Department of Brain and Cognitive Sciences, College of Arts, Science, and Engineering, University of Rochester, Rochester, NY 14627, USA.
Hear Res. 2010 Aug;267(1-2):96-110. doi: 10.1016/j.heares.2010.04.005. Epub 2010 Apr 27.
Linear measures of auditory receptive fields do not always fully account for a neuron's response to spectrotemporally-complex signals such as frequency-modulated sweeps (FM) and communication sounds. A possible source of this discrepancy is cross-frequency interactions, common response properties which may be missed by linear receptive fields but captured using two-tone masking. Using a patterned tonal sequence that included a balanced set of all possible tone-to-tone transitions, we have here combined the spectrotemporal receptive field with two-tone masking to measure spectrotemporal response maps (STRM). Recording from single units in the mustached bat inferior colliculus, we found significant non-linear interactions between sequential tones in all sampled units. In particular, tone-pair STRMs revealed three common features not visible in linear single-tone STRMs: 1) two-tone facilitative interactions, 2) frequency-specific suppression, and 3) post-stimulatory suppression in the absence of spiking. We also found a correlative relationship between these nonlinear receptive field features and sensitivity for different rates and directions of FM sweeps, dynamic features found in many vocalizations, including speech. The overwhelming prevalence of cross-frequency interactions revealed by this technique provides further evidence of the central auditory system's role as a pattern-detector, and underscores the need to include nonlinearity in measures of the receptive field.
线性听觉感受野并不能完全解释神经元对调频扫频和通讯声等复杂时频谱信号的反应。这种差异的一个可能来源是跨频相互作用,这是一种常见的响应特性,可能会被线性感受野所忽略,但可以通过双音掩蔽来捕捉。本研究采用包含所有可能的音对转换的模式化音调序列,将频谱-时间感受野与双音掩蔽相结合,以测量频谱-时间反应图(STRM)。在须蝠下丘记录单个单位,我们发现所有被采样的单位中,连续音调之间存在显著的非线性相互作用。特别是,音调对 STRM 揭示了三个在线性单音调 STRM 中不可见的常见特征:1)双音调促进相互作用,2)频率特异性抑制,以及 3)在没有放电的情况下刺激后的抑制。我们还发现这些非线性感受野特征与不同调频扫频速率和方向的敏感性之间存在相关关系,这些动态特征存在于许多发声中,包括语音。这种技术揭示的大量跨频相互作用进一步证明了中枢听觉系统作为模式探测器的作用,并强调了在感受野测量中纳入非线性的必要性。