Clinical Research Of Speech-Evoked Auditory Brainstem Responses In Normal Adults

Objective:1. This study investigated the electrophysiological characteristics of speech-evoked auditory brainstem response (speech-ABR), including the appearance of the main components, and the relation to the acoustic properties of stimulus sound in terms of the latencies of the main components, and moreover the comparison to conventional click-ABRs to explore their underlying relationships.2. The differences of speech-ABRs between left and right ears'stimulation are studied from the perspective of time domain and frequency domain for exploring the contrasts and possible reasons of neural coding for speech from different ears at the auditory brainstem.3. How the simulation intensity might affect the characteristics of (speech-ABR) from healthy Chinese adults is examed to gain an insight into the mechanism of neural encoding of speech in brainstem.4. The mechanisms and applications of speech-ABR are attemped to explore the speech perception, and establish the standard recording protocols and general analysis methods, throught which to obtain the normal reference values for healthy adults, thus to provide an objective method for basic and clinical research of speech perception. Methods:The speech-ABR and click-ABR were recorded in 31 healthy young adults by audiometric screen processes in 80dB SPL. The ABRs recorded by the same clinic certificated instrument following the established protocols and the stimulus sound were processed and analyzed in both time and frequency domain. Speech-ABRs to syllable da of 31 healthy adults were recorded with ipsilateral recordings. Statistical analysis was performed on time-domain parameters, such as latencies and amplitudes of featured peaks, and frequency-domain parameters, such as amplitudes of the fundamental frequency and the first formant of speech-ABRs ranging from 20-50 ms. A scoring criterion to grade the appearance of featured waves was proposed for waveform evaluation. Speech-ABRs to syllables da of 32 subjects were recorded in four stimulus intensities varied from 20 to 80 dB SPL by 20 dB steps, under 11.1/s stimulus rate. A set of featured peaks were determined under these intensities, and then analyzed the characteristics of their latencies and amplitudes. All the data were analyzed by SPSS 13.0 statistical software with hypothesis testing levelα= 0.05.Results:The main components of speech-ABRs were clearly identified, and were categorized into onset response, frequency following response, transitional response and offset response repectively. And the latencies of them are short than 12 ms. The latency of peak A significantly correlated with that of peaks V, C, D, E and F respectively (P＜0.05), especially with V. Furthermore, the latencies of peak D, E, and F are significantly correlated with each other (P＜0.01). Transitional component peak C is only correlated with peak V and A (P＜0.05); and offset wave O is independent to all the other peaks (P>0.05). The latencies of speech-V and peakⅢare significantly delayed in comparison with their counterparts of click-ABRs (P＜0.01). It is observed that the appearance rate of wave III is significantly lower than that of click-Ⅲ(P＜0.01), while there is no significant delay and low appearance rate for wave-Ⅰobserved in speech-ABR (P＞0.05). There is no significant difference for the latencies of binaural featured peaks, and amplitudes of feature peaks (except for peaks A and O). The waveform scores of right ears are greater than that of left ears. The amplitudes of fundamental frequency of binaural waves are both greater than that of the first formant. Furthermore, there is no significant difference of amplitudes of fundamental frequency and the first formant between two ears. With the reduction of stimulus intensity, the latencies of speech-ABRs delay significantly (P<0.05), but no statiscal change in amplitude (P>0.05). It shows that the intensity variation induced the average changes of prolonged latencies are similar for V, A, C, and O waves, and the sustained FFR (D, E, F) respectively. Under all intensity conditions, latencies of V, A and C waves, D-E complex, and E-F complex are stable in terms of the significant correlations, especially for V-A and A-C complex.Conclusions:1. The main components of speech-ABR are clearly identified, and are categorized into onset response (including V and A waves), transitional C wave, then following frequency-following response (FFR), and the offset O wave, which are originated in the auditory brainstem.2. The main components of speech-ABR are closely related and analogues to the combination of the main waves of click-ABR, SN10 and FFR evoked by low-frequency tone. They well reflect the verbal information of acoustic stimulus.3. There is no significant difference for the latencies and amplitudes of feature peaks, and no significant difference of amplitudes of fundamental frequency and the first formant between two ears, which possibly related to that the origins and distributions of speech-ABR are essentially symmetrical in brainstem in contrast with the hemisphere asymmetry of speech.4. The waveform scores of right ear are greater than that of left ear may be contributed by the language superiority of left hemisphere, but there is not a simple correlation among the anatomic auditory pathways, the functional what-where pathway, and the predominance of brain hemispheres.5. The latencies of speech-ABRs with different stimulus intensities in healthy adults accord with that of auditory brainstem responses. As the reduction of stimulus intensity, the latencies of speech-ABRs delayed significantly, but amplitudes did not reduce statistically.6. The onset response and FFR show different changes of latencies and correlations of latencies suggesting they may come from different neural mechanisms.7. The speech-ABR approach may be a new tool in the study of human auditory evoked potentials and offer a promising method in the study of speech coding and processing mechanism. In this study, speech-ABRs were recorded and analyzed from various perspectives. The recording and analysis methods are established. Moreover, the exploration of the neural mechanisms and applications of speech-ABRs, and the collections of the normal reference values can benefit the basic and clinical research of speech perception.