Klerman Elizabeth B, St Hilaire Melissa
Division of Sleep Medicine, Brigham and Women's Hospital and Harvard Medical School, Boston, MA 02115, USA.
J Biol Rhythms. 2007 Apr;22(2):91-102. doi: 10.1177/0748730407299200.
Mathematical models of neurobehavioral performance and alertness have both basic science and practical applications. These models can be especially useful in predicting the effect of different sleep-wake schedules on human neurobehavioral objective performance and subjective alertness under many conditions. Several relevant models currently exist in the literature. In principle, the development and refinement of any mathematical model should be based on an explicit modeling methodology, such as the Box modeling paradigm, that formally defines the model structure and calculates the set of parameters. While most mathematical models of neurobehavioral performance and alertness include homeostatic, circadian, and sleep inertia components and their interactions, there may be fundamental differences in the equations included in these models. In part, these may be due to differences in the assumptions of the underlying physiology. Because the choice of model equations can have a dramatic influence on the results, it is necessary to consider these differences in assumptions when examining the results from a model and when comparing results across models. This article presents principles of mathematical modeling and examples of how such procedures can be applied to the development and refinement of mathematical models of neurobehavioral performance and alertness. This article also presents several methods of testing and comparing these models, suggests different uses of the models, and discusses problems with current models.
神经行为表现和警觉性的数学模型兼具基础科学和实际应用价值。这些模型在预测不同睡眠-清醒时间表对多种条件下人类神经行为客观表现和主观警觉性的影响方面可能特别有用。目前文献中存在几种相关模型。原则上，任何数学模型的开发和完善都应基于明确的建模方法，如Box建模范式，该范式正式定义模型结构并计算参数集。虽然大多数神经行为表现和警觉性的数学模型都包括稳态、昼夜节律和睡眠惯性成分及其相互作用，但这些模型所包含的方程可能存在根本差异。部分原因可能是基础生理学假设的不同。由于模型方程的选择会对结果产生巨大影响，因此在审视模型结果以及跨模型比较结果时，有必要考虑这些假设差异。本文介绍了数学建模的原则以及此类程序如何应用于神经行为表现和警觉性数学模型的开发与完善的示例。本文还介绍了几种测试和比较这些模型的方法，提出了模型的不同用途，并讨论了当前模型存在的问题。
