Development of Phase Locking and Frequency Representation in the Infant Frequency-Following Response

Longitudinal trajectories of the neural encoding mechanisms of speech-sound features during the first year of life

Infants quickly recognize the sounds of their mother language, perceiving the spectrotemporal acoustic features of speech. However, the underlying neural machinery remains unclear. We used an auditory evoked potential termed frequency-following response (FFR) to unravel the neural encoding maturation for two speech sound characteristics: voice pitch and temporal fine structure. 37 healthy-term neonates were tested at birth and retested at the ages of six and twelve months. Results revealed a reduction in neural phase-locking onset to the stimulus envelope from birth to six months, stabilizing by twelve months. While neural encoding of voice pitch remained consistent across ages, temporal fine structure encoding matured rapidly from birth to six months, without further improvement from six to twelve months. Results highlight the critical importance of the first six months of life in the maturation of neural encoding mechanisms that are crucial for phoneme discrimination during early language acquisition.

The frequency-following response in late preterm neonates: a pilot study

Introduction

Infants born very early preterm are at high risk of language delays. However, less is known about the consequences of late prematurity. Hence, the aim of the present study is to characterize the neural encoding of speech sounds in late preterm neonates in comparison with those born at term.

Methods

The speech-evoked frequency-following response (FFR) was recorded to a consonant-vowel stimulus /da/ in 36 neonates in three different groups: 12 preterm neonates [mean gestational age (GA) 36.05 weeks], 12 "early term neonates" (mean GA 38.3 weeks), and "late term neonates" (mean GA 41.01 weeks).

Results

From the FFR recordings, a delayed neural response and a weaker stimulus F0 encoding in premature neonates compared to neonates born at term was observed. No differences in the response time onset nor in stimulus F0 encoding were observed between the two groups of neonates born at term. No differences between the three groups were observed in the neural encoding of the stimulus temporal fine structure.

Discussion

These results highlight alterations in the neural encoding of speech sounds related to prematurity, which were present for the stimulus F0 but not for its temporal fine structure.

Developmental Trajectory of the Frequency-Following Response During the First 6 Months of Life

PURPOSE

The aim of the present study is to characterize the maturational changes during the first 6 months of life in the neural encoding of two speech sound features relevant for early language acquisition: the stimulus fundamental frequency (fo), related to stimulus pitch, and the vowel formant composition, particularly F1. The frequency-following response (FFR) was used as a snapshot into the neural encoding of these two stimulus attributes.

METHOD

FFRs to a consonant-vowel stimulus /da/ were retrieved from electroencephalographic recordings in a sample of 80 healthy infants (45 at birth and 35 at the age of 1 month). Thirty-two infants (16 recorded at birth and 16 recorded at 1 month) returned for a second recording at 6 months of age.

RESULTS

Stimulus fo and F1 encoding showed improvements from birth to 6 months of age. Most remarkably, a significant improvement in the F1 neural encoding was observed during the first month of life.

CONCLUSION

Our results highlight the rapid and sustained maturation of the basic neural machinery necessary for the phoneme discrimination ability during the first 6 months of age.

Experience-Dependent Plasticity in Nucleus Laminaris of the Barn Owl

Barn owls experience increasing interaural time differences (ITDs) during development, because their head width more than doubles in the month after hatching. We therefore hypothesized that their ITD detection circuit might be modified by experience. To test this, we raised owls with unilateral ear inserts that delayed and attenuated the acoustic signal, then measured the ITD representation in the brainstem nucleus laminaris (NL) when they were adult. The ITD circuit is composed of delay line inputs to coincidence detectors, and we predicted that plastic changes would lead to shorter delays in the axons from the manipulated ear, and complementary shifts in ITD representation on the two sides. In owls that received ear inserts starting around P14, the maps of ITD shifted in the predicted direction, but only on the ipsilateral side, and only in those tonotopic regions that had not experienced auditory stimulation prior to insertion. The contralateral map did not change. Experience-dependent plasticity of the ITD circuit occurs in NL, and our data suggest that ipsilateral and contralateral delays are independently regulated. Thus, altered auditory input during development leads to long-lasting changes in the representation of ITD.

Performance of Statistical Indicators in the Objective Detection of Speech-Evoked Envelope Following Responses

OBJECTIVES

To assess the sensitivity of statistical indicators used for the objective detection of speech-evoked envelope following responses (EFRs) in infants and adults.

