Auditory model of speech intelligibility for electro-acoustic listeners
Ladan Zamaninezhad, Volker Hohmann, Tim Jürgens
Medizinische Physik and Cluster of Excellence Hearing4all, University of Oldenburg
To date, the use of atraumatic electrodes in cochlear implantation allows many severe-to-profound hearing-impaired listeners to retain their acoustic hearing after surgery. These “hybrid“ listeners can receive electric stimulation via their cochlear implant (CI) and low-frequency acoustic stimulation (in most cases amplified using a hearing aid) simultaneously in the same ear. The clinical study of Büchner et al. (2009) in a group of 22 hybrid listeners investigated speech intelligibility in noise in terms of their speech-reception-threshold (SRT, i.e. SNR at 50% recognition score).
Speech intelligibility in noise was found to be much improved in the electro-acoustic listening condition in comparison to electric-only and acoustic-only listening condition. This improvement in speech intelligibility is called “electro-acoustic benefit“. The goal of this study is to develop a model of speech intelligibility for hybrid listeners that (a) is physiologically plausible, (b) shows electro-acoustic benefit, (c) reproduces the SRTs reported in the study of Büchner et al. (2009). This model can be used in later studies to make prediction about the benefits due to specific algorithms for hybrid listeners.