Directional sensitivity of neurons in the primary auditory (AI) cortex: effects of sound-source intensity level.
Directional sensitivity of neurons in the primary auditory (AI) cortex: effects of sound-source intensity level.
复制标题
DOI：
10.1152/jn.00563.2002
复制
发表时间：
2003-02
影响因子：
2.5
通讯作者：
R. Reale;R. Jenison;J. Brugge
中科院分区：
文献类型：
--
作者：
R. Reale;R. Jenison;J. Brugge
在全麻醉猫初级听觉皮层（AI）单个神经元的记录下，在虚拟声空间中从不同方向传递瞬时声音。通过参数化改变声源的强度水平来确定空间感受野的操作特性。空间感受野是根据每个采样方向上声音响应的开始潜伏期来构建的。响应潜伏期的空间梯度构成一个接受野，部分是由于系统的共同依赖于声源方向和强度水平。通常，在任何给定的强度水平下，感受野内反应潜伏期的分布是单峰的，范围约为3-4 ms，尽管对于某些细胞和某些水平，分布可达20 ms或小至2 ms。不同方向上的平均反应潜伏期在相同强度水平下不同神经元之间存在差异，在相同强度水平下不同神经元之间也存在差异。一般来说，强度水平的增加导致均值和方差的减小，这遵循反高斯分布。基于响应延迟的感受野模型使用多个参数（方位角、仰角、强度）开发，通过蒙特卡罗模拟验证，并使用球谐分析描述其空间滤波。通过将高斯反密度与估计声源方向和强度的概率反问题联系起来，获得了模型感受野集合的观测结果。灵敏度的上界是利用Fisher信息从集合中得到的，估计误差的预测模式与心理物理表现有关。
Transient sounds were delivered from different directions in virtual acoustic space while recording from single neurons in primary auditory cortex (AI) of cats under general anesthesia. The intensity level of the sound source was varied parametrically to determine the operating characteristics of the spatial receptive field. The spatial receptive field was constructed from the onset latency of the response to a sound at each sampled direction. Spatial gradients of response latency composing a receptive field are due partially to a systematic co-dependence on sound-source direction and intensity level. Typically, at any given intensity level, the distribution of response latency within the receptive field was unimodal with a range of approximately 3-4 ms, although for some cells and some levels, the spread could be as much as 20 or as little as 2 ms. Response latency, averaged across directions, differed among neurons for the same intensity level, and also differed among intensity levels for the same neuron. Generally, increases in intensity level resulted in decreases in the mean and variance, which follows an inverse Gaussian distribution. Receptive field models, based on response latency, are developed using multiple parameters (azimuth, elevation, intensity), validated with Monte Carlo simulation, and their spatial filtering described using spherical harmonic analysis. Observations from an ensemble of modeled receptive fields are obtained by linking the inverse Gaussian density to the probabilistic inverse problem of estimating sound-source direction and intensity. Upper bounds on acuity is derived from the ensemble using Fisher information, and the predicted patterns of estimation errors are related to psychophysical performance.