DESIGN

Twenty-three adults and 21 infants with normal hearing participated in this study. A modified/susa∫i/speech token was presented at 65 dB SPL monoaurally. Presentation level in infants was corrected using in-ear measurements. EFRs were recorded between high forehead and ipsilateral mastoid. Statistical post-processing was completed using F -test, Magnitude-Square Coherence, Rayleigh test, Rayleigh-Moore test, and Hotelling's T 2 test. Logistic regression models assessed the sensitivity of each statistical indicator in both infants and adults as a function of testing duration.

RESULTS

The Rayleigh-Moore and Rayleigh tests were the most sensitive statistical indicators for speech-evoked EFR detection in infants. Comparatively, Magnitude-Square Coherence and Hotelling's T 2 also provide clinical benefit for infants in all conditions after ~30 minutes of testing, whereas the F -test failed to detect responses to EFRs elicited by vowels with accuracy greater than chance. In contrast, the F-test was the most sensitive for vowel-elicited response detection for adults in short tests (<10 minute) and performed comparatively with the Rayleigh-Moore and Rayleigh test during longer test durations. Decreased sensitivity was observed in infants relative to adults across all testing durations and statistical indicators, but the effects were largest in low frequency stimuli and seemed to be mostly, but not wholly, caused by differences in response amplitude.

CONCLUSIONS

The choice of statistical indicator significantly impacts the sensitivity of speech-evoked EFR detection. In both groups and for all stimuli, the Rayleigh test and Rayleigh-Moore tests have high sensitivity. Differences in EFR detection are present between infants and adults regardless of statistical indicator; however, these effects are largest for low-frequency EFR stimuli and for amplitude-based statistical indicators.

Speech-Evoked Envelope Following Responses in Children and Adults

PURPOSE

Envelope following responses (EFRs) could be useful for objectively evaluating audibility of speech in children who are unable to participate in routine clinical tests. However, relative to adults, the characteristics of EFRs elicited by frequency-specific speech and their utility in predicting audibility in children are unknown.

METHOD

EFRs were elicited by the first (F1) and second and higher formants (F2+) of male-spoken vowels /u/ and /i/ and by fricatives /ʃ/ and /s/ in the token /suʃi/ presented at 15, 35, 55, 65, and 75 dB SPL. The F1, F2+, and fricatives were low-, mid-, and high-frequency dominant, respectively. EFRs were recorded between the vertex and the nape from twenty-three 6- to 17-year-old children and 21 young adults with normal hearing. Sensation levels of stimuli were estimated based on behavioral thresholds.

RESULTS

In children, amplitude decreased with age for /ʃ/-elicited EFRs but remained stable for low- and mid-frequency stimuli. As a group, EFR amplitude and phase coherence did not differ from that of adults. EFR sensitivity (proportion of audible stimuli detected) and specificity (proportion of inaudible stimuli not detected) did not vary between children and adults. Consistent with previous work, EFR sensitivity increased with stimulus frequency and level. The type of statistical indicator used for EFR detection did not influence accuracy in children.

CONCLUSIONS

Adultlike EFRs in 6- to 17-year-old typically developing children suggest mature envelope encoding for low- and mid-frequency stimuli. EFR sensitivity and specificity in children, when considering a wide range of stimulus levels and audibility, are ~77% and ~92%, respectively.

SUPPLEMENTAL MATERIAL

https://doi.org/10.23641/asha.21136171.

Implementation of Machine Learning on Human Frequency-Following Responses: A Tutorial

The frequency-following response (FFR) provides enriched information on how acoustic stimuli are processed in the human brain. Based on recent studies, machine learning techniques have demonstrated great utility in modeling human FFRs. This tutorial focuses on the fundamental principles, algorithmic designs, and custom implementations of several supervised models (linear regression, logistic regression, k -nearest neighbors, support vector machines) and an unsupervised model ( k -means clustering). Other useful machine learning tools (Markov chains, dimensionality reduction, principal components analysis, nonnegative matrix factorization, and neural networks) are discussed as well. Each model's applicability and its pros and cons are explained. The choice of a suitable model is highly dependent on the research question, FFR recordings, target variables, extracted features, and their data types. To promote understanding, an example project implemented in Python is provided, which demonstrates practical usage of several of the discussed models on a sample dataset of six FFR features and a target response label.

