Spectral-Ripple Resolution Correlates with Speech Reception in Noise in Cochlear Implant Users

Abstract

Speech perception ability in noise is one of the most practical measures of success with a cochlear implant; however, with experience, this ability can change dramatically over time, making it a less than ideal tool for comparing performance among different processing strategies. This study examined performance on a spectral discrimination task and compared it to speech perception in noise. An adaptive procedure was used to determine the spectral-ripple density that subjects could discriminate. A closed-set, forced-choice adaptive procedure was used to determine speech reception thresholds for words in twotalker babble and in speech-shaped, steady-state noise. Spectral-ripple thresholds (ripples/octave) were significantly correlated with speech reception thresholds (dB SNR) in noise for 29 cochlear implant users (r = _ 0.55, p = 0.002 in two-talker babble; r = _ 0.62, p = 0.0004 in steady-state noise), demonstrating that better spectral resolution was associated with better speech perception in noise. A significant correlation was also found between the spectralripple discrimination ability and word recognition in quiet (r = 0.50, p = 0.009). In addition, test-retest reliability for spectral-ripple discrimination was good, and no learning was observed. The present study demonstrates that the spectral-ripple discrimination test, which is time efficient and nonlinguistic, would be a useful tool to evaluate cochlear implant performance with different signal processing strategies.

Spectral and Temporal Analysis of Simulated Dead Regions in Cochlear Implants

“…Stimuli with a ripple density of 1.414 ripples/octave and a ripple depth of 30 dB were used for this analysis. This ripple density was chosen because it was similar to the average spectral-ripple discrimination thresholds for the no-dead region condition in the current study as well as to the average performance reported for 31 CI subjects by an earlier report (Won et al 2007). The top panels show the electrode outputs for the redistributed conditions, whereas the bottom panels show the electrode outputs for the dropped conditions.…”

“…When there was no dead region, CI subjects showed an average threshold of 1.80 ripples/octave, which is similar to an earlier report using the same spectralripple test. Using 31 CI subjects, Won et al (2007) reported an average threshold of 1.73 ripples/octave. When NH subjects were tested with a 16-channel vocoder with no-dead region, they showed worse performance than CI subjects.…”

“…Spectral-ripple discrimination thresholds were collected using established techniques (Won et al 2007). Briefly, three rippled noise tokens with a 30-dB peakto-trough ratio, two with standard ripple phase and one with inverted ripple phase, were selected for each trial.…”

“…Three different non-speech psychoacoustic experiments were also performed. Spectral-ripple discrimination was evaluated to assess listeners' sensitivity to spectral envelope cues (Won et al 2007(Won et al , 2014Jones et al 2013). Temporal modulation detection was used to evaluate listeners' sensitivity to temporal envelope cues.…”

Use of Amplitude Modulation Cues Recovered from Frequency Modulation for Cochlear Implant Users When Original Speech Cues Are Severely Degraded

“…From the signal processing stand point, it is generally accepted that two main factors contribute to reduced speech identification in noise for CI users: (1) limited spectral resolution including the negative effect of channel interaction (e.g., Fu et al 1998;Friesen et al 2001;Fu and Nogaki 2005;Won et al 2007; Verschuur 2009), and (2) inability to deliver FM information through current CI speech processors (e.g., Smith et al 2002;Qin and Oxenham 2003;Stickney et al 2005;Zeng et al 2005;Gnansia et al 2010). The latter point should be stated more precisely because speech FM information can be transmitted in the form of recovered AM cues through the CI speech processor and CI users are able to exploit the recovered AM cues for speech identification (Won et al 2012).…”