Biology subjects

de Cheveigne, A.

Publications and source records attributed to de Cheveigne, A..

At least 19 recordsLinked to original sources

bioRxiv2025

Rapid FFR: A rapid method for obtaining Frequency Following Responses

ObjectivesFrequency-Following Responses (FFRs) are typically recorded using stimuli of 40-250 ms in duration separated by silent intervals. Because robust FFRs require averaging across approximately 3000 stimulus repetitions, conventional acquisition is time-intensive. We introduce the Rapid FFR, a technique that minimises recording time by presenting the stimulus continuously (i.e., without inter-stimulus intervals) and by averaging across individual response cycles. This study aimed to: (a) compare Rapid and Conventional FFR performance under time-matched conditions; (b) assess test-retest reliability of both methods, and (c) examine potential neural adaptation during extended Rapid FFR recordings. DesignAll participants (37 in total) were young adults with clinically normal hearing. In Experiment 1, FFRs were elicited from 16 listeners using a 128 Hz sawtooth wave presented in two ways: (1) continuously over approximately 1 minute in each polarity for the Rapid FFR and (2) as discrete tone bursts (1500 trials per polarity) for the standard FFR. Signal-to-Noise ratios (SNRs), response amplitudes, and test-retest reliability were compared across techniques. In Experiment 2, the Rapid FFR was recorded continuously from 21 listeners for nearly nine minutes to assess potential adaptation over time. ResultsThe Rapid FFR produced significantly higher SNRs than the standard FFR, reflecting improved recording efficiency. Both techniques captured inter-individual differences reliably, with comparable frequency-domain response patterns across participants. In particular, measures derived from the Rapid FFR showed strong agreement with those from the standard FFR for lower harmonics (F0-H3). Higher harmonics (H4-H7) exhibited greater variability but remained consistent between techniques. In the prolonged recording condition, FFR amplitudes remained stable over time, with no significant decline across the nine minute recording. This indicates that continuous stimulation did not result in measurable neural adaptation or response fatigue. ConclusionsThe Rapid FFR offers a time-efficient alternative to standard protocols, preserving signal fidelity and reliability while enabling shorter acquisition times. These findings suggest that the Rapid FFR is well-suited for use in populations where long recordings are challenging (e.g., in infants and the clinic) and can facilitate more extensive experimental designs within a single session. The method holds promise for advancing both research and clinical applications of the FFR. Future studies should explore its use across a broader range of sounds (in particular, dynamically varying ones) and listener groups.

neuroscience↗

bioRxiv2024

What and where in electromagnetic brain imaging

To understand the brain, we may record electromagnetic signals and use them to infer the dynamics of brain activity (the "what"), and the location or distribution of its sources within the brain (the "where"). In this paper, these two aspects emerge from the experimental data via distinct analysis pathways. In a first step, recordings of brain activity undergo linear data-driven analysis that results in an analysis matrix, the first column of which isolates a source of interest. In a second step, the remaining columns of the analysis matrix are used to estimate the position of the source, with the help of a source model informed by the anatomical scans and sensor geometry. Specifically, each column is interpreted as a null filter for the source of interest (a spatial filter that suppresses that source). Each such null filter admits a zero set, or subset of the source parameter space for which the filter output is zero. Since the source of interest is necessarily within the zero sets of all of its null filters, its location can be found by taking the intersection of these zero sets. There is no theoretical limit to location accuracy, but practical limits may arise from noise in the data, imperfect calibration, or an incomplete or inaccurate source model. The method is illustrated with simulated and real data.

neuroscience↗

bioRxiv2024

Real-time control of a hearing instrument with EEG-based attention decoding

Enhancing speech perception in everyday noisy acoustic environments remains an outstanding challenge for hearing aids. Speech separation technology is improving rapidly, but hearing devices cannot fully exploit this advance without knowing which sound sources the user wants to hear. Even with high-quality source separation, the hearing aid must know which speech streams to enhance and which to suppress. Advances in EEG-based decoding of auditory attention raise the potential of neurosteering, in which a hearing instrument selectively enhances the sound sources that a hearing-impaired listener is focusing their attention on. Here, we present and discuss a real-time brain-computer interface (BCI) system that combines a stimulus-response model based on canonical correlation analysis (CCA) for real-time EEG attention decoding, coupled with a multi-microphone hardware platform enabling low-latency real-time speech separation through spatial beamforming. We provide an overview of the system and its various components, discuss prospects and limitations of the technology, and illustrate its application with case studies of listeners steering acoustic feedback of competing speech streams via real-time attention decoding. A software implementation code of the system is publicly available for further research and explorations.

