[Simulation of speech perception with cochlear implants : Influence of frequency and level of fundamental frequency components with electronic acoustic stimulation]. / Simulation von Sprachverstehen mit Cochleaimplantat : Einfluss von Frequenz und Pegel der Grundfrequenzkomponente bei elektrisch-akustischer Stimulation.

BACKGROUND: After implantation of cochlear implants with hearing preservation for combined electronic acoustic stimulation (EAS), the residual acoustic hearing ability relays fundamental speech frequency information in the low frequency range. With the help of acoustic simulation of EAS hearing perception the impact of frequency and level fine structure of speech signals can be systematically examined. OBJECTIVE: The aim of this study was to measure the speech reception threshold (SRT) under various noise conditions with acoustic EAS simulation by variation of the frequency and level information of the fundamental frequency f0 of speech. The study was carried out to determine to what extent the SRT is impaired by modification of the f0 fine structure. MATERIAL AND METHODS: Using partial tone time pattern analysis an acoustic EAS simulation of the speech material from the Oldenburg sentence test (OLSA) was generated. In addition, determination of the f0 curve of the speech material was conducted. Subsequently, either the parameter frequency or level of f0 was fixed in order to remove one of the two fine contour information of the speech signal. The processed OLSA sentences were used to determine the SRT in background noise under various test conditions. The conditions "f0 fixed frequency" and "f0 fixed level" were tested under two different situations, under "amplitude modulated background noise" and "continuous background noise" conditions. A total of 24 subjects with normal hearing participated in the study. RESULTS: The SRT in background noise for the condition "f0 fixed frequency" was more favorable in continuous noise with 2.7 dB and in modulated noise with 0.8 dB compared to the condition "f0 fixed level" with 3.7 dB and 2.9 dB, respectively. DISCUSSION: In the simulation of speech perception with cochlear implants and acoustic components, the level information of the fundamental frequency had a stronger impact on speech intelligibility than the frequency information. The method of simulation of transmission of cochlear implants allows investigation of how various parameters influence speech intelligibility in subjects with normal hearing.

Temporal dynamics of pitch strength in regular-interval noises: effect of listening region and an auditory model.

Recently, it was demonstrated that the pitch strength of a stimulus denoted "AABB" differed from rippled noise (RN) despite the fact that their long-term spectra and autocorrelation functions are identical (Wiegrebe et al., 1998). Rippled noise is generated by adding a delayed copy of Gaussian noise back to itself; AABB is generated by concatenating equal-duration, Gaussian-noise segments where every segment is repeated once. It was shown that a simple model based on a two-stage integration process separated by a nonlinear transformation explains the pitch-strength differences quantitatively. Here, we investigate how the spectral listening region influences pitch-strength differences between RN and AABB. Bandpass filtering the two stimuli with a constant bandwidth of 1 kHz revealed a systematic effect of center frequency. For relatively high pitches (corresponding to delays, d, of 4 or 5.6 ms, pitch strength differences between AABB and RN were absent when the pass band was between 0 and 1 kHz. When the pass band was between 3.5 and 4.5 kHz, pitch-strength differences were substantial. For lower pitches (d equal to or longer than 8 ms), AABB had a substantially greater pitch strength independent of the filter center frequency. The model presented in Wiegrebe et al. (1998) cannot capture these effects of center frequency. Here, it is demonstrated that it is possible to simulate the RN-AABB pitch-strength differences, and the effect of listening region, with a computer model of the auditory periphery. It is shown that, in an auditory model, pitch-strength differences are introduced by the nonlinear transformation possibly associated with half-wave rectification or mechanoelectrical transduction. In this experimental context, however, the nonlinearity has perceptual relevance only when the differences in short-term fluctuations of AABB and RN are preserved in auditory-filter outputs. The current experiments relate the purely functional model introduced in the preceding paper to basic properties of the peripheral auditory system. The implication for neural time constants of pitch processing is discussed.

[Effect of venting the ear mold on speech discrimination in masking noise]. / Der Einfluss der Zusatzbohrung im Ohrpassstück auf die Sprachdiskrimination im Störgeräusch.

