In pharmacokinetic (PK) analysis, conventional models are described by ordinary differential equations (ODE) that are generally solved in their Laplace transformed forms. The solution in the Laplace transformed forms is inverse Laplace transformed to derive an analytical solution. However, inverse Laplace transform is often mathematically difficult. Consequently, numerical inverse Laplace transform methods have been developed. In this study, we focus on extending the modeling functions of Nonlinear Mixed Effect Model (NONMEM), a standard software for PK and population pharmacokinetic (PPK) analyses, by adding the Fast Inversion of Laplace Transform (FILT) method, one of the representative numerical inverse Laplace transform methods. We implemented PREDFILT, a specialized PRED subroutine, which functions as an internal model unit in NONMEM to enable versatile FILT analysis with second-order precision. The calculation results of the compartment models and a dispersion model are in good agreement with the ordinary analytical solutions and theoretical values. Therefore, PREDFILT ensures enhanced flexibility in PK or PPK analyses under NONMEM environments.

在药代动力学（PK）分析中，常规模型由常微分方程（ODE）描述，通常以其Laplace变换形式求解。拉普拉斯变换形式的解经逆拉普拉斯变换可得出解析解。但是，拉普拉斯逆变换在数学上通常很困难。因此，已经开发了数值逆拉普拉斯变换方法。在这项研究中，我们着重于扩展非线性混合效应模型（NONMEM）的建模功能，该模型是PK和群体药代动力学（PPK）分析的标准软件，其方法是添加拉普拉斯变换的快速反演（FILT）方法，该方法是代表之一数值拉普拉斯逆变换方法。我们实现了PREDFILT，这是专门的PRED子程序，它在NONMEM中用作内部模型单元，可实现具有二阶精度的通用FILT分析。隔室模型和弥散模型的计算结果与常规解析解和理论值吻合良好。因此，PREDFILT可确保在NONMEM环境下进行PK或PPK分析时提高灵活性。