Cognitive training might have potential to improve speech understanding under adverse listening conditions. Here, we have examined the effects of a 5-week computer-based cognitive training program on speech-in-noise-performance, in normal hearing (NH) participants and in participants with mild-to-moderate sensorineural hearing loss (HI).

Two groups, matched on gender and age (45-65 years), of 20 participants each (HI and NH respectively) are recruited. Participants perform four test-sessions; inclusion (t0), five weeks (t1), ten weeks (t2) and six months (t3). Training is performed either between t0 and t1, or between t1 and t2 (using a cross-over design), using the computer-based Cogmed training program, approximately 30-40 minutes per day, five days per week, during five weeks. At each session participants are tested in three different ways: (a) cognitive testing (KIPS, SICSPAN, TRT); (b) auditory performance (pure tone-audiometry (air- and bone-conduction) and speech audiometry (HINT, Swedish SPIN-test (SNR +4dB))); (c) cortical activation (MR sessions where participants performed a speech-in-noise task using Hagerman-sentences with steady-state speech-spectrum noise (SSN) and with two competing talkers). MR imaging is performed on a Philips Achieva 1.5 Tesla scanner using a sparse imaging technique in which stimuli are presented during the silent period between successive scans. Participants listen to auditory stimuli under eight different conditions: clear speech, SSN or two competing talkers (each at 90%, 50% and 0% intelligibility), and silent rest. Pre- and post-training, hearing disability is assessed by the Speech-Spatial-Qualities-Questionnaire.

The study is on-going and behavioral results as well as results from fMRI will be presented.