CIs work by circumventing the inner ear's damaged organs and directly stimulating auditory nerve fibers with electric current, which leads to the perception of sound. Most implant recipients, with advances in technology and implant candidacy criteria, can now attain high levels of speech understanding in quiet settings, a clear demonstration of the device's success. Despite significant achievements in language perception, however, high levels of music perception, more specifically pitch and melody perception, are rarely attained through conventional implant technology. Music, for many CI users, is frequently perceived as unpleasant and difficult to enjoy. (Acta Otorrinolaringol Esp 2008;59:228; Cochlear Implants Int 2003;4:85.)
Cause of Poor Pitch Perception
Unlike nontonal languages, music is highly dependent on accurate perception of pitch, a property related to the fundamental frequency of sound. Pitch functions as a basic building block of music and forms the foundation for musical melody, harmony, and scale. Impairments in pitch perception therefore translate into broad deficits for many different aspects of music perception. Pitch resolution is typically poor for CI users. Unlike normal hearing listeners who can detect a one percent or less difference for frequencies up to 4 kHz, CI users can only detect a 10 percent to 25 percent difference for frequencies up to 0.5 kHz and have much greater difficulties for frequencies over 0.5 kHz. (J Acoust Soc Am 2005;118:338.) These pitch deficits can be largely attributed to the technological limitations of the device.
Pitch is finely encoded in a normal cochlea by the place of stimulation along the cochlea (i.e., place-pitch), so that the apical end corresponds to lower pitches and the basal end to higher ones and the rate at which the auditory nerve fibers fire (i.e., rate-pitch) affects the pitch, or in other words, a faster rate elicits a higher pitch. These two mechanisms of pitch encoding in electric hearing, however, are severely degraded. Some 3,500 inner hairs cells convey finely tuned frequency information to the auditory nerve fibers based on their location along the cochlea in place-pitch for normal hearing. Comparatively, the implanted electrode array contains only eight to 24 electrodes that are meant to replace the frequency specificity of thousands of inner hair cells.
The broad electric fields delivered by the electrodes also excite a relatively large population of nerve fibers, leading to a less than desirable precision of activation along the place of the cochlea for any given frequency. (Ear Hear 2011;32:221.) A mismatch between the frequency bandwidth assigned to an electrode and the pitch percept produced by stimulation of that electrode occurs in many instances. Often, the electrode place maps do not correspond to the intrinsic tonotopic place map found in the cochlea as a result. (Eur Arch Otorhinolaryngol 2011;268:27.) The ability to perceive complex pitches is highly degraded without an accurate pitch-place map.
Current devices do not effectively transmit the fine timing (also known as fine structure) information of the incoming sound wave, even though CI users can somewhat use rate information to encode pitch. Typically, the electrodes present pulses at a fixed rate unrelated to fine timing of waveform, so that the rate information used to encode pitch is absent. The device, to compound matters, is further limited by the pitch-resolving power of the patient's own auditory system because many CI users have substantial auditory nerve degeneration. These technological and biological constraints severely hinder pitch perception, which translates into deficits in many other aspects of music perception, such as melody perception.
Limitations in Melody Perception
Melody emerges from the pitch relationships formed among a series of notes that are presented sequentially overtime. The listener's ability to perceive the direction of pitch change, interval distances between notes, and the overall contour or shape of a series of notes lies at the foundation of melody perception. Implant-mediated pitch perception is generally poor, so the ability to perceive melodies accurately is subsequently impaired. CI users have difficulty identifying the contour of simple five-note melodies and identifying well known songs such as “Mary Had a Little Lamb” or “Happy Birthday” compared with normal hearing listeners. (Ear Hear 2007;28:302; Ear Hear 2009;30:411.) CI users once again perform significantly worse than normal hearing listeners when melody recognition is tested using real-world songs like those heard on the radio. (Ear Hear 2005;26:237.) The device can more readily transmit speech and temporal information than pitch cues, so the presence of lyrics and notable rhythmic lines can facilitate the recognition of familiar melodies for CI users. (J Music Ther 2012;49:68; Cochlear Implants Int 2002;3:29.)
Techniques that facilitate pitch discrimination between successive notes can lead to improvements in melody perception. One group found that CI users perform better during recognition tasks when melodies are composed of pure tones (i.e., each note contained only one frequency) as opposed to complex harmonic tones that typically form musical melodies (i.e., each note is made of many pure tones, all of which are multiples of a single fundamental frequency). (Ear Hear 2009;30:160-168.) Compared with complex tones that can stimulate several electrodes, pure tones usually activate only one electrode, which can produce more extreme differences between notes and clearer pitch contours. Another group artificially expanded the intervals between successive notes to exaggerate the pitch contours and then tested CI users melody identification. CI users more accurately identified common melodies with expanded contour information, suggesting that limited pitch resolution between notes at least partly contributes to the melody deficits observed.
