Graeme Clark Collection
Permanent URI for this collection
Search Results
1 - 10 of 39
-
ItemA frequency importance function for a new monosyllabic word test( 1998)A frequency importance function, characterising the relative contribution of different frequency bands to speech intelligibility, was determined for a CNC monosyllabic word test designed for Australian usage at the University of Melbourne. The importance function was derived from the phoneme scores of 12 normally-hearing listeners who were tested under various conditions of low-and high-pass filtering presented at signal-to-noise ratios of -8 to +6 dB, using noise which was shaped across frequency 10 match the speech spectrum. The importance function showed a dominant peak at approximately 2000 Hz, which is consistent with previously published word test importance functions. The word test, along with the importance function, will be useful in advanced hearing-aid fitting procedures and research aimed at improving speech perception.
-
ItemThe perceptual dimensions of single-electrode and nonsimultaneous dual-electrode stimuli in cochlear implantees( 1996)Abstract not available due to copyright.
-
ItemThe importance of different frequency bands to the speech perception of cochlear implantees [Abstract]( 1996)It is well known that cochlear implantees exhibit a wide range of speech perception ability. Understanding the reason for this variability may lead to improved speech processors. This study investigates whether implantees rely on different areas of the speech spectrum for speech cues, compared to normally hearing listeners, and whether poor performers rely on different spectral areas than better performers. Six subjects with the Mini System 22 implant and using the SPEAK strategy participated in this experiment. Scores for monosyllabic words were obtained using the full speech spectrum and with selected frequency bands removed from the subjects’ speech processor maps. The Articulation Index (AI) is a measure of the proportion of speech information available to a listener, and the relative contribution to AI from different frequency bands is termed the Importance Function. The five frequency bands studied in this experiment were determined to be of equal importance to normally hearing listeners for the speech material used. The scores for each implantee were transformed into AI values, and hence the relative importance of the bands was determined. This relative importance was compared between the implantee group and normally hearing listeners to determine the way in which speech perception by electrical stimulation varies from that by acoustical stimulation. Comparisons were also made between individual implantees to determine whether correlations exist between their speech perception ability and their use of cues in different parts of the spectrum. Further research will determine whether the differences among implantees are correlated with their ability to perceive changes in stimulation place or temporal characteristics.
-
ItemEvaluation of the Nucleus Spectra 22 Processor and New Speech Processing Strategy (SPEAK) in postlinguistically deafened adults( 1995)A new speech processing strategy (SPEAK) has been compared with the previous Multipeak (MPEAK) strategy in a study with 24 postlinguistically deafened adults. The results show that performance with the SPEAK coding strategy was significantly better for 58.3% of subjects on closed-set consonant identification, for 33.3% of subjects on closed-set vowel identification and open-set monosyllabic word recognition, and for 81.8% of subjects on open-set sentence recognition in quiet and in competing noise (+ 10 dB signal-to-noise ratio). By far the largest improvement observed was for sentence recognition in noise, with the mean score across subjects for the SPEAK strategy twice that obtained with MPEAK.
-
ItemArchitecture of the Spectra 22 speech processor( 1995)The Spectra 22 is a logical extension in the development of the speech processing for the Cochlear Mini 22 system. It can implement the new coding strategy (Speak) that has provided significant improvement in patient benefit.
-
ItemLoudness summation for two channels of stimulation in cochlear implants: effects of spatial and temporal separation( 1995)An experiment with four implantees with the Mini System 22 device was undertaken to measure the loudness summation across two channels of stimulation, with stimuli in which the current pulses were delivered alternately to each channel. The effects of varying spatial separation, temporal separation, and extent of stimulation were investigated. It was found that the absolute amount of summation varied among subjects, and was in general independent of electrode separation, except for a reduction at zero separation. Widening of the spatial extent of the stimulation did not have a consistent effect. There was a reduction in summation for all subjects at zero electrode separation when the time between the two pulses was increased from less than I millisecond to 2 milliseconds. In conclusion, loudness summation did not appear to be highly dependent on parameters that affect the spatial current spread in the cochlea. Further study of the effect of temporal parameters on loudness may help to quantify interaction between stimulation channels.
