York Neuroimaging Centre, University of York, The Biocentre, York Science Park, Heslington, York YO10 5DG, UK.
Eur J Neurosci. 2010 Nov;32(9):1599-607. doi: 10.1111/j.1460-9568.2010.07423.x. Epub 2010 Oct 8.
Previous behavioural studies in human subjects have demonstrated the importance of amplitude modulations to the process of intelligible speech perception. In functional neuroimaging studies of amplitude modulation processing, the inherent assumption is that all sounds are decomposed into simple building blocks, i.e. sinusoidal modulations. The encoding of complex and dynamic stimuli is often modelled to be the linear addition of a number of sinusoidal modulations and so, by investigating the response of the cortex to sinusoidal modulation, an experimenter can probe the same mechanisms used to encode speech. The experiment described in this paper used magnetoencephalography to measure the auditory steady-state response produced by six sounds, all modulated in amplitude at the same frequency but which formed a continuum from sinusoidal to pulsatile modulation. Analysis of the evoked response shows that the magnitude of the envelope-following response is highly non-linear, with sinusoidal amplitude modulation producing the weakest steady-state response. Conversely, the phase of the steady-state response was related to the shape of the modulation waveform, with the sinusoidal amplitude modulation producing the shortest latency relative to the other stimuli. It is shown that a point in auditory cortex produces a strong envelope following response to all stimuli on the continuum, but the timing of this response is related to the shape of the modulation waveform. The results suggest that steady-state response characteristics are determined by features of the waveform outside of the modulation domain and that the use of purely sinusoidal amplitude modulations may be misleading, especially in the context of speech encoding.
先前在人类受试者中进行的行为研究表明，幅度调制对可理解语音感知过程非常重要。在对幅度调制处理进行的功能神经影像学研究中，一个固有假设是所有声音都被分解为简单的构建块，即正弦调制。复杂和动态刺激的编码通常被建模为若干正弦调制的线性叠加，因此，通过研究皮质对正弦调制的反应，实验者可以探测到用于编码语音的相同机制。本文所述的实验使用脑磁图来测量由六个声音产生的听觉稳态响应，所有声音的幅度都在相同频率下调制，但形成了从正弦调制到脉冲调制的连续体。诱发反应的分析表明，包络跟随反应的幅度具有高度的非线性，正弦幅度调制产生的稳态响应最弱。相反，稳态响应的相位与调制波形的形状有关，正弦幅度调制相对于其他刺激产生最短的潜伏期。结果表明，听觉皮层中的一个点对连续体上的所有刺激都产生强烈的包络跟随反应，但该反应的时间与调制波形的形状有关。研究结果表明，稳态响应特征由调制域之外的波形特征决定，并且纯粹的正弦幅度调制的使用可能具有误导性，尤其是在语音编码的背景下。
