Lánský Petr, Lánská Vera, Weiss Michael
Institute of Physiology, Academy of Sciences of the Czech Republic, Vídenská 1082, 142 20 Prague 4, Czech Republic.
J Control Release. 2004 Nov 24;100(2):267-74. doi: 10.1016/j.jconrel.2004.08.021.
A stochastic differential equation describing the process of drug dissolution is presented. This approach generalizes the classical deterministic first-order model. Instead of assuming a constant fractional dissolution rate, it is considered here that the rate is corrupted by a white noise. The half-dissolution time is investigated for the model. The maximum likelihood and Bayes methods for the estimation of the parameters of the model are developed. The method is illustrated on experimental data. As expected, due to the nonlinear relationship between the fractional dissolution rate and the dissolution time, the estimates of the dissolution rate obtained from this stochastic model are systematically lower than the rate calculated from the deterministic model.
提出了一个描述药物溶解过程的随机微分方程。这种方法推广了经典的确定性一阶模型。这里不是假设一个恒定的分数溶解速率，而是认为该速率受到白噪声的干扰。研究了该模型的半溶解时间。开发了用于估计模型参数的最大似然法和贝叶斯方法。并通过实验数据对该方法进行了说明。正如预期的那样，由于分数溶解速率与溶解时间之间的非线性关系，从这个随机模型获得的溶解速率估计值系统地低于从确定性模型计算出的速率。