-
ItemCoding of voice source information in the Nucleus cochlear implant system( 1995)Two studies are reported in which the effectiveness of explicitly coding voicing and fundamental frequency information for the Nucleus cochlear implant was investigated. ln the first study, the voicing perception of a group of three experienced Multipeak users was evaluated when they were using Multipeak and a modified Multipeak in which the explicit fundamental frequency and voicing cues were eliminated and replaced with a 250-Hz constant rate of stimulation. The results of consonant and monosyllabic word tests showed that there was no significant difference in the subjects' ability to discriminate voicing. In the second study, the ability of a group of five experienced users of the constant rate spectral maxima sound processor (SMSP) strategy to discriminate suprasegmental contrasts was evaluated when they were using the SMSP strategy and a modified SMSP strategy that included a rate-encoded representation of the fundamental frequency on the most apical stimulation channel. The results of intonation, roving stress, and question-statement tests showed that there was no significant difference between the scores recorded with these strategies. Since the temporal voicing cue is not a primary cue to voicing discrimination for Multipeak users, and the provision of an additional rate cue to the SMSP strategy does not improve SMSP users' ability to discriminate suprasegmental contrasts, the results of these studies indicate that in the cases investigated, the coding of voice source information by rate of stimulation does not significantly augment the cues present in the spatially distributed constant rate stimulation pattern.
-
ItemPitch matching of amplitude-modulated current pulse trains by cochlear implantees: the effect of modulation depth( 1995)Abstract not available due to copyright.
-
ItemEvaluation of a new Spectral Peak coding strategy for the Nucleus 22 channel cochlear implant system( 1994)Sixty-three postlinguistically deaf adults from four English-speaking countries participated in a 17-week field study of performance with a new speech coding strategy, Spectral Peak (SPEAK), and the most widely used strategy, Multipeak (MPEAK), both of which are implemented on wearable speech processors of the Nucleus 22 Channel Cochlear Implant System; MPEAK is a feature-extraction strategy, whereas SPEAK is a filterbank strategy. Subjects' performance was evaluated with an experimental design in which use of each strategy was reversed and replicated (ABAB). Average scores for speech tests presented sound-only at 70 dB SPL were higher with the SPEAK strategy than with the MPEAK strategy. For tests in quiet, mean scores for medial vowels were 74.8 percent versus 70.1 percent; for medial consonants, 68.6 percent versus 56.6 percent; for monosyllabic words, 33.8 percent versus 24.6 percent; and for sentences, 77.5 percent versus 67.4 percent. For tests in noise, mean scores for Four-Choice Spondees at +10 and +5 dB signal-to-noise ratio (S/N) were 88.5 percent versus 73.6 percent and 80.1 percent versus 62.3 percent, respectively; and for sentences at +15 dB, +10, and +5 dB S/N, 66.5 percent versus 43.4 percent, 61.5 percent versus 37.1 percent, and 60.4 percent versus 31.7 percent, respectively. Subjects showed marked improvement in recognition of sentences in noise with the new SPEAK filterbank strategy. These results agree closely with subjects' responses to a questionnaire on which approximately 80 percent reported they heard best with the SPEAK strategy for everyday listening situations.
-
ItemThe spectral maxima sound processor: recent findings in speech perception and psychophysics(Wien, 1994)The Spectral Maxima Sound Processor (SMSP) was developed at the University of Melbourne for use with the Mini System 22 implant manufactured by Cochlear Pty Ltd. The SMSP has been shown in recent studies to provide improved speech perception to implantees when compared to the currently commercially available processor for this implant (the MSP (MULTIPEAK) processor). In the first of three experiments, the effect on speech perception of increasing the rate of stimulation of the SMSP and of increasing the number of electrodes activated in each stimulation cycle was studied. It was found that these parameter changes made little difference to speech perception in quiet but both changes were advantageous for some subjects when listening in noise. The second and third experiments investigated psychophysically the effects of two aspects of the SMSP strategy which differ from previous processors for this implant. In the second experiment, it was found that concurrent stimulation of two adjacent or nearby electrodes evoked a pitch which was intermediate to that of either electrode. This may explain, in part, the better discrimination of vowel formants by users of the SMSP. In the third experiment, it was found that a pitch related to the modulation frequency was evoked by amplitude-modulating a constant rate stimulus, provided that the rate of stimulation was sufficiently high (four times the modulation frequency) or a multiple of the modulation frequency. This result may explain the equal ability of SMSP and MSP users to perceive speaker differences and intonation patterns, even though the rate of stimulation is constant In the SMSP.