neuroscience↗

bioRxiv2024

Predictive coding in the auditory brainstem

Predictive coding is an influential concept in sensory and cognitive neuroscience. It is often understood as involving top-down prediction of bottom-up sensory patterns, but the term also applies to feed-forward predictive mechanisms, for example in the retina. Here, I discuss a recent model of low-level predictive processing in the auditory brainstem that is of the feed-forward flavor. Past sensory input from the cochlea is delayed and compared with the current input, with a delay that is tuned with the objective of minimizing prediction error. This operation can be performed uniformly across peripheral channels, or independently within each peripheral channel with parameters local to that channel. The result is a sensory representation that is invariant to a certain class of masking sounds (harmonic, quasi-harmonic, or spectrally sparse), thus contributing to Auditory Scene Analysis. The purpose of this paper is to discuss that model in the light of predictive coding, and examine how it might fit within a wider hierarchical model that supports the perceptual representation of objects and events in the world.

neuroscience↗

bioRxiv2023

Virtual electrode, or virtual scalpel?

A spatial filter is a set of weights which, applied to multichannel data, creates a "virtual channel" with desirable properties. Examples range from simple hard-wired spatial filters (re-referencing, gradient, Laplacian, etc.) to more complex data-driven transforms (beamforming, independent component analysis, etc.). The principle is straightforward, but the properties are obscured by the high dimensionality of the data, and the various "spaces" in which a "spatial" filter operates, inhabited by sources, sensors, fields, or signals. This paper reviews the properties and limits of spatial filters, with particular focus on the popular concept of a "virtual electrode" created by combining channels of a non-invasive recording technique such as EEG or MEG. Whereas a spatial filter can perfectly suppress one or more sources (as many as there are sensors, minus one), it cannot select one source and suppress all others, as might a real electrode. This puts hard limits on what to expect of a virtual electrode, and suggests a slightly different perspective, that of a "virtual scalpel". Spatial filtering, like temporal filtering, plays an important role in brain data analysis. Significance statementSpatial filters are ubiquitous in brain data analysis. This paper reviews their properties and the limits of what can be achieved, with emphasis on aspects that are obscured by the complex geometry of sources and sensors, and the high dimensionality of the data.

neuroscience↗

bioRxiv2023

Is EEG best left alone?

A recent article (Delorme, A., 2023, EEG is better left alone. Scientific Reports, 13, 2372. https://doi.org/10.1038/s41598-023-27528-0) proposed a metric to determine the benefit of applying pre-processing methods to EEG data. Using that metric, it concluded that most pre-processing methods do not improve data quality. The question is revisited in this short paper, using a sample of the same data as used by that paper. It is argued that the metric is of limited applicability, and that in some situations pre-processing might be critical to make good use of the data. Note: This paper was submitted to Scientific Reports as a commentary to the paper cited and rejected without review.

neuroscience↗

bioRxiv2023

Simple and efficient auditory-nerve spike generation

This brief paper documents a simple and efficient method to generate auditory-nerve spike trains for the purpose of simulating neural processes of auditory perception. In response to sound, each auditory nerve fiber carries information to the auditory brainstem in the form of a train of spikes (action potentials), the timing and rate of which reflect the sound. The generation process is usually approximated as Poisson process with a time-varying rate, further modified by refractory effects. The purpose, here, is to simulate spike generation as a time- and interval-dependent thinning process applied to a homogenous Poisson process, allowing for fast generation, cheap storage, and unlimited temporal resolution.

neuroscience↗

bioRxiv2022

Why is the perceptual octave stretched? An account based on mismatched time constants within the auditory brainstem.

