Speech Perception in Noise: A Narrative Review of Cognitive Contribution and Hearing-Aid Signal Processing

Environmental noise is one of the most pervasive ecological health issues in modern urban settings, contributing to the onset of hearing damage but also to the development of annoyance, sleep disturbance, cardiovascular diseases, and impaired communication. For people with hearing impairment of any age, the main challenge posed by noisy environments is not merely detecting sound but understanding speech against a competing background-a task heavily related to cognition.

Narrative review; the literature has been identified through searches of PubMed, Scopus, and Google Scholar, combining terms related to cognition, speech perception, speech in noise, hearing-aid signal processing, environmental health, and aging.

The role of cognitive abilities in speech perception in noise has been evaluated, with particular attention to how artificial intelligence supports the interplay between cognition and hearing-aid signal processing in adverse environments. Within the Ease of Language Understanding (ELU) model, this paper describes how degraded acoustic input shifts processing from rapid, implicit lexical access toward explicit working-memory-dependent reconstruction. The actual evidence indicates (i) that working memory, in particular, could predict aided speech recognition under adverse conditions, (ii) that aging dissociates these processes, and (iii) that longitudinal studies link hearing-aid use to slower cognitive decline.

Speech perception in noise is a coordinated neurocognitive process rather than a purely auditory task. Working memory is a central cognitive contributor, with effects that become evident specifically under adverse conditions. In particular, fast-acting compression and frequency lowering may be counterproductive for listeners in case working memory or speech processing are reduced. The success of different algorithms applied to audio signal processing of hearing aid, designed to compensate for difficult comprehension of words and phrases in noisy scenarios is also strictly dependent on working memory, and artificial intelligence significantly contributes to this goal.