Finding tau rhythms in EEG: An independent component analysis approach.

Tau rhythms are largely defined by sound responsive alpha band (~8-13 Hz) oscillations generated largely within auditory areas of the superior temporal gyri. Studies of tau have mostly employed magnetoencephalography or intracranial recording because of tau's elusiveness in the electroencephalogram. Here, we demonstrate that independent component analysis (ICA) decomposition can be an effective way to identify tau sources and study tau source activities in EEG recordings. Subjects (N = 18) were passively exposed to complex acoustic stimuli while the EEG was recorded from 68 electrodes across the scalp. Subjects' data were split into 60 parallel processing pipelines entailing use of five levels of high-pass filtering (passbands of 0.1, 0.5, 1, 2, and 4 Hz), three levels of low-pass filtering (25, 50, and 100 Hz), and four different ICA algorithms (fastICA, infomax, adaptive mixture ICA [AMICA], and multi-model AMICA [mAMICA]). Tau-related independent component (IC) processes were identified from this data as being localized near the superior temporal gyri with a spectral peak in the 8-13 Hz alpha band. These "tau ICs" showed alpha suppression during sound presentations that was not seen for other commonly observed IC clusters with spectral peaks in the alpha range (e.g., those associated with somatomotor mu, and parietal or occipital alpha). The choice of analysis parameters impacted the likelihood of obtaining tau ICs from an ICA decomposition. Lower cutoff frequencies for high-pass filtering resulted in significantly fewer subjects showing a tau IC than more aggressive high-pass filtering. Decomposition using the fastICA algorithm performed the poorest in this regard, while mAMICA performed best. The best combination of filters and ICA model choice was able to identify at least one tau IC in the data of ~94% of the sample. Altogether, the data reveal close similarities between tau EEG IC dynamics and tau dynamics observed in MEG and intracranial data. Use of relatively aggressive high-pass filters and mAMICA decomposition should allow researchers to identify and characterize tau rhythms in a majority of their subjects. We believe adopting the ICA decomposition approach to EEG analysis can increase the rate and range of discoveries related to auditory responsive tau rhythms.

Not fully remembered, but not forgotten: interfering sounds worsen but do not eliminate the representation of pitch in working memory.

Recent research has begun measuring auditory working memory with a continuous adjustment task in which listeners adjust attributes of a sound to match a stimulus presented earlier. This approach captures auditory memory's continuous nature better than standard change detection paradigms that collect binary ("same or different") memory measurements. In two experiments, we assessed the impact of different interference stimuli (multitone complexes vs. white noise vs. silence) on the precision and accuracy of participants' reproductions of pitch from memory. Participants were presented with a target multitone complex stimulus followed by eight successive interference signals. Across trials, these signals alternated between additional multitone complexes, randomly generated white noise samples, or (in Experiment 2) silence. This was followed by a response period where participants adjusted the pitch of a response stimulus using a MIDI touchpad to match the target. Experiment 1 found a significant effect of interference type on performance, with tone interference signals producing the greatest impairments to participants' accuracy and precision compared to white noise. Interestingly, it also found a compression in the participants' responses, with overestimations of low-frequency targets and underestimations for high-frequency targets. Experiment 2 replicated results from Experiment 1, with an additional silence condition showing the best performance, suggesting that non-tonal signals also generate interference. In general, results support a shared resource model of working memory with a limited capacity that can be flexibly allocated to hold items in memory with varying levels of fidelity. Interference does not appear to knock items out of a fixed precision slot, but rather robs a portion of capacity from stored items.

Late ERP correlates of confidence for auditory categorization of complex sounds.