This paper suggests an explanation for listeners greater tolerance to positive than negative mistuning of the higher tone within an octave pair. It hypothesizes a neural circuit tuned to cancel the lower tone, that also cancels the higher tone if that tone is in tune. Imperfect cancellation is the cue to mistuning of the octave. The circuit involves two neural pathways, one delayed with respect to the other, that feed a coincidence-sensitive neuron via excitatory and inhibitory synapses. A mismatch between the time constants of these two synapses results in an asymmetry in sensitivity to mismatch. Specifically, if the time constant of the delayed pathway is greater than that of the direct pathway, there is a greater tolerance to positive mistuning than to negative mistuning. The model is directly applicable to the harmonic octave (concurrent tones), but extending it to the melodic octave (successive tones) requires additional assumptions that are discussed. The paper reviews evidence from auditory psychophysics and physiology in favor - or against - this explanation.

neuroscience↗

bioRxiv2022

In-Channel Cancellation

A model of early auditory processing is proposed in which each peripheral channel is processed by a delay-and-subtract cancellation filter, tuned independently for each channel with a criterion of minimum power. For a channel dominated by a pure tone or a resolved partial of a complex tone, the optimal delay is its period. For a channel responding to harmonically-related partials, the optimal delay is their common fundamental period. Each peripheral channel is thus split into two subchannels, one that is cancellation-filtered and the other not. Perception can involve either or both, depending on the task. The model is illustrated by applying it to the masking asymmetry between pure tones and narrowband noise: a noise target masked by a tone is more easily detectable than a tone target masked by noise. The model is one of a wider class of models, monaural or binaural, that cancel irrelevant stimulus dimensions so as to attain invariance to competing sources. Similar to occlusion in the visual domain, cancellation yields sensory evidence that is incomplete, thus requiring Bayesian inference of an internal model of the world along the lines of Helmholtzs doctrine of unconscious inference.

neuroscience↗

bioRxiv2022

Local Subspace Pruning (LSP) for Multichannel Data Denoising

This paper proposes a simple algorithm to remove noise and artifact from multichannel data. Data are processed trial by trial: for each trial the covariance matrix of the trial is diagonalized together with that of the full data to reveal the subspace that is - locally - most eccentric relative to other trials. That subspace is then projected out from the data of that trial. This algorithm addresses a fundamental limitation of standard linear analysis methods (e.g. ICA) that assume that brain and artifact are linearly separable within the data. That assumption fails if there are more sources, including noise and brain sources, than data channels. The algorithm captitalizes on the fact that, if enough of those sources are temporally sparse, linear separation may succeed locally in time. The paper explains the rationale, describes the algorithm, and evaluates the outcome using synthetic and real brain data.

neuroscience↗

bioRxiv2020

Auditory Stimulus-response Modeling with a Match-Mismatch Task

An auditory stimulus can be related to the brain response that it evokes by a stimulus-response model fit to the data. This offers insight into perceptual processes within the brain and is also of potential use for devices such as Brain Computer Interfaces (BCI). The quality of the model can be quantified by measuring the fit with a regression problem, or by applying it to a classification task and measuring its performance. Here we focus on a match-mismatch (MM) task that entails deciding whether a segment of brain signal matches, via a model, the auditory stimulus that evoked it. The MM task allows stimulus-response models to be evaluated in the limit of very high model accuracy, making it an attractive alternative to the more commonly used task of auditory attention detection (AAD). The MM task does not require class labels, so it is immune to mislabeling, and it is applicable to data recorded in listening scenarios with only one sound source, thus it is cheap to obtain large quantities of training and testing data. Performance metrics from this task, associated with regression accuracy, provide complementary insights into the relation between stimulus and response, as well as information about discriminatory power directly applicable to BCI applications. Using these metrics, we describe a range of models of increasing complexity that we compare to methods in the literature, showing state-of-the-art performance. We document in detail one particular implementation, calibrated on a publicly-available database, that can serve as a robust reference to evaluate future developments.

