Proefschrift_vd_Beek

Table 3. Prediction errors and r-values per individual electrode contact (prediction based upon a randomly chosen subset of 30% of the subjects) a T-levels Electrode Valid, n Prediction errors (dB) Mean ± 1.96 SD Prediction errors (CU) Mean ± 1.96 SD r mean var 95% prediction interval r mean var 95% prediction interval 1 49 0.92 0.61 2.06 –2.2 to 3.4 0.92 3 139 –20 to 26 2 5 0.86 0.10 1.46 –2.3 to 2.5 0.87 0 35 –12 to 12 3 49 0.94 0.38 1.62 –2.1 to 2.9 0.94 2 107 –19 to 22 4 39 0.94 0.38 1.02 –1.6 to 2.4 0.96 2 32 –9 to 13 5 16 0.98 0.06 0.82 –1.7 to 1.8 0.99 –1 82 –19 to 17 6 39 0.98 0.21 0.33 –0.9 to 1.3 0.99 1 8 –4 to 7 7 50 1.00 0.00 0.00 0.0 to 0.0 1.00 0 0 0 to 0 8 39 0.98 0.00 0.44 –1.3 to 1.3 0.98 0 15 –8 to 7 9 16 0.99 0.13 0.42 –1.1 to 1.4 1.00 0 22 –9 to 9 10 39 0.98 –0.09 0.49 –1.5 to 1.3 0.97 –1 27 –11 to 9 11 50 0.97 –0.18 0.84 –2.0 to 1.6 0.98 –2 54 –16 to 13 12 39 0.95 –0.38 0.93 –2.3 to 1.5 0.95 –3 49 –17 to 11 13 50 0.94 –0.24 1.65 –2.8 to 2.3 0.96 –3 128 –25 to 19 14 6 0.94 0.57 0.74 –1.1 to 2.2 0.93 3 35 –8 to 15 15 50 0.85 –0.38 3.86 –4.2 to 3.5 0.88 –4 370 –42 to 34 16 37 0.90 –0.64 1.69 –3.2 to 1.9 0.89 –6 151 –30 to 18

b M-levels Electrode Valid, n

Prediction errors (dB)

Mean ± 1.96 SD Prediction errors (CU)

Mean ± 1.96 SD 95% prediction interval

r

mean var

95% prediction interval

r

mean var

1 2 3 4 5 6 7 8 9

49

0.96 –0.01 0.92 –0.34

0.77 0.89 0.58 0.61 0.37 0.42 0.00 0.31 0.16 0.35 0.19 0.38 0.28 0.53 0.47 0.67

–1.7 to 1.7 –2.2 to 1.5 –1.5 to 1.5 –1.3 to 1.7 –1.3 to 1.1 –1.2 to 1.3 0.0 to 0.0 –1.0 to 1.2 –0.9 to 0.7 –1.3 to 1.1 –1.1 to 0.6 –1.4 to 1.1 –1.3 to 0.7 –2.1 to 0.8 –1.7 to 1.0 –1.8 to 1.4

0.95 –3 0.91 –4 0.97 –1

786 220 542 281 507 214 134 135 235 444 493 557 122 663 611 0

–58 to 52 –33 to 25 –47 to 44 –29 to 36 –46 to 42 –29 to 28 –21 to 25 –27 to 19 –32 to 28 –47 to 35 –45 to 42 –54 to 38 –35 to 8 –59 to 42 –50 to 47 0 to 0

5

49 39 16 39 50 39 16 39 50 39 50 50 37 6

0.97 0.95 0.96 1.00 0.97

0.00 0.21 0.04 0.00 0.09

0.97

3

0.99 –0.08

0.98 –2

0.98 1.00 0.99

0 0 2

1.00 –0.09 0.97 –0.11 0.99 –0.25 0.97 –0.14 0.99 –0.31 0.95 –0.67 0.98 –0.34 0.96 –0.18

1.00 –4 0.98 –2 0.98 –6 0.98 –2 0.98 –8 0.97 –13 0.97 –8 0.98 –2

10 11 12 13 14 15 16

var = variance.

DISCUSSION The present paper demonstrates how the group profile of Tand M-levels in a relatively large population can be described in closed-set formulas and how this can serve as a starting point for fitting individual cochlear implant recipients. With the help of equations 2–5, the measure-ment of the Tand M-level at just one electrode contact along the array suffices to obtain a prediction of Tand M-levels along the array (fig.6; table 3), which can be ap-plied in a simplified and time-efficient fitting procedure. In particular, it can be a Population-Based Prediction of Fitting Levels Audiol Neurotol 2015;20:1–16 DOI: 10.1159/000362779 11

Downloaded by: LeidenUniversity 145.88.209.33 -11/23/20142:34:38PM

106 | Chapter 5