In this study the importance of in situ measurement during the fitting of a hearing aid is clearly emphasized. Reliable evaluation of the real ear gain can be achieved only with this method, and hence the assessment of the advantages of earmold modifications on the hearing aid. In particular, the gain can be evaluated by the person fitting the aid. The insertion of a vent into an earmold raises the listening comfort of the hearing-impaired person. The speech intelligibility in background noise may also improve. We analysed the effect of earmold venting on speech intelligibility under different background noise conditions. We found that venting improves the speech intelligibility, especially in background noise simulating modulated speech. Our example clearly demonstrates the importance of an exact control of the hearing aid fitting by the physician. In one case a vent ended at the ear canal wall so that no improvement of hearing comfort could be expected. A new earmold was made and the effect on insertion gain was demonstrated when enlarging the vent step by step.

The hearing sensation roughness and neuronal responses to AM-tones.

The hearing sensation roughness as a function of modulation frequency shows a bandpass characteristic similar to that of modulation transfer functions (MTFs) obtained from neurons. However, while the MTFs may change their characteristics from bandpass to low-pass at low levels or with masking noise, roughness shows a bandpass characteristic irrespective of level or masking noise.

Temporal integration in normal hearing, cochlear impairment, and impairment simulated by masking.

To assess temporal integration in normal hearing, cochlear impairment, and impairment simulated by masking, absolute thresholds for tones were measured as a function of duration. Durations ranged from 500 ms down to 15 ms at 0.25 kHz, 8 ms at 1 kHz, and 2 ms at 4 and 14 kHz. An adaptive 2I, 2AFC procedure with feedback was used. On each trial, two 500-ms observation intervals, marked by lights, were presented with an interstimulus interval of 250 ms. The monaural signal was presented in the temporal center of one observation interval. The results for five normal and six impaired listeners show: (1) normal listeners' thresholds decrease by about 8 to 10 dB per decade of duration, as expected; (2) listeners with cochlear impairments generally show less temporal integration than normal listeners; and (3) listeners with impairments simulated using masking noise generally show the same amount of temporal integration as normal listeners tested in the quiet. The difference between real and simulated impairments indicates that the reduced temporal integration observed in impaired listeners probably is not due to splatter of energy to frequency regions where thresholds are low, but reflects reduced temporal integration per se.

Searching for neural correlates of the hearing sensation fluctuation strength in the auditory cortex of squirrel monkeys.

Sounds with slow (less than 20 Hz) fluctuations may elicit the hearing sensation fluctuation strength. For AM tones, neural correlates of fluctuation strength were searched in the auditory cortex of unanesthetized squirrel monkeys. To enable a comparison of psychophysical and physiological data, the 'modulation' of the peristimulus time histogram was fitted by a sinusoidal function. The dependence of the amplitude of this function on modulation frequency, modulation depth and sound pressure level was often comparable to the dependence of fluctuation strength on the same stimulus parameters. In particular, as a function of modulation frequency, the neural data also show a bandpass characteristic at low modulation frequencies as was found for the hearing sensation fluctuation strength.

Suppression in simultaneous masking.

Suppression, i.e., the decrease of masked threshold caused by the addition of a second masker M2 to a first masker M1, is measured for the case of simultaneous masking. The magnitude of suppression decreases with increasing test tone duration; pulsed maskers elicit somewhat more suppression than continuous maskers. In comparison to suppression effects obtained in nonsimultaneous masking (post-masking, pulsation threshold) suppression in simultaneous masking is considerably smaller and was found only at the lower slopes of the two maskers. Suppression in simultaneous masking would not be predicted by those models of suppression which require nonsimultaneous presentation of maskers and test sound.

A free-field equalizer for TDH 39 earphones.

The free-field response of TDH 39 earphones, mounted in MX 41/AR cushions, is determined by loudness comparisons in an anechoic chamber. Based on these data, a passive equalizing network with two resonances at 720 and 6000 Hz is developed and realized. When used with this free-field equalizer, the earphone TDH 39 produces a free-field equivalent level independent of frequency within +/- 2.5 dB in the frequency range 100 Hz to 10 kHz. Because of the small differences between TDH 39 and TDH 49 earphones it is expected that the equalizer can also be successfully applied with the TDH 49.

