Averaging evoked potentials with an improved weighting algorithm.

The procedure of averaging generally applied for noise reduction of evoked potentials can probably be improved by weighting the sweeps in dependence on their noise energy, which fluctuates considerably. This article demonstrates that weighting factors must be free of the signal part in the sweeps if underestimation and distortion of the reconstructed evoked potential are to be avoided. This can be achieved by pairing the sweeps and weighting the sum of a pair by the reciprocal sample variance from its difference. In order to enhance precision in variance determination, pre-whitening the data has been suggested (Wernicke, 1992). In a feasibility study, this technique is compared with several other algorithms known from the literature.

Masking and pitch shift of tone bursts and clicks by low-frequency tones.

From experiments in animals and investigations in humans it is known that the normally phase-dependent masking of a short stimulus by a low-frequency continuous tone does not occur in the case of endolymphatic hydrops. The recording of the masked threshold of short tone stimuli in a loud tone of 30 Hz is to be evaluated for the clinical diagnostics of Ménière's disease. To this purpose, the main parameters of the measurements (type, frequency, duration of the stimulus, and intensity of the masker) and their effect of phase-dependent masking and pitch-shift are investigated. Stimuli above 2 kHz are masked less than those of lower frequencies. Wide-band stimuli are less useful, since only the low-frequency component of their spectrum is masked. The tone stimuli should be short (1 - 2 ms) in order to make the measurement of the phase dependence more accurate. With increasing masker level the masking at phase 0 degree corresponds to the increase in level, at phase 270 degrees the amount is twice as much. The pitch shift which is perceived in low-tone masking depends on the phase of the stimulus, and on the levels of the stimulus and the masking tone. The use of brain stem recordings in the investigation of phase-dependent low tone masking is problematic since well-synchronizing stimuli with high frequency spectral components are masked poorly.

[Cepstrum analysis in voice disorders]. / Cepstrumanalyse bei Stimmerkrankungen.

Objective data and reproducible procedures are increasingly demanded for assessment and long-term comparison in voice disturbances. Well-known electroacoustic methods like formant analysis, autocorrelation and inverse filtering have been applied for this purpose, though without leaving experimental stage. The cepstrum analysis presented here permits an easy and distinct separation of glottal pitch and filter function of the vocal tract. In this study, 112 normal and hoarse voices were analyzed. The sum of amplitudes of the first cepstral pitch peaks differed significantly between the different degrees of hoarseness (0-3). Disturbances producing hoarseness, however, could not be sufficiently differentiated by the method in its present state.

Olfactory evoked potentials and contingent negative variation simultaneously recorded for diagnosis of smell disorders.

Objective diagnosis of olfaction can be performed by registration of cortical olfactory evoked potentials (OEP) and of contingent negative variation (CNV). The CNV is a negative voltage developing at the vertex after discrimination of one of two smells while the patient is expecting a second stimulus. By an adequate procedure, including a long time window for averaging (2.56 seconds) with appropriate filters, the two tests can be performed simultaneously in a single session of less than 10 minutes. Anosmia is determinable by both OEP and CNV, although CNV shows less variability. On the other hand, CNV requires attention and some cooperation of the patient. Parosmia is accessible by CNV only; two odor qualities presented in random order have to be distinguished. Hyposmia can also be detected; just above the discrimination threshold, CNV amplitudes tend to be large--even enhanced--whereas OEP amplitudes may still be undetectable.

[Effect of decreased and lost high frequency hearing on latency of acoustically evoked brain stem reaction]. / Der Einfluss von Hochtonsenken und -abfällen auf die Latenz der akustisch evozierten Hirnstammreaktion.

The diagnosis of retrocochlear damage is supported by an increased latency difference between peak I and peak V. In case of high frequency hearing loss peak I is often hard to determine, and peak V latency may be shifted not only by neural delay, but also by missing basal hair cells. The amount of cochlear delay can be estimated by a procedure presented here. High frequency decay was simulated by steep high-pass noise masking. Peak V latency turned out to be established by the highest unaffected frequency components of the click stimulus. Thus, in case of a high frequency gap (with normalization towards higher frequencies) latency may be almost normal. In case of prolonged latency the amount ascribable to the cochlea may be rather precisely be estimated. If the whole lag is explained this way, unnecessary further diagnostics can be avoided.

