Combining an upward shifting CI with a hearing aid

When a CI is combined with a hearing aid, and is fitted in a standard way with a basalward shift of frequency to place map, this mapping will inevitably be in conflict with the frequency to place mapping of the acoustic modality. The presence of two conflicting frequency to place maps could pose far more of a problem than when the listener only hears speech through a cochlear implant.

Work in HearCom has been using acoustic simulations of a combined fitting of a cochlear implant and a hearing aid to investigate the effects of mismatches of frequency to place map between the two ears
Here we have used an acoustic simulation that, like the examples above, uses 6 sine waves, but presented with only 3 sine waves to each ear, and with the basalward shift applied only to one ear.
 
 
 
 
The figure shows the 6 analysis filters at the bottom. The middle panel represents the 3 upward shifted sine waves whose amplitudes are controlled by the speech envelopes extracted in filters 1 3 and 5. The top panel shiws the 3 unshifted sine waves controlled by envelopes from filters 2, 4, 6.
 
The following examples are stereo sound files, and the intended effect requires the use of headphones so that the left and right channels are each audible only in one ear.

The first example contains only the 3 unshifted sine waves shown in blue

Next, the 3 upward shifted sine waves as shown in pink are presented to the opposite ear.

Third, you can hear both ears together, with the upward shift only in one ear.

Finally, the upward shift is removed so that there are 6 unshifted tones, 3 to each ear.

 
 
If it was easy to make use of the information carried by the 3 upward shifted sine waves when these are heard together with the 3 unshifted tones, speech should be easier to understand when this combination is heard than when the 3 unshifted tones are heard by themselves since more spectral information is available.. And if this was indeed the case it would seem that a frequency-to-place mismatch between a cochlear implant and a hearing aid would pose no problems. But several experiments run at UCL in the Phonetics and Linguistics department have shown that adding the 3 shifted tones never improves performance, and that this remains true after at least 10 hours of intensive training. From this we conclude that it is likely to be important to avoid this form of mismatch of frequency-to-place maps when fitting a cochlear implant together with a contralateral hearing aid
 
Further details of this research can be found in this poster presented at the 2007 Conference on Implantable Auditory Protheses
 
 
The simulations outlined above refer to a mismatch across the whole speech frequency range. In considering a typical CI recipient, the hearing aided ear will have only low frequency hearing. Read more about simulations that more accurately reflect this situation.