Recent research suggests that confidence judgments relate to the quality of early sensory representations and later modality independent processing stages. It is not known whether the nature of this finding might vary based on task and/or stimulus characteristics (e.g., detection vs. categorization). The present study investigated the neural correlates of confidence using electroencephalography (EEG) in an auditory categorization task. This allowed us to examine whether the early event-related potentials (ERPs) related to confidence in detection also apply to a more complex auditory task. Participants listened to frequency-modulated (FM) tonal stimuli going up or down in pitch. The rate of FM tones ranged from slow to fast, making the stimuli harder or easier to categorize. Tone-locked late posterior positivity (LPP) but not N1 or P2 amplitudes were larger for (correct-only) trials rated with high than low confidence. These results replicated for trials presenting stimuli at individually identified threshold levels (rate of change producing ∼71.7% correct performance). This finding suggests that, in this task, neural correlates of confidence do not vary based on difficulty level. We suggest that the LPP is a task general indication of the confidence for an upcoming judgment in a variety of paradigms.

Auditory short-term memory for pitch loses precision over time.

The impact of retention interval duration on the fidelity of pitch memory was investigated. Listeners heard "target" pure tones, followed by a retention interval (2-8 s), then a response period in which the frequency of a novel sound was adjusted to match their memory of the target. The variability of pitch matches increased with retention interval duration. Supplemental analyses of the most accurate trials and temporal dynamics of matching suggest that decreasing precision was not due to differences in complete forgetting among intervals. Results suggest that the precision of short-term memory for pitch may continuously degrade over time.

Learning to detect auditory signals in noise: Active top-down selection and stable change in signal representations.

Training can improve detection of auditory signals in noise. This learning could potentially occur through active top-down selection mechanisms or stable changes in signal representations. Here, participants were trained and tested (pretest vs. posttest design) on abilities to detect pure tone signals in noise. Auditory evoked potentials (AEPs) to tones were gathered under dichotic listening conditions where participants attended to nontonal stimuli in the opposite ear. Improvements in detection sensitivity were observable regardless of tested tone frequency. This was true in generalization between 861 Hz and 1058-Hz tones (Experiment 1a), and when testing a frequency range > 1 octave (Experiment 2). Such learning was not apparent without training (Experiment 1b). In contrast to behavior, AEP amplitude increases from pre- to posttest were partially specific to trained tone frequencies, even when selective attention was diverted to the opposite ear of tone presentation. Placed in the context of previous work, results suggest that changes in active top-down selection mechanisms and stable signal representations both play a role in auditory detection learning. The mismatch between AEP and behavioral effects suggests a need to consider how these different learning processes can impact detection performance in the variety of listening scenarios a listener may face. (PsycInfo Database Record (c) 2023 APA, all rights reserved).

Effortful listening produces both enhancement and suppression of alpha in the EEG.

Introduction: Adverse listening conditions can drive increased mental effort during listening. Neuromagnetic alpha oscillations (8-13 Hz) may index this listening effort, but inconsistencies regarding the direction of the relationship are abundant. We performed source analyses on high-density EEG data collected during a speech-on-speech listening task to address the possibility that opposing alpha power relationships among alpha producing brain sources drive this inconsistency. Methods: Listeners (N=20) heard two simultaneously presented sentences of the form: Ready go to now. They either reported the color/number pair of a "Baron" call sign sentence (active: high effort), or ignored the stimuli (passive: low effort). Independent component analysis (ICA) was used to segregate temporally distinct sources in the EEG. Results: Analysis of independent components (ICs) revealed simultaneous alpha enhancements (e.g., for somatomotor mu ICs) and suppressions (e.g., for left temporal ICs) for different brain sources. The active condition exhibited stronger enhancement for left somatomotor mu rhythm ICs, but stronger suppression for central occipital ICs. Discussion: This study shows both alpha enhancement and suppression to be associated with increases in listening effort. Literature inconsistencies could partially relate to some source activities overwhelming others in scalp recordings.

Familiarization with meaningless sound patterns facilitates learning to detect those patterns among distracters.

