Share this article on:

Making Room in the Brain for Speech Understanding

Chung, King PhD

doi: 10.1097/01.HJ.0000459745.86628.ec
Journal Club

Dr. Chung is an associate professor of audiology at Northern Illinois University, with expertise in amplification and wind noise research. She also leads humanitarian research and service programs to a different country every year.

Figure

Figure

The effects of cognitive function on audition and speech understanding continue to fascinate the hearing community. One attribute of cognitive function is working memory capacity, which describes a person's ability to keep relevant information in mind while performing a task.

Figure

Figure

Many studies have examined the effects of working memory on the auditory performance of people listening in environments with signal degradation (e.g., steady-state background noise, modulated background noise, or accented speech), using different amplification technologies (e.g., hearing aids, cochlear implants, or noise-reduction algorithms), or possessing certain characteristics (e.g., musical training, older versus younger age, or target speech in the listener's second language).

While there are numerous models for working memory, Jerker Rönnberg ( Int J Audiol 2003;42[suppl 1]:S68-S76http://informahealthcare.com/doi/abs/10.3109/14992020309074626) and colleagues ( Front Syst Neurosci 2013;7:31http://journal.frontiersin.org/Journal/10.3389/fnsys.2013.00031/full) proposed the Ease of Language Understanding model to explain the complex cognitive processes and the multimodal sensory binding processes involved in speech understanding.

In the model, speech understanding is accomplished by a fast episodic buffer and a slow language-processing loop. The fast episodic buffer matches phonological information received by the listener with existing language representations stored in the long-term memory to derive the meaning of the information. This episodic buffer is generally rapid and automatic, and it is most effective in optimal listening conditions.

The slow language-processing loop, on the other hand, is elicited in challenging listening situations. In such suboptimal conditions, the listener may be presented with mismatches between the auditory input and the mental lexicon stored in the long-term memory, requiring a conscious effort to summon higher-level executive-processing functions and decode the lexicon.

These executive functions may include: 1) inhibiting irrelevant information to keep it out of the working memory, 2) updating old stored information with new relevant information in the working memory, and 3) shifting expectations to reconstruct fragmented mismatches and make sense of incoming auditory signals. Engaging these executive-processing functions generally requires effort and increases the cognitive load.

Cognitive Spare Capacity and Speech Communication: A Narrative Overview

Rudner M, Lunner T

BioMed Res Int

2014;869726http://www.hindawi.com/journals/bmri/2014/869726/

In this review article, Mary Rudner and Thomas Lunner summarized the neural pathways for different cognitive functions, as well as the effects of hearing impairment, background noise, and hearing aid use on audition and speech understanding.

They also elaborated on the concepts of the Ease of Language Understanding model and the cognitive spare capacity for communication, which refers to the cognitive resources left for higher-level speech processing (e.g., remembering the last words of a series of sentences) when a person is performing a task (e.g., repeating the sentences one by one).

In addition, the authors explained the rationale for the Cognitive Spare Capacity Test (CSCT), which was recently developed by Sushmit Mishra and colleagues ( J Speech Lang Hear Res 2013;56[4]:1120-1132http://jslhr.pubs.asha.org/article.aspx?articleid=1797124).

The CSCT was designed to assess the executive functions of updating and inhibition using high or low memory load tasks presented in the audio-only or audiovisual mode. Double-digit numbers from 13 to 99 were grouped to form 48 lists of 13 numbers each.

There were three to four inhibition or updating occasions per list. The tests can be administered in quiet or in noise.

Prior to the presentation of stimuli in the updating test, the listener is asked to recall either the highest or the lowest number from the list. As a result, information in the working memory must be updated; the listener needs to discard a previously remembered number when a number with a higher or lower value is heard.

In the inhibition test, the listener is required to recall either two odd or two even numbers spoken by the female or the male speaker. Inhibition is required to ignore all the nonrelevant information presented and to remember the odd or even numbers spoken by the target speaker.

In high memory load trials, the listener recalls the first number plus the two odd or even numbers in the list (i.e., recalls three numbers). In low memory load trials, the listener is asked to recall the two odd or even numbers only (i.e., recall two numbers).

Back to Top | Article Outline

INFLUENCING SPEECH UNDERSTANDING

Further, the authors reviewed the relationship among working memory/cognitive spare capacity, speech characteristics, and other factors affecting audition. For example:

  1. Background noise increases cognitive load and reduces cognitive spare capacity.
  2. Working memory and cognitive spare capacity are related, but they are not correlated.
  3. Adults with normal hearing obtain lower CSCT scores in the audiovisual condition than in the audio-only condition in quiet, likely because visual cues are distractions in such a favorable condition.
  4. Older adults with hearing loss obtain similar CSCT scores in the audio-only and audiovisual conditions in quiet when the memory load is low. Visual cues improve their scores in quiet when the memory load is high and in noise in general.
  5. Noise-reduction algorithms can increase cognitive spare capacity by improving word recall of adults with normal hearing and with hearing loss.
  6. Phonological representation in the mental lexicon is affected by factors such as hearing impairment, speaker characteristics, use of amplification devices, and distortion or enhancement from signal processing. Good working memory can compensate for part of any losses. Methods to enhance phonological distinctiveness of target speech and phonological representations are needed to improve cognitive function and communication in challenging listening situations.
  7. Semantic and text cues can help people with hearing impairment understand speech in challenging situations. Future studies are needed to examine the relationship among semantic/text cues, signal-processing strategies, and cognitive abilities.
  8. Older adults have lower CSCT scores than younger adults because they may need to engage more cognitive resources to compensate for sensory degradation or cognitive decline.
  9. Hearing loss is associated with cognitive decline and poorer long-term memory, but not with lower working memory.
  10. Training in multitasking abilities may improve working memory. Keeping the brain healthy and providing rich sensory input may help older adults understand speech in challenging situations.

Working memory and cognitive spare capacity play key roles in speech understanding. Learning how to optimize these resources is an important area for further research.

Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.