gms | German Medical Science

Different speech training methods are currently available for hearing impaired listeners. These methods can be categorized either as being more top-down, relying on speech content associations (using words or sentences), or more bottom-up, where the attention is concentrated on spectro-temporal, low-level speech details (phonemes). Previously, we have shown that supervised phoneme training in laboratory conditions has the potential to improve speech in noise reception thresholds in hearing impaired listeners [1] [2]. Here, we describe a phoneme training method developed for unsupervised phoneme training at home, in the form of a mobile phone app. The training procedure comprises different stages, ranging from finding an optimised training sweet spot for the subject, to changing the training level difficulty (e.g. consonant environment, speakers and signal to noise ratio). Recently, we evaluated a prototype of this app in a home-use study with mild to moderate hearing impaired listeners (N = 15, median age = 63) and found significant effects on phoneme discrimination (trained material) and speech in noise understanding (untrained material).

A long version of this article is available here:

https://dga.cloud/s/mwtmjExgseigFps