neuroscience↗

bioRxiv2020

Shared Component Analysis

This paper proposes Shared Component Analysis (SCA) as an alternative to Principal Component Analysis (PCA) for the purpose of dimensionality reduction of neuroimaging data. The trend towards larger numbers of recording sensors, pixels or voxels leads to richer data, with finer spatial resolution, but it also inflates the cost of storage and computation and the risk of overfitting. PCA can be used to select a subset of orthogonal components that explain a large fraction of variance in the data. This implicitly equates variance with relevance, and for neuroimaging data such as electroencephalography (EEG) or magnetoencephalography (MEG) that assumption may be inappropriate if (latent) sources of interest are weak relative to competing sources. SCA instead assumes that components that contribute to observable signals on multiple sensors are of likely interest, as may be the case for deep sources within the brain as a result of current spread. In SCA, steps of normalization and PCA are applied iteratively, linearly transforming the data such that components more widely shared across channels appear first in the component series. The paper explains the motivation, defines the algorithm, evaluates the outcome, and sketches a wider strategy for dimensionality reduction of which this algorithm is an example. SCA is intended as a plug-in replacement for PCA for the purpose of dimensionality reduction.

neuroscience↗

bioRxiv2020

Neurophysiological indices of audiovisual speech integration are enhanced at the phonetic level for speech in noise

Seeing a speakers face benefits speech comprehension, especially in challenging listening conditions. This perceptual benefit is thought to stem from the neural integration of visual and auditory speech at multiple stages of processing, whereby movement of a speakers face provides temporal cues to auditory cortex, and articulatory information from the speakers mouth can aid recognizing specific linguistic units (e.g., phonemes, syllables). However it remains unclear how the integration of these cues varies as a function of listening conditions. Here we sought to provide insight on these questions by examining EEG responses to natural audiovisual, audio, and visual speech in quiet and in noise. Specifically, we represented our speech stimuli in terms of their spectrograms and their phonetic features, and then quantified the strength of the encoding of those features in the EEG using canonical correlation analysis. The encoding of both spectrotemporal and phonetic features was shown to be more robust in audiovisual speech responses then what would have been expected from the summation of the audio and visual speech responses, consistent with the literature on multisensory integration. Furthermore, the strength of this multisensory enhancement was more pronounced at the level of phonetic processing for speech in noise relative to speech in quiet, indicating that listeners rely more on articulatory details from visual speech in challenging listening conditions. These findings support the notion that the integration of audio and visual speech is a flexible, multistage process that adapts to optimize comprehension based on the current listening conditions. Significance StatementDuring conversation, visual cues impact our perception of speech. Integration of auditory and visual speech is thought to occur at multiple stages of speech processing and vary flexibly depending on the listening conditions. Here we examine audiovisual integration at two stages of speech processing using the speech spectrogram and a phonetic representation, and test how audiovisual integration adapts to degraded listening conditions. We find significant integration at both of these stages regardless of listening conditions, and when the speech is noisy, we find enhanced integration at the phonetic stage of processing. These findings provide support for the multistage integration framework and demonstrate its flexibility in terms of a greater reliance on visual articulatory information in challenging listening conditions.

neuroscience↗

bioRxiv2019

ZapLine: a simple and effective method to remove power line artifacts

Power line artifacts are the bane of animal and human electrophysiology. A number of methods are available to help attenuate or eliminate them, but each has its own set of drawbacks. In this brief note I present a simple method that combines the advantages of spectral and spatial filtering, while minimizing their downsides. This method is applicable to multichannel data such as electroencephalography (EEG), magnetoencephalography (MEG), or multichannel local field potentials (LFP). I briefly review past methods, pointing out their drawbacks, describe the new method, and evaluate the outcome using synthetic and real data.

neuroscience↗

bioRxiv2019

Cortical encoding of melodic expectations in human temporal cortex

Humans engagement in music rests on underlying elements such as the listeners cultural background and general interest in music, all shaping the way music is processed in the brain and perceived. Crucially, these factors modulate how listeners anticipate musical events, a process inducing instantaneous neural responses as the music confronts these expectations. Measuring such neural correlates would represent a direct window into high-level brain processing of music. Here we recorded electroencephalographic and electrocorticographic brain responses as participants listened to Bach melodies. We assessed the relative contributions of the acoustic versus melodic components of the music to the neural signal. Acoustic features included envelope and its derivative. Melodic features included information on melodic progressions (pitch) and their tempo (onsets), which were extracted from a Markov model predicting the next note based on a corpus of Western music and the preceding proximal musical context. We related the music to brain activity with a linear temporal response function, and demonstrated that cortical responses to music encode melodic expectations. Specifically, individual-subject neural signals were better predicted by a combination of acoustic and melodic expectation features than by either alone. This effect was most pronounced at response latencies up to 350ms, and in both planum temporale and Heschls gyrus. Finally, expectations of pitch and onset-time of musical notes exerted independent cortical effects, and such influences were modulated by the listeners musical expertise. Overall, this study demonstrates how the interplay of experimental and theoretical approaches can yield novel insights into the cortical encoding of melodic expectations.