Neonatal Frequency-Following Responses: A Methodological Framework for Clinical Applications

The frequency-following response (FFR) to periodic complex sounds is a noninvasive scalp-recorded auditory evoked potential that reflects synchronous phase-locked neural activity to the spectrotemporal components of the acoustic signal along the ascending auditory hierarchy. The FFR has gained recent interest in the fields of audiology and auditory cognitive neuroscience, as it has great potential to answer both basic and applied questions about processes involved in sound encoding, language development, and communication. Specifically, it has become a promising tool in neonates, as its study may allow both early identification of future language disorders and the opportunity to leverage brain plasticity during the first 2 years of life, as well as enable early interventions to prevent and/or ameliorate sound and language encoding disorders. Throughout the present review, we summarize the state of the art of the neonatal FFR and, based on our own extensive experience, present methodological approaches to record it in a clinical environment. Overall, the present review is the first one that comprehensively focuses on the neonatal FFRs applications, thus supporting the feasibility to record the FFR during the first days of life and the predictive potential of the neonatal FFR on detecting short- and long-term language abilities and disruptions.

Non-negative matrix factorization improves the efficiency of recording frequency-following responses in normal-hearing adults and neonates

OBJECTIVE

One challenge in extracting the scalp-recorded frequency-following response (FFR) is related to its inherently small amplitude, which means that the response cannot be identified with confidence when only a relatively small number of recording sweeps are included in the averaging procedure.

DESIGN

This study examined how the non-negative matrix factorisation (NMF) algorithm with a source separation constraint could be applied to improve the efficiency of FFR recordings. Conventional FFRs elicited by an English vowel/i/with a rising frequency contour were collected. Study sample: Fifteen normal-hearing adults and 15 normal-hearing neonates were recruited.

RESULTS

The improvements of FFR recordings, defined as the correlation coefficient and root-mean-square differences across a sweep series of amplitude spectrograms before and after the application of the source separation NMF (SSNMF) algorithm, were characterised through an exponential curve fitting model. Statistical analysis of variance indicated that the SSNMF algorithm was able to enhance the FFRs recorded in both groups of participants.

CONCLUSIONS

Such improvements enabled FFR extractions in a relatively small number of recording sweeps, and opened a new window to better understand how speech sounds are processed in the human brain.

The emergence of idiosyncratic patterns in the frequency-following response during the first year of life

The frequency-following response (FFR) is a scalp-recorded signal that reflects phase-locked activity from neurons across the auditory system. In addition to capturing information about sounds, the FFR conveys biometric information, reflecting individual differences in auditory processing. To investigate the development of FFR biometric patterns, we trained a pattern recognition model to recognize infants (N = 16) from FFRs collected at 7 and 11 months. Model recognition scores were used to index the robustness of FFR biometric patterns at each time. Results showed better recognition scores at 11 months, demonstrating the emergence of robust FFR idiosyncratic patterns during this first year of life.

Deficient neural encoding of speech sounds in term neonates born after fetal growth restriction

Infants born after fetal growth restriction (FGR)-an obstetric condition defined as the failure to achieve the genetic growth potential-are prone to neurodevelopmental delays, with language being one of the major affected areas. Yet, while verbal comprehension and expressive language impairments have been observed in FGR infants, children and even adults, specific related impairments at birth, such as in the ability to encode the sounds of speech, necessary for language acquisition, remain to be disclosed. Here, we used the frequency-following response (FFR), a brain potential correlate of the neural phase locking to complex auditory stimuli, to explore the encoding of speech sounds in FGR neonates. Fifty-three neonates born with FGR and 48 controls born with weight adequate-for-gestational age (AGA) were recruited. The FFR was recorded to the consonant-vowel stimulus (/da/) during sleep and quantified as the spectral amplitude to the fundamental frequency of the syllable and its signal-to-noise ratio (SNR). The outcome was available in 45 AGA and 51 FGR neonates, yielding no differences for spectral amplitudes. However, SNR was strongly attenuated in the FGR group compared to the AGA group at the vowel region of the stimulus. These findings suggest that FGR population present a deficit in the neural pitch tracking of speech sounds already present at birth. Our results pave the way for future research on the potential clinical use of the FFR in this population, so that if confirmed, a disrupted FFR recorded at birth may help deriving FGR neonates at risk for postnatal follow-ups.

The Accuracy of Envelope Following Responses in Predicting Speech Audibility

OBJECTIVES

The present study aimed to (1) evaluate the accuracy of envelope following responses (EFRs) in predicting speech audibility as a function of the statistical indicator used for objective response detection, stimulus phoneme, frequency, and level, and (2) quantify the minimum sensation level (SL; stimulus level above behavioral threshold) needed for detecting EFRs.

