Proefschrift_vd_Beek
El 1
El 2
El 3
El 4
–18
–12
–6
El 5
El 6
El 7
El 8
–18
–12
–6
El 9
El 10
El 11
El 12
–18
Measrued T-level (dB) –12
–6
El 13
El 14
El 15
El 16
–18
–12
–6
5
0
–18
–12 –6
–18 –12 –6
–18 –12 –6
–18 –12 –6
c
Predicted T-level (dB)
Fig. 5. c Scatter plot of the measured T-levels vs. predicted T-levels for the remaining 30% of the population for each electrode. Fig. 5. c Scatter plot of the measured T-levels vs. predicted T-levels for the remaining 30% of the population for each electrode.
predicted T-levels versus the measured T-levels for all 16 electrode contacts, while table 3a provides the associated numerical data. Again, it is clear that the predictions are slightly better for the center region of the array. This pro- cedure was repeated with a number of other random se- lections of 30% of the population, with essentially the same result. To obtain a T-level profile expressed in clinical units, equation 2 can be reformulated as follows:
ray) was given by the T-level of electrode 7, yielding the following prediction formula: T-level ( electrode ) = 0.01 ( electrode 2 – 7 2 ) + 0.025 ( electrode – 7) + T-level electrode 7 ( in dB ). (2) As shown in figure 5b, the neighboring electrodes in the center of the array provided comparable results, while the mean r was reduced at both ends of the array (to r = 0.86). R-values for electrodes 2, 5, 9 and 14 are not shown in figure 5b, since these electrodes were active in only less than 33% of the subjects (table 3). The goodness of fit of equation 2 for the individual T- levels was tested in the remaining 30% of the measured data (fig. 5c; table 3a). Figure 5c shows scatter plots of the
7 T level lectrode T-level T-level × × e e - lectrode lectrode e 7
1 001
2 2
e
.
lectrode
7 0025 .
lectrode
7
e
10 10
20
2 2
02 .
lectrode
7 05 .
lectrode
7
e
e
(3)
(in CU).
10
van der Beek/Briaire/Frijns
Audiol Neurotol 2015;20:1–16 DOI: 10.1159/000362779
Downloadedby: LeidenUniversity 145.88.209.33 - 11/23/2014 2:34:38PM
105
Made with FlippingBook