98 / 162
96 | Chapter 5
populationbased fitting, universal templates for Tand M-levels were constructed to limit the amount of
behavioral information required. Additionally, to serve as guidance for fitting children or other difficult-to-
fit cochlear implant recipients, normative data are reported. Furthermore, the T-/M-level ratio was explored
as well as the course of the levels in the first year. Finally, the predictive value of Tand M-levels for speech
understanding in our relatively large study group was examined.
SUBJECTS AND METHODS
Subjects
Clinical data for 151 postlingually deafened adult cochlear implant recipients were analyzed for this study.
All used either a CII or an electrically identical HiRes 90K cochlear implant with a HiFocus 1/1 J electrode
array, which was fully inserted into the cochlea (Advanced Bionics). These subjects were implanted between
2002 and 2008 in the Leiden University Medical Center. All implantations during this period were
performed by only two surgeons. Subjects younger than 16 years were not included in this study. The subject
demographics are shown in table 1. All subjects used the HiRes processing strategy. Fifteen postlingually
deafened adult subjects additionally implanted during this period were not included in the study, for a
variety of reasons (table 2).
Stimulation Levels
T-levels were measured for each active electrode contact separately while delivering a 300-ms pulse train of
biphasic pulses in the following up-down-up procedure. Per electrode contact, stimulus levels were increased,
starting at 0 clinical units (CU), until the subjects indicated that they heard a sound. Next, the current was
increased above this approximate T-level to provide a clearly audible percept on which the subject could
focus. Subsequently, the level was decreased again until the subject indicated that he/ she did not hear the
sound anymore. Then, the level was decreased somewhat further to reach a definitely subthreshold level.
Finally, the level was raised again to find the final T-level. For the M-levels, at initial fitting, a profile was
introduced with an up to 25% (in linear clinical units) emphasis for the more basal electrode contacts
[the electrode numbering in Advanced Bionics devices is from apical (1) to basal (16)]. Subsequently, the
processor was set in live speech mode, and live speech at normal voice level was then administered to the
subject while all of the M-levels were increased simultaneously until speech was reported to be comfortably
loud. At this time, the subject was asked to assess the sound quality. First, an open question was asked, but,
if needed, adjectives (low-pitched, muffled, high-pitched, sharp) were suggested to facilitate the description
of the sound quality for the patient. If the percept had a very low or muffled quality, the M-levels of the
apical electrodes were reduced while maintaining a smooth M-level profile. If the sound was described as
too sharp, the slope of the M-level profile was lowered until the patient accepted the sound quality but never