Initially "meaningless" and randomly generated sounds can be learned over exposure. This is demonstrated by studies where repetitions of randomly determined sound patterns are detected better if they are the same sounds presented on previous trials than if they are novel. This experiment posed two novel questions about this learning. First, does familiarization with a sound outside of the repetition detection context facilitate later performance? Second, does familiarization enhance performance when repeats are interleaved with distracters? Listeners were first trained to categorize a unique pattern of synchronous complex tone trains (210 ms in duration) from other tone trains with similar qualities (familiarization phase). They were then tasked to detect repeated pattern presentations interleaved with similar distracters in 4.2 s long excerpts (repetition detection phase). The familiarized pattern (Familiar Fixed - FF), an unfamiliar pattern that remained fixed throughout (Unfamiliar Fixed - UF), or patterns that were uniquely determined on each trial (Unfamiliar Unfixed - UU) could be presented as repeats. FF patterns were learned at a faster rate and achieved higher repetition detection sensitivity than UF and UU patterns. Similarly, FF patterns also showed steeper learning slopes in their response times (RTs) than UF patterns. The data show that familiarity with a "meaningless" sound pattern on its own (i.e., without repetition) can facilitate repetition detection even in the presence of distracters. Familiarity effects become most apparent in the potential for learning.

EEG power spectral dynamics associated with listening in adverse conditions.

Adverse listening conditions increase the demand on cognitive resources needed for speech comprehension. In an exploratory study, we aimed to identify independent power spectral features in the EEG useful for studying the cognitive processes involved in this effortful listening. Listeners performed the coordinate response measure task with a single-talker masker at a 0-dB signal-to-noise ratio. Sounds were left unfiltered or degraded with low-pass filtering. Independent component analysis (ICA) was used to identify independent components (ICs) in the EEG data, the power spectral dynamics of which were then analyzed. Frontal midline theta, left frontal, right frontal, left mu, right mu, left temporal, parietal, left occipital, central occipital, and right occipital clusters of ICs were identified. All IC clusters showed some significant listening-related changes in their power spectrum. This included sustained theta enhancements, gamma enhancements, alpha enhancements, alpha suppression, beta enhancements, and mu rhythm suppression. Several of these effects were absent or negligible using traditional channel analyses. Comparison of filtered to unfiltered speech revealed a stronger alpha suppression in the parietal and central occipital clusters of ICs for the filtered speech condition. This not only replicates recent findings showing greater alpha suppression as listening difficulty increases but also suggests that such alpha-band effects can stem from multiple cortical sources. We lay out the advantages of the ICA approach over the restrictive analyses that have been used as of late in the study of listening effort. We also make suggestions for moving into hypothesis-driven studies regarding the power spectral features that were revealed.

Effects of auditory training on low-pass filtered speech perception and listening-related cognitive load.

Studies supporting learning-induced reductions in listening-related cognitive load have lacked procedural learning controls, making it difficult to determine the extent to which effects arise from perceptual or procedural learning. Here, listeners were trained in the coordinate response measure (CRM) task under unfiltered (UT) or degraded low-pass filtered (FT) conditions. Improvements in low-pass filtered CRM performance were larger for FT. Both conditions showed training-related reductions in cognitive load as indexed by a secondary working memory task. However, only the FT condition showed a correlation between CRM improvement and secondary task performance, suggesting that effects can be driven by perceptual and procedural learning.

Auditory detection learning is accompanied by plasticity in the auditory evoked potential.

Auditory detection can improve with practice. These improvements are often assumed to arise from selective attention processes, but longer-term plasticity as a result of training may also play a role. Here, listeners were trained to detect either an 861-Hz or 1058-Hz tone (counterbalanced across participants) presented in noise at SNRs varying from -10 to -24 dB. On the following day, they were tasked with detecting 861-Hz and 1058-Hz tones at an SNR of -21 dB. In between blocks of this active task, EEG was recorded during passive presentation of trained and untrained frequency tones in quiet. Detection accuracy and confidence ratings were higher for trials at listeners' trained, than untrained-frequency (i.e., learning occurred). During passive exposure to sounds, the P2 component of the auditory evoked potential (∼150 - 200 ms post tone onset) was larger in amplitude for the trained compared to the untrained frequency. An analysis of global field power similarly yielded a stronger response for trained tones in the P2 time window. These effects were obtained during passive exposure, suggesting that training induced improvements in detection are not solely related to changes in selective attention. Rather, there may be an important role for changes in the long-term neural representations of sounds.