DESIGN

In 21 participants with normal hearing, EFRs were elicited by 8 band-limited phonemes in the male-spoken token /susa∫i/ (2.05 sec) presented between 20 and 65 dB SPL in 15 dB increments. Vowels in /susa∫i/ were modified to elicit two EFRs simultaneously by selectively lowering the fundamental frequency (f0) in the first formant (F1) region. The modified vowels elicited one EFR from the low-frequency F1 and another from the mid-frequency second and higher formants (F2+). Fricatives were amplitude-modulated at the average f0. EFRs were extracted from single-channel EEG recorded between the vertex (Cz) and the nape of the neck when /susa∫i/ was presented monaurally for 450 sweeps. The performance of the three statistical indicators, F-test, Hotelling's T, and phase coherence, was compared against behaviorally determined audibility (estimated SL, SL ≥0 dB = audible) using area under the receiver operating characteristics (AUROC) curve, sensitivity (the proportion of audible speech with a detectable EFR [true positive rate]), and specificity (the proportion of inaudible speech with an undetectable EFR [true negative rate]). The influence of stimulus phoneme, frequency, and level on the accuracy of EFRs in predicting speech audibility was assessed by comparing sensitivity, specificity, positive predictive value (PPV; the proportion of detected EFRs elicited by audible stimuli) and negative predictive value (NPV; the proportion of undetected EFRs elicited by inaudible stimuli). The minimum SL needed for detection was evaluated using a linear mixed-effects model with the predictor variables stimulus and EFR detection p value.

RESULTS

of the 3 statistical indicators were similar; however, at the type I error rate of 5%, the sensitivities of Hotelling's T (68.4%) and phase coherence (68.8%) were significantly higher than the F-test (59.5%). In contrast, the specificity of the F-test (97.3%) was significantly higher than the Hotelling's T (88.4%). When analyzed using Hotelling's T as a function of stimulus, fricatives offered higher sensitivity (88.6 to 90.6%) and NPV (57.9 to 76.0%) compared with most vowel stimuli (51.9 to 71.4% and 11.6 to 51.3%, respectively). When analyzed as a function of frequency band (F1, F2+, and fricatives aggregated as low-, mid- and high-frequencies, respectively), high-frequency stimuli offered the highest sensitivity (96.9%) and NPV (88.9%). When analyzed as a function of test level, sensitivity improved with increases in stimulus level (99.4% at 65 dB SPL). The minimum SL for EFR detection ranged between 13.4 and 21.7 dB for F1 stimuli, 7.8 to 12.2 dB for F2+ stimuli, and 2.3 to 3.9 dB for fricative stimuli.

CONCLUSIONS

EFR-based inference of speech audibility requires consideration of the statistical indicator used, phoneme, stimulus frequency, and stimulus level.

Effect of Auditory Maturation on the Encoding of a Speech Syllable in the First Days of Life

(1) Background: In neonates and infants, the physiological modifications associated with language development are reflected in their Frequency Following Responses (FFRs) in the first few months of life. (2) Objective: This study aimed to test the FFRs of infants in the first 45 days of life in order to evaluate how auditory maturation affects the encoding of a speech syllable. (3) Method: In total, 80 healthy, normal-hearing infants, aged 3 to 45 days old, participated in this study. The sample was divided into three groups: GI, 38 neonates from 3 to 15 days; GII, 25 infants from 16 to 30 days; and GIII, 17 infants from 31 to 45 days. All participants underwent FFR testing. Results: With age, there was a decrease in the latency of all FFR waves, with statistically significant differences among the groups studied for waves V, A, E, F, and O. The mean amplitudes showed an increase, with a statistically significant difference only for wave V. The slope measure increased over the 45 days, with a statistically significant difference between GIII and GI and between GIII and GII. (4) Conclusions: The encoding of a speech sound changes with auditory maturation over the first 45 days of an infant’s life.

