Speech-in-noise detection is related to auditory working memory precision for frequency

Abstract

Speech-in-noise (SiN) perception is a critical aspect of natural listening, deficits in which are a major contributor to the hearing handicap in cochlear hearing loss. Studies suggest that Sin perception correlates with cognitive skills, particularly phonological working memory: the ability to hold and manipulate phonemes or words in mind. We consider here the idea that Sin perception is linked to a more general ability to hold sound objects in mind, auditory working memory, irrespective of whether the objects are speech sounds. This process might help combine foreground elements, like speech, over seconds to aid their separation from the background of an auditory scene. We investigated the relationship between auditory working memory precision and Sin thresholds in listeners with normal hearing. We used a novel paradigm that tests auditory working memory for non-speech sounds that vary in frequency and amplitude modulation (AM) rate. the paradigm yields measures of precision in frequency and AM domains, based on the distribution of participants' estimates of the target. Across participants, frequency precision correlated significantly with SiN thresholds. Frequency precision also correlated with the number of years of musical training. Measures of phonological working memory did not correlate with SiN detection ability. Our results demonstrate a specific relationship between working memory for frequency and Sin. We suggest that working memory for frequency facilitates the identification and tracking of foreground objects like speech during natural listening. Working memory performance for frequency also correlated with years of musical instrument experience suggesting that the former is potentially modifiable. Speech-in-noise (SiN) perception is the ability to identify spoken words when background noise is present. Deficits in SiN are one of the most common problems in patients with cochlear hearing loss, but there has been increasing interest in cognitive abilities that determine SiN perception 1. Akeroyd 2 summarised studies describing the relationship of ccognitive measures to speech-in-noise performance. Phonological working measures such as the reading span and digit span were found to have an effect on SiN detection after accounting for hearing loss. However, other studies have suggested that phonological working memory only comes into play in older participants or when a participant has high-frequency hearing loss 3. We consider here the idea that more fundamental forms of working memory that apply to all sounds, including speech, are relevant to SiN perception. From first principles, the ability to hold in mind sound features that are characteristic of particular sources, including voices, might aid SiN perception by allowing sequential outputs from a particular source to be grouped. Previous work has shown that fundamental auditory grouping processes involved in separating non-speech figures from an acoustic background ('figure-ground perception') explains a sizable portion of individual differences in ...

A specific relationship between musical sophistication and auditory working memory

“…Alternatively, the additional demands imposed by musical listening in comparison to listening in general could lead to the acquisition of better general listening skills and different behavioural and neural mechanisms for listening. The correlation between auditory working memory and years of musical experience is parsimoniously explained by the acquisition of listening skills 32 , 33 . Different listening strategies are suggested by functional imaging of musicians carrying out listening tasks that suggest different neural systems for these in musicians 34 .…”

“…Our previous work 33 , using the same working memory paradigm described in Experiment 1, has indicated that sample size of around 50 is needed to obtain an expected correlation of around 0.40, with years of musical training as a variable, with a two-tailed α of 0.05 and a β of 0.20 and so we aimed to recruit around 50 participants each to each individual study. 102 participants were recruited for two online experiments, designed in the form of a web application, that were created using HTML, CSS and Javascript and hosted freely on Firebase, a platform developed by Google for web developers to create bespoke web applications.…”

A specific relationship between musical sophistication and auditory working memory

“…Alternatively, the additional demands imposed by musical listening in comparison to listening in general could lead to the acquisition of better general listening skills and different behavioural and neural mechanisms for listening. The correlation between auditory working memory and years of musical experience is parsimoniously explained by the acquisition of listening skills (Kraus et al, 2012; Lad et al, 2020). Different listening strategies are suggested by functional imaging of musicians carrying out listening tasks that suggest different neural systems for these in musicians (Gaab and Schlaug, 2003).…”

Performance on stochastic figure-ground perception varies with individual differences in speech-in-noise recognition and working memory capacity

“…In addition to the acoustic properties of the stimulus, interactions among individual difference factors, such as age, working memory capacity (WMC), and musical experience also contribute to differences in auditory perception, stream segregation, and temporal auditory acuity that underlie speech-in-noise recognition. For example, auditory WMC has been positively associated with speech-in-noise recognition (Bidelman & Yoo, 2020;Lad et al, 2020), although this relationship is sometimes not observed in younger adults (e.g., Füllgrabe & Rosen, 2016a, 2016bVermeire et al, 2019) who tend to have better speech-in-noise perception than older adults (Presacco et al, 2016;Vermeire et al, 2019). Although the relationship between WMC and speech-in-noise tasks is sometimes dependent on age, studies investigating musical expertise have found that, for both younger (Bidelman & Yoo, 2020;Parbery-Clark et al, 2009) and older adults (Parbery-Clark et al, 2011), individuals with greater musical experience perform better on auditory WMC and speech-innoise tasks (see Coffey et al, 2017 for a review on musician advantages in SIN tasks).…”