[Electromyography of the larynx with skin surface electrodes]. / Zur Elektromyographie des Larynx mit Hautoberflächenelektroden.

A new method of skin surface electromyography (EMG) was evaluated. Twelve electrodes were placed symmetrically in the cricothyroid and thyroid lamina region. Raw EMG signals were continuously recorded, filtered, and rectified. Phonation of the vowel /a/ triggered a computer, which thereafter stored a 625-ms EMG recording, starting 375 ms before and ending 250 ms after onset of the audible phonation. Sixteen to 64 of these recordings were averaged to eliminate random noise. Twenty-six healthy subjects and 9 patients with unilateral laryngeal paralysis were studied. EMG activity exhibited great intra- and interindividual variation. Often, even in healthy subjects, EMG levels were different between the left and right side of the larynx. Thus, this method is not suitable for diagnosis or follow-up of laryngeal paralysis. Sources of variation in EMG activity are discussed.

Low-frequency masking of brainstem potentials.

The phase and intensity dependence of masking a click by a loud low-frequency tone was examined with brainstem potentials. Wave V latency is practically unaffected; its amplitude, however, is maximally suppressed at a phase of 270 degrees, i.e. rarefaction in the ear canal and basilar membrane displacement towards the scala tympani. Corresponding to subjective threshold, a minor suppression is also observed for the opposite deflection. In the phase of maximal suppression, wave V can be cancelled by a 30 Hz tone of 115 dB SPL up to click intensities of 40 dB HL. With cochlear damage, total suppression can be achieved at even higher click intensities as long as they are close to the subjective threshold. Low-frequency suppression is discussed as a tool applicable for early diagnosis of endolymphatic hydrops.

Auditory brain stem response monitoring during middle ear surgery.

Conductive hearing loss can be determined in the course of middle ear surgery by sound probe stimulation and simultaneous recording of auditory brain stem responses. Mechanical properties of the ossicular chain, usually judged subjectively by the surgeon's visual and tactile senses, can be quantified objectively. Thus, disarticulation of ossicles can be localized precisely, and fixation of the head of the malleus can be differentiated from stapes fixation. Moreover, the function of the reconstructed ossicular chain or prosthesis may be ascertained.

[Brain stem evoked response audiometry study with a bone conduction receiver in atresia of the auditory canal and microtia]. / BERA-Untersuchung mit Knochenleitungshörer bei Gehörgangsatresie und Mikrotie.

Conductive hearing loss may be determined from the latency delay of the brainstem auditory evoked response Jewett V, if other reasons that may prolong latency are ruled out. Therefore it is important to determine the bone conduction threshold when examining infants. Two bone vibrators have been chosen, which, with the help of "inverse filtering", provide the possibility of a good sound conduction of the click stimulus and thus a measurement near the hearing threshold. When applying bone vibrators to infants it must be borne in mind, that the sound transfer is reduced because the cranial sutures are not yet closed. The measurement of bone conduction is suitable for securing the result of high-grade combined hearing defects. In two cases of atresia of the external auditory canal and microtia brainstem evoked response audiometry is presented with evaluation of both air and bone conduction.

[Is measuring impedance following ear operations an aid in the analysis of postoperative sequelae?]. / Ist die Impedanzmessung nach Ohroperationen eine Hilfe bei der Analyse postoperativer Folgezustände?

