Department of Biostatistics, University at Buffalo, State University of New York , Buffalo, NY, USA.
Department of Chemistry, Buffalo State, State University of New York , Buffalo, NY, USA.
J Biopharm Stat. 2020 Jul 3;30(4):704-720. doi: 10.1080/10543406.2020.1730871. Epub 2020 Mar 4.
Estimating the area under a curve (AUC) is an important subject in many fields of medicine and science. The regression model using B-spline functions provides flexibility in curve fitting, making it suitable for AUC estimation with various types of nonlinear trends. Despite the versatility of the B-spline approach, comprehensive discussions regarding relevant AUC estimation techniques using B-spline functions and their comparison with existing methods cannot be found in extant literature. In this paper, we investigate AUC estimation using B-spline regression and B-spline regression with several penalties, as well as discuss corresponding inferences. We carry out an extensive Monte Carlo study to evaluate the performance of the proposed methods in various realistic pharmacokinetics and analytical chemistry data settings. We show that the proposed methods provide robust and reliable AUC estimation regardless of different types of nonlinear models from scientific and medical research areas. Our proposed method is appropriate for general AUC estimation since it does not require nonlinear model specifications and inference techniques corresponding to the specified model.
估算曲线下面积（AUC）是医学和科学许多领域的重要课题。使用 B 样条函数的回归模型在曲线拟合方面提供了灵活性，使其适用于具有各种非线性趋势的 AUC 估计。尽管 B 样条方法具有多功能性，但在现有文献中，无法找到关于使用 B 样条函数进行相关 AUC 估计技术的综合讨论及其与现有方法的比较。在本文中，我们研究了使用 B 样条回归和具有多个惩罚项的 B 样条回归进行 AUC 估计，并讨论了相应的推断。我们进行了广泛的蒙特卡罗研究，以评估所提出方法在各种现实的药代动力学和分析化学数据环境中的性能。我们表明，所提出的方法提供了稳健和可靠的 AUC 估计，无论来自科学和医学研究领域的非线性模型类型如何。我们提出的方法适用于一般的 AUC 估计，因为它不需要与指定模型相对应的非线性模型规范和推断技术。
