Graeme Clark Collection
Permanent URI for this collection
Search Results
1 - 10 of 15
-
Item
-
ItemElectrical stimulation of the auditory nerve: the coding of frequency, the perception of pitch and the development of cochlear implant speech processing strategies for profoundly deaf people( 1996)1. The development of speech processing strategies for multiple-channel cochlear implants has depended on encoding sound frequencies and intensities as temporal and spatial patterns of electrical stimulation of the auditory nerve fibres so that speech information of most importance for intelligibility could be transmitted. 2. Initial physiological studies showed that rate encoding of electrical stimulation above 200 pulses/s could not reproduce the normal response patterns in auditory neurons for acoustic stimulation in the speech frequency range above 200 Hz and suggested that place coding was appropriate for the higher frequencies. 3. Rate difference limens in the experimental animal were only similar to those for sound up to 200 Hz. 4. Rate difference limens in implant patients were similar to those obtained in the experimental animal. 5. Satisfactory rate discrimination could be made for durations of 50 and 100 ms, but not 25 ms. This made rate suitable for encoding longer duration suprasegmental speech information, but not segmental information, such as consonants. The rate of stimulation could also be perceived as pitch, discriminated at different electrode sites along the cochlea and discriminated for stimuli across electrodes. 6. Place pitch could be scaled according to the site of stimulation in the cochlea so that a frequency scale was preserved and it also had a different quality from rate pitch and was described as tonality. Place pitch could also be discriminated for the shorter durations (25 ms) required for identifying consonants. 8. As additional speech frequencies have been encoded as place of stimulation, the mean speech perception scores have continued to increase and are now better than the average scores that severely-profoundly deaf adults and children with some residual hearing obtain with a hearing aid.
-
ItemContinuing improvements in speech processing for adult cochlear implant patients( 1995)The Cochlear 22-channel cochlear implant has employed a succession of improved speech-processing strategies since its first use in an adult patient in Melbourne in 1982. 1 The first patients received the F0F2 coding strategy developed by the University of Melbourne, in the Wearable Speech Processor (WSP). The F0F2 coding scheme presented the implant user with three acoustic features of speech. These were 1) the amplitude of the waveform, presented as the amount of current charge, 2) fundamental frequency (F0) or voice pitch, presented as rate of biphasic pulsatile stimulation, and 3) the spectral range of the second formant frequency (F2), which was represented by varying the site of stimulation along the electrode array.
-
ItemImplant designs for future coding strategies( 1995)This paper briefly describes the history of speech processing developments leading to the presently available Speak processing strategy. The similarities and differences of the Speak and Continuous Interleaved Sampling (CIS) strategies are then discussed and some recent key experimental observations are examined as a guide to potential future coding strategies. Key issues for future coding strategies and implant designs are the number of electrodes and stimulation rates in use. Consideration of these issues has led to development of a prototype implant to be used for advanced speech-processing research.
-
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.
-
ItemComparison of the SPEAK (Spectral Maxima) and multipeak speech processing strategies and improved speech perception in background noise( 1995)As more is known about speech processing for Cochlear Implant patients, results should continue to improve. It now appears possible that Cochlear Implant patients may, in some instances, reach performance levels that are better than those obtained by most severely deaf people who use hearing aids.
-
ItemMultichannel cochlear implant speech processing: further variations of the spectral maxima sound processor strategy( 1995)The spectral maxima sound processor (SMSP) was first developed at the University of Melbourne in 1989. A full description of the SMSP has been given by McDermott et al.1 In short, the SMSP utilizes an ear-level microphone to measure acoustic sound pressure. A 16-channel band-pass filter bank is used to analyze the sound spectrum at discrete time intervals. Each of the 16 filters is assigned to one of the 16 intracochlear electrodes according to frequency. Within each time interval the six channels with the largest band-pass filter amplitudes are selected and used to stimulate six corresponding electrodes in quick succession. The current implementation of the SMSP2 differs from the original in that a digital signal processor is used in place of the analog filter bank and the microprocessor. The filter bank has been implemented with a discrete Fourier transform. Also, the input dynamic range has been improved by increasing the resolution of the analog-to-digital converter from 8 to 12 bits.
