. The similarity of profiles of discriminatory abilities and rule learning across groups suggests that insights from previous studies of implicit sequence learning in vascular aphasia are likely to prove applicable in nfvPPA.

The Beat to Read: A Cross-Lingual Link between Rhythmic Regularity Perception and Reading Skill.

This work assesses one specific aspect of the relationship between auditory rhythm cognition and language skill: regularity perception. In a group of 26 adult participants, native speakers of 11 different native languages, we demonstrate a strong and significant correlation between the ability to detect a "roughly" regular beat and rapid automatized naming (RAN) as a measure of language skill (Spearman's rho, -0.47, p < 0.01). There was no such robust relationship for the "mirror image" task of irregularity detection, i.e., the ability to detect ongoing small deviations from a regular beat. The correlation between RAN and regularity detection remained significant after partialling out performance on the irregularity detection task (rho, -0.41, p, 0.022), non-verbal IQ (rho, -0.37, p < 0.05), or musical expertise (rho, -0.31, p < 0.05). Whilst being consistent with the "shared resources model" in terms of rhythm as a common basis of language and music, evolutionarily as well as in individual development, the results also document how two related rhythm processing abilities relate differently to language skill. Specifically, the results support a universal relationship between rhythmic regularity detection and reading skill that is robust to accounting for differences in fluid intelligence and musical expertise, and transcends language-specific differences in speech rhythm.

Dysrhythmia: a specific congenital rhythm perception deficit.

Why do some people have problems "feeling the beat"? Here we investigate participants with congenital impairments in musical rhythm perception and production. A web-based version of the Montreal Battery of Evaluation of Amusia was used to screen for difficulties with rhythmic processing in a large sample and we identified three "dysrhythmic" individuals who scored below cut-off for the rhythm subtest, but not the pitch-based subtests. Follow-up testing in the laboratory was conducted to characterize the nature of both rhythm perception and production deficits in these dysrhythmic individuals. We found that they differed from control participants when required to synchronize their tapping to an external stimulus with a metrical pulse, but not when required to tap spontaneously (with no external stimulus) or to tap in time to an isochronous stimulus. Dysrhythmics exhibited a general tendency to tap at half the expected tempo when asked to synchronize to the beat of strongly metrical rhythms. These results suggest that the individuals studied here did not have motor production problems, but suffer from a selective rhythm perception deficit that influences the ability to entrain to metrical rhythms.

The basal ganglia in perceptual timing: timing performance in Multiple System Atrophy and Huntington's disease.

The timing of perceptual events depends on an anatomically and functionally connected network comprising basal ganglia, cerebellum, pre-frontal cortex and supplementary motor area. Recent studies demonstrate the cerebellum to be involved in absolute, duration-based timing, but not in relative timing based on a regular beat. Conversely, functional involvement of the striatum is observed in relative timing, but its role in absolute timing is unclear. This work tests the specific role of the basal ganglia in the perceptual timing of auditory events. It aims to distinguish the hypothesised unified model of time perception (Teki, Grube, & Griffiths, 2012), in which the striatum is a mandatory component for all timing tasks, from a modular system in which they subserve relative timing, with absolute timing processed by the cerebellum. Test groups comprised individuals with Multiple System Atrophy, a disorder in which similar pathology can produce clinical deficits associated with dysfunction of the cerebellum (MSA-C, n = 8) or striatum (MSA-P, n = 10), and early symptomatic Huntington's disease (HD, n = 14). Individuals with chronic autoimmune peripheral neuropathy (n = 11) acted as controls. Six adaptive tasks were carried out to assess perceptual thresholds for absolute timing through duration discrimination for sub- and supra-second time intervals, and relative timing through the detection of beat-based regularity and irregularity, detection of a delay within an isochronous sequence, and the discrimination of sequences with metrical structure. All three patient groups exhibited impairments in performance in comparison with the control group for all tasks, and severity of impairment was significantly correlated with disease progression. No differences were demonstrated between MSA-C and MSA-P, and the most severe impairments were observed in those with HD. The data support an obligatory role for the basal ganglia in all tested timing tasks, both absolute and relative, as predicted by the unified model. The results are not compatible with models of a brain timing network based upon independent modules.

Exploring the role of auditory analysis in atypical compared to typical language development.

The relationship between auditory processing and language skills has been debated for decades. Previous findings have been inconsistent, both in typically developing and impaired subjects, including those with dyslexia or specific language impairment. Whether correlations between auditory and language skills are consistent between different populations has hardly been addressed at all. The present work presents an exploratory approach of testing for patterns of correlations in a range of measures of auditory processing. In a recent study, we reported findings from a large cohort of eleven-year olds on a range of auditory measures and the data supported a specific role for the processing of short sequences in pitch and time in typical language development. Here we tested whether a group of individuals with dyslexic traits (DT group; n = 28) from the same year group would show the same pattern of correlations between auditory and language skills as the typically developing group (TD group; n = 173). Regarding the raw scores, the DT group showed a significantly poorer performance on the language but not the auditory measures, including measures of pitch, time and rhythm, and timbre (modulation). In terms of correlations, there was a tendency to decrease in correlations between short-sequence processing and language skills, contrasted by a significant increase in correlation for basic, single-sound processing, in particular in the domain of modulation. The data support the notion that the fundamental relationship between auditory and language skills might differ in atypical compared to typical language development, with the implication that merging data or drawing inference between populations might be problematic. Further examination of the relationship between both basic sound feature analysis and music-like sound analysis and language skills in impaired populations might allow the development of appropriate training strategies. These might include types of musical training to augment language skills via their common bases in sound sequence analysis.