Rapid-FFR: Developing a rapid method of measuring the auditory frequency following response for use in children and infants

Lead Research Organisation: University College London
Department Name: Speech Hearing and Phonetic Science

Abstract

The most commonly used electrophysiological measure in the audiological clinic is that of the click-evoked Auditory Brainstem Response (ABR). Used widely in newborn hearing screening programmes, it has the advantage of taking only 2-3 minutes to measure (once the electrodes are placed), an important consideration when testing infants and children. Although of great utility, the information one can gain in a short time from such a simple stimulus is limited, and click-evoked ABRs have mostly been used to detect hearing loss.

Recently, there been a growing interest in another auditory response, the so-called frequency following response (FFR). The FFR is a scalp-recorded electrophysiological measure which reflects sustained synchronous neural firing in response to periodic auditory stimuli along the auditory pathway. Recent evidence suggests that by varying the fundamental frequency (F0) of the stimuli, it is possible to favour neural generators at different levels of the auditory pathway, with relatively low frequency F0s leading to more cortical contributions, and higher frequency ones weighting lower level brainstem contributions. Also, with such a technique, evoking sounds can be much more acoustically complex than clicks, even including speech sounds. Because of the wider variety of acoustic attributes that can be explored, it is possible to target various speech-related features more specifically than is possible with a relatively crude click ABR. Most importantly, both human and animal studies show that the FFR and ABR measure complementary aspects of neurophysiological functioning, relating to behaviour differently, as well as in reflecting different neural populations. In short, FFRs could be a useful tool to use in auditory diagnoses in infants and children, and might allow the probing of different stages in the auditory pathway.

One important aspect of measuring FFRs that impedes their adoption in the clinic is the time it takes to measure them. FFRs are typically recorded to discrete silence-separated stimuli with a duration of 40 - 150 ms and sometimes up to as much as 2 s. Collecting a robust FFR can take up to 30 minutes as it requires approximately 2000-3000 repetitions of the stimulus, and it is often considered desirable to record both positive and negative polarities of the signal. Such measures are onerous enough in adult participants. For infants and children, the time required for FFRs poses an important obstacle to their more general use.

We have been developing a new method for measuring FFRs, in which the stimulus is presented continuously (i.e., without an interstimulus interval) and averaging across a single cycle of the response. For the particular stimulus we used, FFR recording times were reduced from approximately 2.5 minutes for the standard technique to only 35 seconds using our new one (Rapid-FFR), with comparable data quality. This reduction in acquisition time means that that it becomes easier to test a larger number of conditions within one session and to test participants who have difficulty remaining still for a long time.

In this project, we propose to investigate more fully the use of this technique in adults, varying, for example, the F0 of the stimuli, as well as the acoustic features present, with a particular emphasis on features crucial in speech perception. Secondly, we plan to apply it in children and infants, with a view towards making the technique viable for measurements of the FFR to be included in audiological test batteries.

Publications

10 25 50