While the diagnostic value of pre-operative tympanometry is undisputed, there are very few publications on its application in postoperative control of sound conduction development. We measured otoadmittance several times after middle ear surgery in 42 cases. A significant increase of maximal admittance is found over a period of up to 24 months in case of tympanoplasty type I (16 cases, Fig. 1), and after stapedectomy (8 cases, Fig. 3). After a tympanoplasty of type III (Wullsteins classification) many tympanograms were flat, even in the follow-up examination, but a tendency to improvement can also be observed (Fig. 2). In none of the groups was hearing loss correlated with otoadmittance. Thus, tympanometry is not suitable for postoperative evaluation of conductive hearing loss. Nevertheless, our results indicate that the healing process after tympanoplasty modifies middle ear sound transmission for quite a long time.

Auditory brainstem responses to single-slope stimuli. The influence of steepness and polarity.

Click polarity has little influence on brainstem potentials. We applied an auditory stimulus similar to a step function generated in a closed acoustic system. The influence of stimulus onset steepness (comprising rise time and intensity) on wave V latency and amplitude was investigated. A remarkable latency prolongation was observed for condensation (C) compared with rarefaction (R), if a sharp bend at the foot of the slope was avoided. The C latency lag was nearly the amount of rise time. The effect can be explained by cochlear travel time. Wave V amplitude for R slopes was significantly enhanced. At high intensity and short rise time, it reached twice the values found with C slopes, or with clicks of either polarity. Although the explanations found are not yet satisfactory, a clinical application in cochlear diagnosis is predictable.

A long-term study of hearing in children following neonatal hyperbilirubinemia.

We conducted a long-term study of 85 children with known transient neonatal hyperbilirubinemia to determine if their hearing had been affected. None of the children had neural symptoms such as kernicterus. The children ranged in age from birth to 9 months and were studied by means of brainstem evoked response audiometry (BERA). Thirty-four of the children were studied sequentially between 15 and 80 months after the initial examination. Our results showed that there was no significant correlation between serum bilirubin concentrations and BERA thresholds or latencies. These findings indicate that, unlike manifest cases of kernicterus, neonatal hyperbilirubinemia does not affect neonatal hearing when treated promptly.

[Latency behavior of early acoustic evoked potentials in inner ear hearing loss]. / Latenzverhalten früher akustisch evozierter Potentiale bei Innenohrschwerhörigkeit.

In 101 patients suffering from sensory hearing loss the latency of the auditory evoked brain stem potentials was investigated. In case of minor and moderate pancochlear deafness latency is found to be within normal range. In severe pancochlear hearing loss above 80 dB latency is significantly delayed. In basocochlear hearing disorders the latency-delay is depending on the extent of the frequency loss and on the degree of pure tone threshold shift. If the patient is suffering from a high frequency hearing loss, normal latency will be found in the higher stimulus levels whereas within low levels latency tends to be delayed depending on the cut-off frequency. A general prolongation of latency across all sound pressure levels will take place in cases with steep forms of high frequency hearing loss. For the purpose of investigating the wave-I-latency we used an earcanal electrode in some patients. It can be shown, that in cochlear hearing disorders the latency of wave V is delayed to the same extent as the latency of wave V. After discussing model studies reference values of wave-V-delay as a function of the degree of high frequency disorders are listed

Brainstem and cochlea potentials evoked by rarefaction and condensation single-slope stimuli. A preliminary report.

Influences of stimulus polarity on Jewett wave V are rather small when using clicks which produce two or more polarity changes within a short time interval. In order to separate pressure changes towards rarefaction (R) and towards condensation (C) we applied steep single-slope stimuli returning very slowly to baseline. Brainstem responses recorded from 8 human subjects differed markedly for R and C onset. Amplitudes were much higher for R than for C onset. The main C response was double-peaked with the first peak appearing about 0.5 ms earlier and the second, higher one, 1 ms later than the predominant R wave. The transition from single-slope to click stimulation was investigated by combining R and C slopes. For large time intervals, independent responses to either slope were observed. Down to an on/off interval of 1 ms, the on-response predominated. For still shorter intervals an equalization of R and C responses and a graduation towards click responses was found. Using the same stimuli, cochlear microphonics (CM) and compound action potentials (CAP) were recorded from 7 guinea pigs. The CM did not replicate the slow off-motion of the single-slope stimulus, but returned back to baseline after 0.6 ms. The C compared with R latency of the CAP (peak N1) was also delayed by 0.6 ms. This delay, and that of human peak V, may be explained by CAP initiation only by one direction of basilar membrane motion.