Higher Cortical Processing Deficits
The listener must process a sequence of individual pitches in temporal coherence to one another to appreciate melodies, allowing a singular musical phrase to emerge from these discrete elements. Evidence suggests that the normal auditory system encodes a series of pitches more accurately than sequences of other acoustic features, such as timbre or loudness. (J Acoust Soc Am 2009;126:3179.) This increased perceptual salience for pitch sequences, particularly pitches with resolved harmonics, presumably facilitates our ability to perceive complex melodies accurately. Unlike normal hearing listeners, however, CI users do not exhibit any advantage processing sequences of pitches over other acoustic features such as loudness. (Hear Res 2010;269[1-2]:34.) These results suggest that melody perception impairments result not only from poor pitch resolution but also from higher cognitive processing deficits. The inability of current implants to support harmonic resolvability is hypothesized to at least partially contribute to limitations in pitch sequence processing.
Functional neuroimaging in CI users has been another method to increase our understanding of cortical mechanisms of music perception in CI users. Positron emission tomography scan, an instrument that traces cerebral blood flow, provides a useful, implant-compatible tool for visualizing brain activity in CI users while they listen to auditory stimuli. One study compared cortical activity in CI users with normal hearing controls during language, rhythm, and melody listening tasks. (J Assoc Res Otolaryngol 2010; 11:133.)
Compared with normal hearing controls, CI users demonstrated greater auditory cortex activation for all listening conditions. The difference between groups was greatest for the language condition and weakest for the melody condition. Considering that CI users are much more accurate at language perception than melody perception, the authors suggested a possible link between auditory performance and degree of cortical activity, so that increased activation may facilitate auditory stimuli processing. The limited activation of cortical regions during melody listening tasks may provide a neural basis for melody processing difficulties for CI users.
Measures Related to Melody Perception
Several individual subject factors have been correlated to melody recognition abilities. Patients with residual low-frequency acoustic hearing in addition to their implant (referred to as electrical-acoustical stimulation or bimodal hearing) exhibit better melody recognition and pitch-ranking abilities over patients who only hear electric. (Ear Hear 2007;28:412; J Acoust Soc Am 2005;117[3 Pt 1]:1351.) Experts hypothesize that this additional acoustic hearing provides important fine structure and fundamental frequency percepts that facilitate music perception.
Higher speech perception scores and increased amount of focused music listening have also been positively associated with melody recognition abilities. (Ear Hear 2005;26:237.) Conversely, age has been negatively correlated to most music listening tasks, including melody identification. Older implant users typically have an increased incidence of nerve degeneration and cortical processing impairments, such as temporal processing and visual monitoring, which likely contribute to impaired melody processing. (J Acoust Soc Am 2010;118:338; J Am Acad Audiol 2009;20:71.) No objective evidence supports that any one implant device provides better pitch or melody perception over another, despite advertised differences among implant manufacturers.
Training May Improve Music Perception
Currently, most CI rehabilitation programs focus on acquiring necessary language skills with little (if any) emphasis on music-related hearing. Musical perceptual skills, unlike language skills, do not significantly improve over time as a result of regular auditory exposure; focused music rehabilitation programs therefore become even more of a necessity for CI users to achieve their full musical potential. (J Am Acad Audiol 2010;21:28; Laryngoscope 2007;117:1183.) Accumulating evidence exists to support the implementation of such programs. (J Am Acad Audiol 2009;20:71; Int J Audiol 2010;49:116.) One group, for example, found that CI users who participated in only a one-week to two-month training program demonstrated significantly improved abilities to recognize semitone distances, identify melodic contours, and recognize familiar melodies compared with baseline performance. (Ear Hear 2007;28:302; Ann N Y Acad Sci 2009; 1169:518.)
Improvements in performance increased as the training period continued in duration. Follow-up testing one month after training revealed that performance was still significantly higher than baseline, suggesting possible long-term benefits of music rehabilitation. Overall, more effort is still needed to explore the advantages of various music programs, and these programs still need to become widely available for CI users. Music rehabilitation still offers great promise for CI users to improve pitch, melody perception, and music enjoyment and appreciation.
Increasing attention has been placed on identifying and improving areas of poor music perception because many CI users attain good speech scores. Pitch perception impairments form the foundation of many implanted-mediated musical deficits, including limitations in melody perception. CI users, in addition to clear pitch and melody perceptual deficits, also exhibit higher cortical processing impairments during musical tasks. Focused music rehabilitation programs offer great potential for music perception improvements in CI users, despite these limitations. Considering that music is one of the most difficult acoustic stimuli that CI users must process, the achievement of satisfactory music perception represents the pinnacle of implant-mediated listening. It is hoped that improvements in CI technology in conjunction with musical training programs will result in reaching this goal in the future.
© 2012 Lippincott Williams & Wilkins, Inc.
- Click and Connect! Access the links in The Hearing Journal by reading this issue on our website or in our new iPad app, both available at thehearingjournal.com.
- Comments about this article? Write to HJ at HJ@wolterskluwer.com.
- Follow us on Twitter at twitter.com/hearingjournal and like us on Facebook at www.facebook.com/HearingJournal.