Fluctuation strength and temporal masking patterns of amplitude-modulated broadband noise.

Fluctuation strength of sinusoidally amplitude modulated broadband noise was measured by a magnitude-estimation procedure. Fluctuation strength shows, as a function of modulation frequency, a bandpass characteristic with a maximum around 4 Hz. With increasing modulation depth and sound pressure level, fluctuation strength increases. For the same stimulus parameters as used in the experiments on fluctuation strength, temporal masking of amplitude modulated broadband noise was determined. The level difference delta L between the maximum and the minimum of the resulting temporal masking patterns shows a low-pass characteristic as a function of modulation frequency and increases with increasing modulation depth and sound pressure level. A model of fluctuation strength F based on the depth delta L of the temporal masking pattern and the modulation frequency fmod in relation to 4 Hz is proposed as follows: F approximately delta L/((fmod/4 Hz)+(4 Hz/fmod)). The model accounts for the dependence of fluctuation strength on some essential stimulus parameters. The usefulness of temporal masking patterns as an intermediate value for the description of hearing sensations is confirmed.

Discrimination of level differences by hearing-impaired patients.

A technique for measuring the discrimination of level differences with pulsed tones is proposed. Alternating tone bursts, each of 500 ms duration but of different sound pressure levels, are separated by 200-ms pauses. The advantages of this method in comparison with traditional methods (SISI test, Lüscher-Zwislocki test) which use modulated tones are that level discrimination is independent of frequency and that it shows very little dependence on sensation level. With this technique, the level discrimination of patients with conductive hearing loss, 'toxic' hearing impairment, noise-induced hearing loss, sudden deafness, presbyacusis, Meniere's disease, and retrocochlear impairment was determined and compared to the results obtained in normal ears. Using the level discrimination test proposed, retrocochlear disorders can be diagnosed with great reliability.

Scaling of pitch strength.

Pitch strength of the following equally loud sounds was scaled by magnitude estimation procedures: pure and complex tones, band-pass filtered complex tones, AM-tones, low-pass, high-pass and band-pass noise, comb-filtered noise. AM-noise and Zwicker-tone. At the test frequencies 125, 250 and 500 Hz pure tones were assigned a pitch strength of 100%. Relative to this value spectral pitches reach 100--75%, virtual pitches 50% and noise pitches 25--0% pitch strength. At 125 Hz no consistent data were found for the Zwicker-tone, while at 500 Hz it elicits on the average the same pitch strength as a pure tone (100%). Implications concerning pitch mechanisms would seem to suggest a prevalence of the 'place principle' in comparison to the 'time principle'.

Frequency discrimination for pulsed versus modulated tones.

Estimates of frequency discrimination for pulsed modulated tones were obtained by 11 observers at 350, 500, 1000, 4000, and 8000 Hz. At low frequencies, frequency DL's are larger for modulated than for pulsed tones; at 8000 Hz the contrary was found. Frequency DL's (difference limens) determined by different methods and procedures differed by a factor up to four; extreme individual frequency DL's, however, by a factor up to 27.

Simulation of a hearing loss at long versus short test tones.

At long test tones, a hearing loss (abrupt drop at 3 kHz) is simulated in good approximation by presenting a continuous masking noise of appropriate spectral distribution to a normal ear. When shortening the test tones, the increase in threshold is systematically smaller for the impaired ear than for the ear with simulated hearing loss, showing normal temporal integration. Comparing threshold curves at long versus short test tones, for the hearing loss simulation only an upward shift is found. For the impaired ear, however, the magnitude of the hearing loss apparently increases with increasing test tone duration; moreover, the spectral characteristic of the hearing loss is quite different at long versus short test tones.

Influence of test tone duration on auditory masking patterns.

Temporal integration functions of tone bursts masked by critical band noises become shallower with increasing test tone frequency. This holds for tone bursts in the centre as well as at the slopes of the masking pattern of critical band maskers. When shortening the tone bursts, the horizontal masking pattern of uniform masking noise changes into a pattern with decreasing slope for increasing frequency. The shape of the masking pattern of critical band maskers remains rather independent of test tone duration.