Werner Reichardt Center for Integrative Neuroscience, Systems Neuroscience, Eberhard Karls University, 72076 Tübingen, Germany.
Department of Cognitive Neurology, Hertie Institute for Clinical Brain Research, Eberhard Karls University, 72076 Tübingen, Germany.
Cereb Cortex. 2018 Jun 1;28(6):2015-2027. doi: 10.1093/cercor/bhx108.
Although whisker-related perception is based predominantly on local, near-instantaneous coding, global, intensive coding, which integrates the vibrotactile signal over time, has also been shown to play a role given appropriate behavioral conditions. Here, we study global coding in isolation by studying head-fixed rats that identified pulsatile stimuli differing in pulse frequency but not in pulse waveforms, thus abolishing perception based on local coding. We quantified time locking and spike counts as likely variables underpinning the 2 coding schemes. Both neurometric variables contained substantial stimulus information, carried even by spikes of single barrel cortex neurons. To elucidate which type of information is actually used by the rats, we systematically compared psychometric with neurometric sensitivity based on the 2 coding schemes. Neurometric performance was calculated by using a population-encoding model incorporating the properties of our recorded neuron sample. We found that sensitivity calculated from spike counts sampled over long periods (>1 s) matched the performance of rats better than the one carried by spikes time-locked to the stimulus. We conclude that spike counts are more relevant to tactile perception when instantaneous kinematic parameters are not available.
虽然与胡须相关的感知主要基于局部的、即时的编码，但全局的、密集的编码也已被证明在适当的行为条件下发挥作用，这种编码会随时间整合振动触觉信号。在这里，我们通过研究固定在头部的大鼠来研究全局编码，这些大鼠识别的脉冲刺激在脉冲频率上有所不同，但在脉冲波形上没有差异，从而消除了基于局部编码的感知。我们量化了时间锁定和尖峰计数，作为支持这两种编码方案的潜在变量。这两种神经测量变量都包含大量的刺激信息，甚至单个桶状皮层神经元的尖峰也携带这些信息。为了阐明大鼠实际上使用了哪种类型的信息，我们根据这两种编码方案系统地比较了心理测量和神经测量的敏感性。神经测量性能是通过使用包含我们记录的神经元样本特性的群体编码模型来计算的。我们发现，从长时间（>1 秒）采样的尖峰计数中计算出的敏感性比与刺激时间锁定的尖峰携带的敏感性更能匹配大鼠的表现。我们的结论是，当无法获得即时运动学参数时，尖峰计数与触觉感知的相关性更强。
