74 / 162
72 | Chapter 4
Theoretically, electrodes that produce more localized excitation would be more likely to provide better
pitch acuity and less channel interaction. However, previous research showed no correlation between pitch
ranking and eCAP SOE (Hughes & Abbas, 2006a).
The attraction of eCAP measurements, in particular, is that these can usually be performed in young
children, either intraor postoperatively, whereas psychophysical testing would often be impossible. Even in
adults, psychophysical testing can be time-consuming and clinical eCAP procedures can be semi-automated
and thereby made straightforward to perform clinically. Thus, it may be possible to identify the electrode
contacts with narrower SOE, as part of the goal to identify an ideal subset of electrode contacts in an
individual.
eCAP recordings can be used in several ways to evaluate the SOE and are usually made using the back
telemetry capabilities of the CI. A relatively basic method (termed “scanning” in this paper) is to stimulate
one electrode contact and then measure the evoked response on all the other contacts along the array (Frijns
et al, 2002; Cohen et al, 2004) (Figure 1, A). The amplitude profile of the responses thus obtained is not
a direct measure of the SOE (as even a very narrow region of excitation can be detected some distance
away) but wider regions of excitation would be expected to result in wider scanning profiles. For artefact
reduction alternating polarity is incorporated in some clinical fitting software (used with Advanced Bionics
cochlear implants), because it is a faster method than the subtraction method with masker and probe
(Klop et al, 2004). A theoretically more precise method (termed “selectivity” in this paper) is analogous
to the production of psychophysical forward-masking tuning curves, whereby a masker is applied to each
electrode contact in turn, while a fixed “probe” contact is stimulated subsequently. The response amplitude
indicates the amount of overlap between masker and probe (Figure 1, B). A limitation of this method is
that the SOE of an electrode is derived from the neural excitation of two different electrodes (the masker
and probe electrode). Residual charge from the masker pulse on the neural membranes of non-excited
fibers could potentially change the region responding to the probe pulse, leading to a systematic deviation
from the actual region of overlap. Additionally, the SOE of the masker contacts can vary along the array,
which will affect the derived SOE of the electrode contact of interest. However, eCAP selectivity curves
so produced have been shown to match behaviorally-obtained psychophysical tuning curves (Cohen et al,
2003), which theoretically suffer from the same limitation.
Besides eCAP thresholds, eCAP-derived SOE measures are not a totally accurate reflection of the neural
excitation. There are several characteristics of eCAP measures that might account for this. First, the position
of the recording contact can affect certain characteristics of SOE measures. One limitation of intracochlear
measurements is that the recording electrode needs to be located some distance away from the stimulating
electrode contact. As the electrical pulse generated at the stimulating electrode is much larger than the
neural response, the stimulus introduces a large artefact when the recording electrode is nearby. Another
parameter, which may affect comparisons between different studies, is that not all researchers have used
the same position of the recording contact. Indeed, there is even some discrepancy between publications