Korn Christoph W, Bach Dominik R
J Vis. 2016;16(3):28. doi: 10.1167/16.3.28.
Pupil size is often used to infer central processes, including attention, memory, and emotion. Recent research has spotlighted its relation to behavioral variables from decision-making models and to neural variables such as locus coeruleus activity and cortical oscillations. As yet, a unified and principled approach for analyzing pupil responses is lacking. Here we seek to establish a formal, quantitative forward model for pupil responses by describing them with linear time-invariant systems. Based on empirical data from human participants, we show that a combination of two linear time-invariant systems can parsimoniously explain approximately all variance evoked by illuminance changes. Notably, the model makes a counterintuitive prediction that pupil constriction dominates the responses to darkness flashes, as in previous empirical reports. This prediction was quantitatively confirmed for responses to light and darkness flashes in an independent group of participants. Crucially, illuminance- and nonilluminance-related inputs to the pupillary system are presumed to share a common final pathway, composed of muscles and nerve terminals. Hence, we can harness our illuminance-based model to estimate the temporal evolution of this neural input for an auditory-oddball task, an emotional-words task, and a visual-detection task. Onset and peak latencies of the estimated neural inputs furnish plausible hypotheses for the complexity of the underlying neural circuit. To conclude, this mathematical description of pupil responses serves as a prerequisite to refining their relation to behavioral and brain indices of cognitive processes.
瞳孔大小常被用于推断包括注意力、记忆和情绪在内的中枢过程。最近的研究聚焦于它与决策模型中的行为变量以及诸如蓝斑核活动和皮层振荡等神经变量的关系。然而，目前仍缺乏一种统一且有原则的方法来分析瞳孔反应。在此，我们试图通过用线性时不变系统来描述瞳孔反应，从而建立一个正式的、定量的瞳孔反应前向模型。基于来自人类参与者的经验数据，我们表明两个线性时不变系统的组合能够简洁地解释由光照变化引发的几乎所有方差。值得注意的是，正如之前的经验报告所示，该模型做出了一个违反直觉的预测，即瞳孔收缩在对黑暗闪光的反应中占主导地位。这一预测在另一组独立参与者对光和黑暗闪光的反应中得到了定量证实。至关重要的是，瞳孔系统中与光照和非光照相关的输入被认为共享一条由肌肉和神经末梢组成的共同最终通路。因此，我们可以利用基于光照的模型来估计在听觉奇偶数任务、情感词汇任务和视觉检测任务中这种神经输入的时间演变。估计的神经输入的起始和峰值潜伏期为潜在神经回路的复杂性提供了合理的假设。总之，这种对瞳孔反应的数学描述是完善其与认知过程的行为和脑指标之间关系的先决条件。
