Proefschrift_vd_Beek

practical situations. The use of a microphone array does not cause any decrement in speech perception scores in quiet or at high SNRs relative to scores with the headpiece microphone. The listening tests show an improvement of 8.2 dB and 5.9 dB based on the SRT for the Handymic and the Linkit respectively. This average improvement is higher than improvements of 1–2 dB found in other studies with more electrodes, higher rates or bilateral implantation (Friesen et al., 2001; Frijns et al. 2003; Turner 2004). The absolute word score at 0 dB reaches now 44% for the Handymic and 38% for the Linkit. These values are higher than the average value of 5% found by Friesen et al. (2001). The improvement may be experienced by all CIusers and does, based on the regression analysis of the data (Figure 5), not depend on the personal SRT. From the results in Table 4, it may be concluded that for the cochlear implant users, the average gradient of the psychometric function, based on the curvefitting, equals 4.9%/dB. This means that a typical cochlear implant user, in a listening situation with SNRs just around the SRT, may expect a large average improvement of the phoneme score. For example, in a typical cocktail party or restaurant, SNR values of 0 dB and worse may be expected. The results of the listening tests in the diffuse noise set-up show that in situations with such an SNR, the intelligibility increases from poor (group average 42% in this test) to fair with a level above 62% (CVC score 67% for the Handymic and 62% for the Linkit). This change from below to above 50% might be of significant help to understand what is said and to ease conversation. However, we must not forget that this is still lower than the 5 listeners with normal hearing who can understand more than 90% of what is said, at this same SNR. The trend of an extra benefit of 2.3 dB of the Handymic over the Linkit as found in this study can partly be explained by the closer positioning of the Handymic to the speech loudspeaker. This resulted in a better SNR in the position of the Handymic of +1.5 dB. This difference will also exist in real life and can be significantly more when the Handymic is held nearer to the mouth of the speaker. However, it must be kept in mind that the Handymic and the Linkit are designed for different applications and use. The choice of which device to use should be made on required improvement, the daily situation and personal appreciation. This was also found in our group. This study was initiated by the question as to whether our patients could benefit from directional microphone systems in practical situations in daily life. Only two subjects showed interest in evaluating the systems in daily practice. Both of them experienced severe problems in their daily work and welcomed any improvement in being able to focus on their tasks instead of having to be constantly engaged with communication only. Other subjects showed less interest, although they could expect a significant benefit. It is most likely that they manage to communicate in noisy social settings by the combined use of their cochlear implants and many years of experience of lip-reading.

2

35