-
ItemInvestigations on a curved intracochlear array( 1995)The electrode array of a multiple-channel cochlear implant lies against the outer wall of the scala tympani. From this position electrical current spreads to excite residual neural elements, particularly spiral ganglion cells within the modiolus. It is not clear whether the spread of current from the outer wall is optimal for multiple-channel speech processing, but placement closer to the target nerves could result in lower thresholds. This could have benefits through the use of shorter pulse durations and extended battery life. Computer modeling studies and animal experiments have suggested that for localized current the optimal electrode position is adjacent to the modiolus. At the University of Melbourne it was felt that an electrode with a curve matching the internal cochlear spiral would remain close to the modiolus after insertion. A curved electrode was developed and an inserting tool was designed and produced (Treaba et al, this suppl, this section). Preliminary studies suggested that the electrode array did indeed remain close to the modiolus. Before further development of this type of electrode design, it was necessary to determine whether modifications to the surgical technique for its insertion were required. It was also important to ensure that the curved electrode fabricated for clinical trial would lie closer to the modiolus than to the outer wall of the scala tympani. This study was undertaken to examine these issues.
-
ItemSpeech perception in children using the advanced Speak speech-processing strategy( 1995)The Speak speech-processing strategy, developed by the University of Melbourne and commercialized by Cochlear Pty Limited for use in the new Spectra 22 speech processor, has been shown to provide improved speech perception for adults in both quiet and noisy situations. The present study evaluated the ability of children experienced in the use of the Multipeak (Mpeak) speech-processing strategy (implemented in the Nucleus Minisystem-22 cochlear implant) to adapt to and benefit from the advanced Speak speech-processing strategy (implemented in the Nucleus Spectra 22 speech processor). Twelve children were assessed using Mpeak and Speak over a period of 8 months. All of the children had over 1 year's previous experience with Mpeak, and all were able to score significantly on open-set word and sentence tests using the cochlear implant alone. Children were assessed with both live-voice and recorded speech materials, including Consonant-Nucleus-Consonant monosyllabic words and Speech Intelligibility Test sentences. Assessments were made in both quiet and in noise. Assessments were made at 3-week intervals to investigate the ability of the children to adapt to the new speech-processing strategy. For most of the children, a significant advantage was evident when using the Speak strategy as compared with Mpeak. For 4 of the children, there was no decrement in speech perception scores immediately following fitting with Speak. Eight of the children showed a small (10% to 20%) decrement in speech perception scores for between 3 and 6 weeks following the changeover to Speak. After 24 weeks' experience with Speak, 11 of the children had shown a steady increase in speech perception scores, with final Speak scores higher than for Mpeak. Only 1 child showed a significant decrement in speech perception with Speak, which did not recover to original Mpeak levels.
-
ItemTemporal coding of frequency: neuron firing probabilities for acoustic and electric stimulation of the auditory nerve( 1995)A better understanding of the temporal coding of frequency, and its application to electrical stimulation of auditory nerve fibers, should lead to advances in cochlear implant speech processing. Past research studies have suggested that the intervals between nerve action potentials are important in the temporal coding of frequency. For sound frequencies up to approximately 500 Hz, the shortest or predominant intervals between the nerve action potentials are usually the same as the periods of the sound waves. The intervals between each nerve action potential can be plotted as an interval histogram. Although there is evidence that the intervals between spikes are important in the temporal coding of frequency, it is not known up to what frequency this applies. It is also not known whether the information transmitted along individual fibers or an ensemble of fibers is important, to what extent the coding of frequency is interrelated with the coding of intensity, the relative importance of temporal and place coding for different frequencies, and finally, how well electrical stimulation can simulate the temporal coding of sound.