Proefschrift_vd_Beek

for the middle electrode. For this individual p for the highest current level compared to th However, across all subjects the second linea that the curve widths of the medium and h significantly different from those found at the and p 0.8 respectively). Recording electrode The third linear mixed model with data of rized in Table 4) revealed significant effect on the width of the selectivity measures. tended to shift in the direction of the rec the apical electrode (EA-B) the width wa smaller when measured apically compared basal electrode (EB-A) the width was 1.7 measured apically. The results from the el of the array (EM) did not show significan the recording contacts.

ent stimulus levels. There were, however, some notable exceptions,

Table 3. Significance (p-value) of difference in width of selectivity measures along the array (t-statistic and degrees of freedom between

brackets). Selectivity

EA-B

EM-A

EM-B

EB-A

EA-B EM-A

p 0.01 (t 2.56;

dF 289.7)

EM-B

p 0.01 (t 5.43;

p 0.01 (t 7.95;

dF 289.7)

dF 289.3)

EB-A

p 0.01 (t 2.86;

p 0.01 (t 5.21;

p 0.03 (t 2.17;

dF 292.2)

dF 291.2)

dF 291.2)

Masker or probe fixed In five subjects two sets of selectivity measures were performed, one with the position of the masker fixed and one with the probe contact fixed. Apart from some outliers at EB-A, the widths along the array and for different current levels showed similar profiles. Figure 4, C, illustrates the small differences in masker- fixed versus probefixed curves across location and current level in a typical subject. The third linear mixed model using the data of all five subjects is summarized in Table 4. Results showed very similar curve widths for masker or probe position fixed for all electrode contacts (mean differences 0.3 to 0.2 spacing, not significant). Artefact rejection method A linear mixedmodel was generated using the data from the five subjects in which the scanning measurements were performed using the forward masking and the alternating polarity artefact rejection schemes. The curves of those subjects showed the same shape for both methods, as illustrated by an example in Figure 4, D, and differences did not reach significance levels.

F. B. van der Beek et al.

472

Table 4. Mean differences between widths (contact spacing) obtained for selectivity measures with the masker fixed or the probe fixed, with significance (Sig.) levels (left columns). The right columns list the mean differences (with significance levels) of the widths obtained with the recording contact 2 apical or 2 basal relative to the contact of interest.

The finding that scanning produces wider measures has been previously reported by C Hughes & Stille (2010), and is in line with th This can be explained as follows: in the sc recordings are made at several points along t current from the stimulating contact (i.e. sti current generated by the nerve fibers (neural easily through the fluid of the cochlea and res recording electrode some distance from the st selectivity method, however, using a fixed re shows the spread of excitation of the stimulati for scanning were evident at all positions an tion of the middle contact measured in the exception is likely a consequence of the met that many curves for both methods did not r The rationale for the choice of this 0.6 criter methodology section. The limitation of the introduction of arbitrary, fixed values at the proposed by Abbas et al (2004) and complic

Difference masker vs. probe fixed (contact spacing)

Difference apical vs. basal recording (contact spacing)

Measure

Sig.

Sig.

EA-B EM-A EM-B EB-A

0.62 0.75 0.69 0.60

0.3 0.2 0.2 0.3

1.5 1.0

0.01 0.08 0.99 0.01

0.01

1.7

82 | Chapter 4

Masker or probe fixed

11/27/13