Auditory Brainstem Response elicited by CE chirps and an investigation of their time-frequency properties

Abstract

Chirps are a family of stimuli designed to provide enhanced neural synchronicity and hence improve ABR amplitudes. It is known that NB CE chirps can achieve greater wave V amplitude waveforms compared to the traditional tone pip (Ferm et al 2013, 2015; Rodrigues et al 2013; Cobb et al 2016). However, it is unclear whether this is due to the wider bandwidth of the NB CE chirp or the frequency-delay characteristics of the stimuli. The aim of this study was to evaluate CE chirps and investigate their time-frequency properties in three ways: 1. comparing the acoustic spectrum of the NB CE chirps with traditional tone pips 2. analysing the time-frequency properties of the NB CE chirps through spectrograms and 3. measurement of wave V amplitudes stimulated by forward and reverse CE chirps.

The acoustical spectral bandwidth for CE chirp stimuli and Blackman-gated tone pips were compared at 500 Hz and 4 kHz. The 500 Hz CE chirp was 43% wider than the 500 Hz tone pip and the 4 kHz CE chirp was 40% wider than the 4 kHz tone pip. The widened acoustic spectra for the NB CE chirp could be a contributing factor to the increased wave V amplitude shown for NB CE chirps over tone pips in clinical practice.

Spectrograms were generated to assess the time-frequency properties of the chirp that should show a nonlinearly sweep in time from low to high frequency compared to a time reversed chirp that should show the opposite. This frequency sweep appears clear for 4 kHz and 2 kHz NB CE chirps, 1 kHz NB CE chirp frequency sweep appears to be on the limit of being resolved and the 500 Hz CE chirp sweep could not be resolved by spectrogram analysis. The stimuli were also tested in an auditory nerve model. From the model outputs it appears that the lowest frequency chirps may not be well resolved by the auditory system.

An issue with producing effective band limited chirps is that the short stimulus must be frequency specific but still follow the frequency-delay function of the cochlea by sweeping in frequency over time. Dau (2000) and Petoe (2010) both showed that a time reversed broadband chirp produces a smaller wave V ABR response compared to a forward broadband chirp, therefore showing that the time-frequency properties of broadband chirp can be resolved by the cochlea. With the NB CE chirps we may be near the limit of time-frequency resolution but a comparison of forward and reverse NB CE chirps has not been performed previously. An initial experiment explored this comparison, but due to limitations in the data quality obtained, the results should be treated with caution. For Experiment 2, data quality was improved by monitoring Fmp and residual noise levels, using Bayesian weighting and increasing our cohort of participants. ABRs were recorded in 28 normal hearing adults in response to forward and time-reverse broadband and 0.5, 1, 2 and 4 kHz NB CE chirp stimuli at 50 dB HL using ER-3A insert earphones. ABR traces were recorded on the Interacoustic Eclipse and 20 subjects met an inclusion criterion for data quality. There was a small but significant increase in wave V ABR amplitudes when comparing forward and reverse CE chirps at 1, 2 and 4 kHz (p<0.005). There was a larger effect for the broadband CE chirp (p<0.005) but no significant effect at 500 Hz CE chirp (p=0.814). Overall it appears that there is an effect of the forward over the reversed chirps for one-octave wide 1, 2 and 4 kHz NB CE chirps. However, there was not a benefit of direction for the 500 Hz CE chirp, so this stimulus may not be resolved as a chirp by the cochlea.

Finally, we explored the effect of high pass filter setting on the quality of ABR responses elicited by broadband and NB CE chirps, using high pass filter settings ranging from 10 – 100 Hz. There was a statistically significant reduction in the ABR wave V amplitude response with increasing high pass filter setting. From the results it appears that energy in the frequency band 10-100 Hz is responsible for approximately half of wave V ABR amplitudes.

More information

Identifiers

Catalogue record

Export record