[Objective gustometry with adequate stimulation by registration of contingent negative variation]. / Objektive Gustometrie mit adäquaten Reizen durch Registrierung der Contingent Negative Variation.

Cortical evoked potentials require a steep and temporally reproduceable onset of sensory stimulation. As both items are difficult with gustatory stimuli, we tried to record the slow expectancy potential or "contingent negative variation" (CNV), which has turned out to be useful in objective olfactometry and speech audiometry. Sour, sweet, bitter or salty fluids were embedded in a water stream, which was sucked through a device well fitting to the tongue with holes at adequate locations. The negative vertex potential developped reproduceably between the recognition of taste and the acoustic signal, which could be expected only after one of the two alternatively offered gustatory stimuli. Clinical use of the method is demonstrated in a case of asymmetric loss of smell due to interruption of the chorda tympani during cholesteatoma operation. The CNV appeared only when stimulating the side of the tongue opposite to the ear operated on. In addition to the total or unilateral loss of taste, the ability to differentiate between two gustatory qualities can be diagnosed with an objective method, provided the patient cooperates by calm and attentive behaviour.

Brain stem potentials evoked by binaural click stimuli with differences in interaural time and intensity.

Auditory-evoked brain stem potentials were recorded from 12 adults with normal hearing using click stimuli with differences in interaural time and intensity. Almost independent superimposed Jewett V peaks were produced, whose latency and amplitude depended on the parameters of the stimulus applied to either ear. This indicates that separate binaural information for the evaluation of sound source direction is still available at the brain stem level where wave V originates. We demonstrate that the normal nonlinear latency/intensity function may be responsible for the subjective compensation of time and intensity differences, since the well-known trading functions show similar intensity-dependent gradients.

[Evoked response audiometry in circulatory disorders in the brainstem (author's transl)]. / Hörbahndiagnostik bei Zirkulationsstörungen im Hirnstamm.

In the case of retrocochlear lesions evoked potentials from different stages of the auditory pathway were recorded. Four Patients with nearly normal hearing threshold and normal cortical potentials show reduced medium latency potentials show reduced medium latency potentials due to circulatory dysfunction in the brainstem. Desynchronisation effects on the brainstem level are completely compensated in the cortex. The damage of the auditory pathway could not be found by X-ray or other usual clinical methods. In the two cases described improvement under medical treatment can be demonstrated.

Diagnosis of abnormal hearing adaptation by brainstem potentials.

Adaptation of the human brainstem potentials was investigated using three types of stimulation: paired clicks, varied click rates, and a gap in a continuous tone. Brainstem responses to the second of a pair of clicks is hardly influenced by pathological adaptation. In patients with a normal subjective threshold decay (Carhart test) peak V latencies of the brainstem potential for high click rates and for the tone gap were within the normal range. For a threshold decay of more than 20 dB, latencies were significantly prolonged. In cases of an unlimited Békésy decay the brainstem response was missing for the tone gap stimulus. Fast rate and tone gap stimulation are suitable for a sensitive illustration of the threshold tone decay.

[Olfactory stimulation and contingent negative variation (CNV) (author's transl)]. / Contingent negativae variation bei olfaktorischer Reizung.

In addition to olfactory evoked responses (Fig. 1a) the vertex-negative voltage shift was registered in 20 normal adults. Fig. 2 shows this potential for an olfactory indicative stimulus followed by a tone burst or flash sequence, which are stopped by the test person's manual reaction. In Fig. 3 two alternating odorous stimuli are presented, one of which is followed by the tone. The selective expectancy wave is followed by the tone. The selective expectancy wave is only built up by the marked olfactory stimulus. For the first time CNV can be used as a clinical investigation tool to confirm central perception of odorous stimuli and furthermore to objectivate impaired central odour discrimination (parosmia).