Speech perception using maps based on neural response telemetry measures

To determine how speech perception is affected for Nucleus cochlear implant users when their speech processor is programmed using neural response telemetry (NRT) measures rather than traditional behavioral estimates of threshold and maximum comfort level.

Electrically evoked compound action potential (EAP) thresholds were measured for a group of 10 adult Nucleus cochlear implant users. These physiologic threshold estimates were used to create three SPEAK MAPs. One MAP ("+10/-20 MAP") was created using only the NRT data. The second MAP ("Combined MAP") was created using a combination of both EAP thresholds and a single behavioralmeasure of threshold and maximum comfort level from electrode 10. A third MAP ("Measured MAP") was created using standard programming techniques. Speech perception was then assessed using each of these three MAPs for either CUNY or HINT sentences presented at two different presentation levels (70 dB SPL and 55 dB SPL).

On average, at the higher presentation level (70 dB SPL) subjects performed significantly better when using the Measured MAP than whenusing either of the NRT-based MAPs. No significant difference in speech perception was obtained for either of the three MAPs when the lower presentation level (55 dB SPL) was used.

This study demonstrated that at relatively high presentation levels, speech perception scores obtained by subjects who use MAPs based solely or primarily on EAP thresholds ar eslightly lower than similar scores obtained using a more traditional MAP. This difference did not reach statistical significance at lower presentation levels. We interpret these findings in a positive light to suggest that although NRT-based MAPs may not be optimal, they are of sufficient quality to support reasonable levels of speech understanding. This is important to establish because these MAPs may prove to be most useful when very young children receive a cochlear implant.