Localization cues with bilateral cochlear implants investigated in virtual space - 
a case study

Bernhard U. Seeber and Hugo Fastl

AG Technische Akustik, MMK, TU Muenchen, Arcisstr. 21, 80333 Muenchen, Germany
+ Now: Auditory Percept. Lab, UC Berkeley, Berkeley, CA 94720-1650

Cochlear implants stimulate spiral ganglion cells in few, wide areas using discrete,
amplitude-quantized pulses. This results in reduced spectral, temporal, and amplitude 
resolution. Nevertheless, several studies could show outstanding localization ability 
with selected bilateral cochlear implant subjects. A previous free-field study showed 
that interaural time differences could not be evaluated for localization with 
low-frequency carriers (Seeber et al., DAGA 2003). As the relative weighting of 
interaural temporal and level cues for localization can not be tested for
high-frequency carriers in the free-field, a study using virtual sound presentation 
was conducted. Interaural level cues (ILDs) were calculated from test stimuli, which 
incorporated the compressive pre-processing of the cochlear implant. Interaural 
temporal cues (ITDs) were deduced from directional transfer functions of the speech
processor when being placed on a head. The ILD and ITD cues of the test stimuli were 
then independently modified. A localization test using these modified virtual stimuli 
showed high spacial sensitivity to changes in ILDs, but no or small sensitivity to 
ITD changes, even for high frequency carriers. It can thus be followed that interaural 
temporal cues are not used by this subject for localization. In opposite to normal 
hearing, localization is entirely based on monaural spectral cues and interaural level 
cues for our cochlear implant subject.