Correction: Artificial neural networks reveal individual differences in metacognitive monitoring of memory.

[This corrects the article DOI: 10.1371/journal.pone.0220526.].

Artificial neural networks reveal individual differences in metacognitive monitoring of memory.

Previous work supports an age-specific impairment for recognition memory of pairs of words and other stimuli. The present study tested the generalization of an associative deficit across word, name, and nonword stimulus types in younger and older adults. Participants completed associative and item memory tests in one of three stimulus conditions and made metacognitive ratings of perceptions of self-efficacy, task success ("postdictions"), strategy success, task effort, difficulty, fatigue, and stamina. Surprisingly, no support was found for an age-related associative deficit on any of the stimulus types. We analyzed our data further using a multilayer perceptron artificial neural network. The network was trained to classify individuals as younger or older and its hidden unit activities were examined to identify data patterns that distinguished younger from older participants. Analysis of hidden unit activities revealed that the network was able to correctly classify by identifying three different clusters of participants, with two qualitatively different groups of older individuals. One cluster of older individuals found the tasks to be relatively easy, they believed they had performed well, and their beliefs were accurate. The other cluster of older individuals found the tasks to be difficult, believed they were performing relatively poorly, yet their beliefs did not map accurately onto their performance. Crucially, data from the associative task were more useful for neural networks to discriminate between younger and older adults than data from the item task. This work underscores the importance of considering both individual and age differences as well as metacognitive responses in the context of associative memory paradigms.

Easy-to-hard effects in perceptual learning depend upon the degree to which initial trials are "easy".

Starting perceptual training at easy levels before progressing to difficult levels generally produces better learning outcomes than constantly difficult training does. However, little is known about how "easy" these initial levels should be in order to yield easy-to-hard effects. We compared five levels of initial training block difficulty varying from very easy to hard in two auditory-discrimination learning tasks-a frequency modulation rate discrimination (Experiment 1) and a frequency range discrimination (Experiment 2). The degree of difficulty was based on individualized pretraining ~71% correct discrimination thresholds. Both experiments revealed a sweet spot for easy-to-hard effects. Conditions where initial blocks were either too easy or too difficult produced less benefit than did blocks of intermediate difficulty. Results challenge assumptions that sequencing effects in learning are related to attentional spotlighting of task-relevant dimensions. Rather, they support incremental learning models that account for easy-to-hard effects. Further, the results have implications for how perceptual training regimens should be designed to maximize the benefits of rehabilitative perceptual training.

Pre-stimulus brain state predicts auditory pattern identification accuracy.

Recent studies show that pre-stimulus band-specific power and phase in the electroencephalogram (EEG) can predict accuracy on tasks involving the detection of near-threshold stimuli. However, results in the auditory modality have been mixed, and few works have examined pre-stimulus features when more complex decisions are made (e.g. identifying supra-threshold sounds). Further, most auditory studies have used background sounds known to induce oscillatory EEG states, leaving it unclear whether phase predicts accuracy without such background sounds. To address this gap in knowledge, the present study examined pre-stimulus EEG as it relates to accuracy in a tone pattern identification task. On each trial, participants heard a triad of 40-ms sinusoidal tones (separated by 40-ms intervals), one of which was at a different frequency than the other two. Participants' task was to indicate the tone pattern (low-low-high, low-high-low, etc.). No background sounds were employed. Using a phase opposition measure based on inter-trial phase consistencies, pre-stimulus 7-10 Hz phase was found to differ between correct and incorrect trials ∼200 to 100 ms prior to tone-pattern onset. After sorting trials into bins based on phase, accuracy was found to be lowest at around π-+ relative to individuals' most accurate phase bin. No significant effects were found for pre-stimulus power. In the context of the literature, findings suggest an important relationship between the complexity of task demands and pre-stimulus activity within the auditory domain. Results also raise interesting questions about the role of induced oscillatory states or rhythmic processing modes in obtaining pre-stimulus effects of phase in auditory tasks.

