Musical instrument identification in simulated electric acoustic hearing and in cochlear implant users with contralateral hearing aids.

Abstract

While cochlear implant (CI) users achieve high levels of speech perception in quiet conditions, music perception is a challenging task. Gfeller et al. (2000) found that CI users rated music as second most important acoustic stimulus in their lives next to speech. Music is considered to be a central part of human experience, and many CI users can only dream of one day being able to appreciate music as they had before losing their hearing. For those with congenital deafness, music remains a distant hope, something experienced only through reactions of others, as indicated by poor performance of prelingually deaf children with CIs on measures of music perception (Stabej et al., 2012). Some recent work in CI field has focused on improving quality of music perception, and in doing so, improving overall quality of life for implant users (Limb & Roy, 2014; Limb & Rubinstein, 2012; Petersen, Mortensen, Hansen, & Vuust, 2012). This has posed a great challenge to researchers, as CI devices have a long history of design principles solely focused on delivering aspects of acoustic signal that lead to improved speech communication without much consideration of music perception (Loizou, 1998, 2006).A number of factors may limit music perception for CI users (McDermott, 2004). The substantial reduction in spectral resolution, limited dynamic range, and channel interactions between neighboring electrodes may render most music noisy and unpleasant to CI users (Limb & Rubinstein, 2012). Although CI users' rhythm perception is comparable with normal-hearing (NH) listeners (Gfeller & Lansing, 1991; Gfeller, Woodworth, Robin, Witt, & Knutson, 1997; Phillips-Silver et al., 2015), melody and timbre perception remain quite poor. Implant users rely heavily on rhythmic cues in identification of melodies, performing better when recognizing melodies with a distinctive rhythm than recognizing melodies with less rhythmic structure (Kong, Cruz, Jones, & Zeng, 2004; Schulz & Kerber, 1994). Their increased reliance on temporal cues may be a consequence of reduction in spectral resolution owing to a limited number of stimulating electrodes. Pitch perception has been found to vary greatly across CI users. Some listeners can discriminate a difference in pitch of one semitone reliably, whereas others require interval differences as large as two octaves for reliable judgments (Drennan & Rubinstein, 2008; Ping, Yuan, & Feng, 2012). Timbre, defined as the attribute of auditory sensation which enables a listener to judge that two nonidentical sounds, similarly presented and having same loudness and pitch, are dissimilar (ASA 111-1994) has also been found to be problematic for CI users. An important dimension of musical expressivity, timbre is primarily defined through temporal envelope and spectral shape of a sound (Dowling & Harwood, 1986; Handel, 1995), which enable identification and tracking of a musical sound source over time (McAdams & Giordano, 2016). While temporal envelopes are encoded fairly well by CIs, spectral shape is degraded (Drennan & Rubinstein, 2008). Two major limitations include less spectral resolving power (typically Timbre is an essential element in enjoying musical sound. Composers use timbre in form of different musical instruments. …