Frequency following responses to tone glides: Effects of frequency extent, direction, and electrode montage

Abstract

The spectral (frequency) and amplitude cues in speech change rapidly over time. Study of the neural encoding of these dynamic features may help to improve diagnosis and treatment of speechperception difficulties. This study uses tone glides as a simple approximation of dynamic speech sounds to better our understanding of the underlying neural representation of speech. The frequency following response (FFR) was recorded from 10 young normal-hearing adults using six signals varying in glide direction (rising and falling) and extent of frequency change ( 1 3 , 2 3 , and 1 octave). In addition, the FFR was simultaneously recorded using two different electrode montages (vertical and horizontal). These factors were analyzed across three time windows using a measure of response strength (signal-to-noise ratio) and a measure of temporal coherence (stimulus-toresponse correlation coefficient). Results demonstrated effects of extent, montage, and a montageby-window interaction. SNR and stimulus-to-response correlation measures differed in their sensitivity to these factors. These results suggest that the FFR reflects dynamic acoustic characteristics of simple tonal stimuli very well. Additional research is needed to determine how neural encoding may differ for more natural dynamic speech signals and populations with impaired auditory processing.

Neural tracking of the fundamental frequency of the voice: the effect of voice characteristics

“…larger correlations for lower f0 and lower rate of f0 change. This observation is in line with the behaviour of classic EFRs which have a lower amplitude for higher stimulus frequencies and rapid frequency changes (Purcell et al, 2004;Billings et al, 2019;Van Canneyt et al, 2020). One major cause for this is that it is harder for neurons to phase-lock when the stimulus frequency is high and highly variable, resulting in weaker population responses.…”

“…Cortical contributions to f0-tracking are not unlikely as the more meaningful and less predictable stimuli could solicit more top-down processing than the traditional EFR stimuli. The second finding is that higher fundamental frequencies and higher rates of fundamental frequency change in the stimulus result in smaller EFRs (Purcell et al, 2004;Billings et al, 2019;Van Canneyt et al, 2020). We aim to investigate whether similar effects are likely to occur for f0-tracking.…”

Spectrally specific temporal analyses of spike-train responses to complex sounds: A unifying framework

“…Speech is a nonstationary signal and conveys substantial information in its dynamic spectral trajectories (e.g., Fig 1A ). A number of studies have investigated the robustness of the neural representation of dynamic spectral trajectories using frequency glides and frequency-modulated tones as the stimulus [ 61 – 64 ]. These studies have usually employed a spectrogram analysis.…”

From Modulated Noise to Natural Speech: the Effect of Stimulus Parameters on the Frequency Following Response

“…In studies that evoke FFRs with sweeping pure tones, less robust FFRs are found for higher rates of change, i.e. in the range of 900-6600 Hz/s (Billings et al, 2019; Clinard and Cotter, 2015). Another possibility is that the reference for the Fourier Analyzer represented these rapid frequency changes less precisely, impairing response analysis.…”

“…These simple stimuli are often unnatural, but they allow for precise manipulation of the acoustic cues. Examples are tonebursts (Clinard et al, 2010; Gardi et al, 1979; Glaser et al, 1976; Tichko and Skoe, 2017), pure tones (Gockel et al, 2015; Holmes et al, 2018), tone sweeps (Billings et al, 2019; Clinard and Cotter, 2015; Krishnan and Parkinson, 2000; Purcell et al, 2004) and (sinusoidally) amplitude modulated (AM) stimuli (Bidelman and Patro, 2016; Dimitrijevic et al, 2016; Van Canneyt et al, 2019). It is unclear whether findings for these non-speech stimuli can be generalized to responses evoked by speech stimuli.…”