Confidence tracks sensory- and decision-related ERP dynamics during auditory detection.

Recent research has focused on measuring neural correlates of metacognitive judgments in decision and post-decision processes during memory retrieval and categorization. However, many tasks (e.g., stimulus detection) may require monitoring of earlier sensory processing. Here, participants indicated which of two intervals contained an 80-ms pure tone embedded in white noise. One frequency (e.g., 1000 Hz) was presented on ∼80% of all trials (i.e., 'primary' trials). Another frequency (e.g., 2500 Hz) was presented on ∼20% of trials (i.e., 'probe' trials). The event-related potential (ERP) was used to investigate the processing stages related to confidence. Tone-locked N1, P2, and P3 amplitudes were larger for trials rated with high than low confidence. Interestingly, a P3-like late positivity for the tone-absent interval showed high amplitude for low confidence. No 'primary' vs. 'probe' differences were found. However, confidence rating differences between primary and probe trials were correlated with N1 and tone-present P3 amplitude differences. We suggest that metacognitive judgments can track both sensory- and decision-related processes (indexed by the N1 and P3, respectively). The particular processes on which confidence judgments are based likely depend upon the task an individual is faced with and the information at hand (e.g., presence or absence of a signal).

Sustained frontal midline theta enhancements during effortful listening track working memory demands.

Recent studies demonstrate that frontal midline theta power (4-8 Hz) enhancements in the electroencephalogram (EEG) relate to effortful listening. It has been proposed that these enhancements reflect working memory demands. Here, the need to retain auditory information in working memory was manipulated in a 2-interval 2-alternative forced-choice delayed pitch discrimination task ("Which interval contained the higher pitch?"). On each trial, two square wave stimuli differing in pitch at an individual's ∼70.7% correct threshold were separated by a 3-second ISI. In a 'Roving' condition, the lowest pitch stimulus was randomly selected on each trial (uniform distribution from 840 to 1160 Hz). In a 'Fixed' condition, the lowest pitch was always 979 Hz. Critically, the 'Fixed' condition allowed one to know the correct response immediately following the first stimulus (e.g., if the first stimulus is 979 Hz, the second must be higher). In contrast, the 'Roving' condition required retention of the first tone for comparison to the second. Frontal midline theta enhancements during the ISI were only observed for the 'Roving' condition. Alpha (8-13 Hz) enhancements were apparent during the ISI, but did not differ significantly between conditions. Since conditions were matched for accuracy at threshold, results suggest that frontal midline theta enhancements will not always accompany difficult listening. Mixed results in the literature regarding frontal midline theta enhancements may be related to differences between tasks in regards to working memory demands. Alpha enhancements may reflect task general effortful listening processes.

Theta- and alpha-power enhancements in the electroencephalogram as an auditory delayed match-to-sample task becomes impossibly difficult.

Recent studies have related enhancements of theta- (∼4-8 Hz) and alpha-power (∼8-13 Hz) to listening effort based on parallels between enhancement and task difficulty. In contrast, nonauditory works demonstrate that, although increases in difficulty are initially accompanied by increases in effort, effort decreases when a task becomes so difficult as to exceed one's ability. Given the latter, we examined whether theta- and alpha-power enhancements thought to reflect effortful listening show a quadratic trend across levels of listening difficulty from impossible to easy. Listeners (n = 14) performed an auditory delayed match-to-sample task with frequency-modulated tonal sweeps under impossible, difficult (at ∼70.7% correct threshold), and easy (well above threshold) conditions. Frontal midline theta-power and posterior alpha-power enhancements were observed during the retention interval, with greatest enhancement in the difficult condition. Independent component-based analyses of data suggest that theta-power enhancements stemmed from medial frontal sources at or near the anterior cingulate cortex, whereas alpha-power effects stemmed from occipital cortices. Results support the notion that theta- and alpha-power enhancements reflect effortful cognitive processes during listening, related to auditory working memory and the inhibition of task-irrelevant cortical processing regions, respectively. Theta- and alpha-power dynamics can be used to characterize the cognitive processes that make up effortful listening, including qualitatively different types of listening effort.