Frequency-Following Response in Newborns and Infants: A Systematic Review of Acquisition Parameters

Purpose The purpose of this study is to characterize parameters used for frequency-following response (FFR) acquisition in children up to 24 months of age through a systematic review. Method The study was registered in PROSPERO and followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses' recommendations. Search was performed in six databases (LILACS, LIVIVO, PsycINFO, PubMed, Scopus, and Web of Science) and gray literature (Google Scholar, OpenGrey, ProQuest)as well as via manual searches in bibliographic references. Observational studies using speech stimuli to elicit the FFR in infants with normal hearing on the age range from 0 until 24 months were included. No restrictions regarding language and year of publication were applied. Risk of bias was assessed with the Joanna Briggs Institute Critical Appraisal Checklist. Data on stimulus, presentation rate, time window for analysis, number of sweeps, artifact rejection, online filters, stimulated ear, and examination condition were extracted. Results Four hundred fifty-nine studies were identified. After removing duplicates and reading titles and abstracts, 15 articles were included. Seven studies were classified as low risk of bias, seven as moderate risk, and one as high risk. Conclusions There is a consensus in the use of some acquisition parameters of the FFR with speech stimulus, such as the vertical mounting, the use of alternating polarity, a sampling rate of 20000 Hz, and the /da/ synthesized syllable of 40 ms in duration as the preferred stimulus. Although these parameters show some consensus, the results disclosed lack of a single established protocol for FFR acquisition with speech stimulus in infants in the investigated age range.

Relationships between click auditory brainstem response and speech frequency following response with development in infants born preterm

The speech evoked frequency following response (sFFR) is used to study relationships between neural processing and functional aspects of speech and language that are not captured by click or toneburst evoked auditory brainstem responses (ABR). The sFFR is delayed, deviant, or weak in school age children having a variety of disorders, including autism, dyslexia, reading and language disorders, in relation to their typically developing peers. Much less is known about the developmental characteristics of sFFR, especially in preterm infants, who are at risk of having language delays. In term neonates, phase locking and spectral representation of the fundamental frequency is developed in the early days of life. Spectral representation of higher harmonics and latencies associated with transient portions of the stimulus are still developing in term infants through at least 10 months of age. The goal of this research was to determine whether sFFR could be measured in preterm infants and to characterize its developmental trajectory in the time and frequency domain. Click ABR and sFFR were measured in 28 preterm infants at ages 33 to 64 weeks gestational age. The sFFR could be measured in the majority of infants at 33 weeks gestational age, and the detectability of all sFFR waves was 100% by 64 weeks gestational age. The latency of all waves associated with the transient portion of the response (waves V, A, and O), and most waves (waves D and E) associated with the quasi-steady state decreased with increasing age. The interpeak wave A-O latency did not change with age, indicating that these waves share a neural generator, or the neural generators are developing at the same rate. The spectral amplitude of F₀ and the lower frequencies of the first formant increased with age, but that for higher frequencies of the first formant and higher harmonics did not. The results suggest that the sFFR can be reliably recorded in preterm infants, including those cared for in the neonatal intensive care unit. These findings support that in preterm infants, F₀ amplitude continues to develop within the first 6 months of life and develops before efficient representation of higher frequency harmonics. Further research is needed to determine if the sFFR in preterm infants is predictive of long-term language or learning disorders.

Characteristics of Speech-Evoked Envelope Following Responses in Infancy

Envelope following responses (EFRs) may be a useful tool for evaluating the audibility of speech sounds in infants. The present study aimed to evaluate the characteristics of speech-evoked EFRs in infants with normal hearing, relative to adults, and identify age-dependent changes in EFR characteristics during infancy. In 42 infants and 21 young adults, EFRs were elicited by the first (F1) and the second and higher formants (F2+) of the vowels /u/, /a/, and /i/, dominant in low and mid frequencies, respectively, and by amplitude-modulated fricatives /s/ and /∫/, dominant in high frequencies. In a subset of 20 infants, the in-ear stimulus level was adjusted to match that of an average adult ear (65 dB sound pressure level [SPL]). We found that (a) adult-infant differences in EFR amplitude, signal-to-noise ratio, and intertrial phase coherence were larger and spread across the frequency range when in-ear stimulus level was adjusted in infants, (b) adult-infant differences in EFR characteristics were the largest for low-frequency stimuli, (c) infants demonstrated adult-like phase coherence when they received a higher (i.e., unadjusted) stimulus level, and (d) EFR phase coherence and signal-to-noise ratio changed with age in the first year of life for a few F2+ vowel stimuli in a level-specific manner. Together, our findings reveal that development-related changes in EFRs during infancy likely vary by stimulus frequency, with low-frequency stimuli demonstrating the largest adult-infant differences. Consistent with previous research, our findings emphasize the significant role of stimulus level calibration methods while investigating developmental trends in EFRs.