neuroscience↗

bioRxiv2018

Accurate Modeling of Brain Responses to Speech

Perceptual processes can be probed by fitting stimulus-response models that relate measured brain signals such as electroencephalography (EEG) to the stimuli that evoke them. These models have also found application for the control of devices such as hearing aids. The quality of the fit, as measured by correlation, classification, or information rate metrics, indicates the value of the model and the usefulness of the device. Models based on Canonical Correlation Analysis (CCA) achieve a quality of fit that surpasses that of commonly-used linear forward and backward models. Here, we show that their performance can be further improved using several techniques that capture the time-varying and context-dependent relationships within the data, including adaptive beamforming, CCA weight optimization, and recurrent neural networks that capture the time-varying and context-dependent relationships within the data. We demonstrate these results using a match-vs-mismatch classification paradigm, in which the classifier must decide which of two stimulus samples produced a given EEG response and which is a randomly chosen stimulus sample. This task captures the essential features of the more complex auditory attention decoding (AAD) task explored in many other studies.

neuroscience↗

bioRxiv2018

Cortical responses to natural speech reflect probabilistic phonotactics

Humans comprehend speech despite the various challenges of real-world environments, such as loud noise and mispronunciation. Our auditory system is robust to these thanks to the integration of the upcoming sensory input with prior knowledge and expectations built on language-specific regularities. One such regularity regards the permissible phoneme sequences, which determine the likelihood that a word belongs to a given language (phonotactic probability; \"blick\" is more likely to be an English word than \"bnick\"). Previous research suggested that violations of these rules modulate brain evoked responses such as the N400 and the late positive complex. Yet several fundamental questions remain unresolved, especially regarding the neural encoding and integration strategy of phonotactic information. Here, we used linear modelling approaches to assess the influence of phonotactic probabilities on the brain responses to narrative speech measured with non-invasive EEG. We found that the relationship between continuous speech and EEG responses is best described when the speech descriptor includes phonotactic probabilities. This provides us with a methodology to isolate and measure the brain responses to phonotactics using natural speech at the individual subject-level. Furthermore, such low-frequency signals showed the strongest speech-EEG interactions at latencies of 100-400 ms, supporting a pre-lexical role of phonotactic information.\n\nSignificance StatementSpeech is composed of basic units, called phonemes, whose combinations comply with language-specific regularities determining whether a sequence \"sounds\" as a plausible word. Our ability to detect irregular combinations requires matching incoming sequences with our internal expectations, a process that supports speech segmentation and learning. However, the neural mechanisms underlying this phenomenon have not yet been established. Here, we examine this in the human brain using narrative speech. We identified a brain signal reflecting the likelihood that a word belongs to the language, which may offer new opportunities to investigate speech perception, learning, development, and impairment. Our data also suggest a pre-lexical role of this phenomenon, thus supporting and extending current mechanistic perspectives.

neuroscience↗

bioRxiv2018

Multiway Canonical Correlation Analysis of Brain Signals

Brain signals recorded with electroencephalography (EEG), magnetoencephalography (MEG) and related techniques often have poor signal-to-noise ratio due to the presence of multiple competing sources and artifacts. A common remedy is to average over repeats of the same stimulus, but this is not applicable for temporally extended stimuli that are presented only once (speech, music, movies, natural sound). An alternative is to average responses over multiple subjects that were presented with the same identical stimuli, but differences in geometry of brain sources and sensors reduce the effectiveness of this solution. Multiway canonical correlation analysis (MCCA) brings a solution to this problem by allowing data from multiple subjects to be fused in such a way as to extract components common to all. This paper reviews the method, offers application examples that illustrate its effectiveness, and outlines the caveats and risks entailed by the method.

neuroscience↗