Benefits of fading in perceptual learning are driven by more than dimensional attention.

Individuals learn to classify percepts effectively when the task is initially easy and then gradually increases in difficulty. Some suggest that this is because easy-to-discriminate events help learners focus attention on discrimination-relevant dimensions. Here, we tested whether such attentional-spotlighting accounts are sufficient to explain easy-to-hard effects in auditory perceptual learning. In two experiments, participants were trained to discriminate periodic, frequency-modulated (FM) tones in two separate frequency ranges (300-600 Hz or 3000-6000 Hz). In one frequency range, sounds gradually increased in similarity as training progressed. In the other, stimulus similarity was constant throughout training. After training, participants showed better performance in their progressively trained frequency range, even though the discrimination-relevant dimension across ranges was the same. Learning theories that posit experience-dependent changes in stimulus representations and/or the strengthening of associations with differential responses, predict the observed specificity of easy-to-hard effects, whereas attentional-spotlighting theories do not. Calibrating the difficulty and temporal sequencing of training experiences to support more incremental representation-based learning can enhance the effectiveness of practice beyond any benefits gained from explicitly highlighting relevant dimensions.

Predicting favorable and unfavorable consequences of perceptual learning: worsening and the peak shift.

Discrimination learning can cause improved and worsened ability to perceive differences. This subsequently affects how stimuli are associated with meanings and behaviors. Here, human listeners were trained with frequency-modulated (FM) tonal sweeps (500-1000 Hz) in a paradigm where one FM rate (8.29 octaves per second) required a 'Target' response, while a rate either slower (5.76 octaves per second) or faster (11.94 octaves per second) required a 'Non-Target' response. Training led to a shift in 'Target' responding along the FM rate dimension away from the 'Target' in a direction opposite the trained 'Non-Target'. This peak shift was paralleled by an asymmetry in acuity along the FM rate dimension in an untrained ABX task (a.k.a. match-to-sample). Performance improved relative to pre-training on trials where the 'Target' was contrasted with stimuli nearer the trained 'Non-Target'. Performance worsened on trials containing stimuli displaced along the FM dimension further from the trained 'Non-Target'. A connectionist model of perceptual learning containing non-associative representational modification and associative-based task-specific reweighting was able to simulate behavior. Simulations generated novel testable predictions regarding peak shift and worsening as a result of discrimination learning. Data have theoretical and practical consequences for predicting trends in the generalization of learned behaviors and modifiable perceptual acuities.

Indices of Effortful Listening Can Be Mined from Existing Electroencephalographic Data.

OBJECTIVES: Studies suggest that theta (~4 to 7 Hz), alpha (~8 to 12 Hz), and stimulus-evoked dynamics of the electroencephalogram index effortful listening. Numerous auditory event-related potential datasets exist, without thorough examination of these features. The feasibility of mining those datasets for such features is assessed here. DESIGN: In a standard auditory-oddball paradigm, 12 listeners heard deviant high-frequency tones (10%) interspersed among low-frequency tones (90%) "near" or "far" separated in frequency. RESULTS: During active listening (deviance detection; experiment 1), sustained frontal midline theta power, and gamma-band inter-trial phase coherence, were greater for the near condition. No significant "near"/"far" differences were observable during passive exposure to the same sounds (experiment 2). CONCLUSIONS: Increased theta power likely reflects increased utilization of cognitive-control processes (e.g., working memory) that rely on frontal cortical networks. Inter-trial phase coherence differences may reflect differences in attention-modulated stimulus encoding. Reanalysis of existing datasets can usefully inform future work on listening effort.