Pharmacokinetic modelling of the anti-malarial drug artesunate and its active metabolite dihydroartemisinin Adam J. Hall, Michael J. Chappell, John A.D. Aston, Stephen A. Ward Computer Methods and Programs in Biomedicine Volume 112, Issue 1, Pages 1-15 (October 2013) DOI: 10.1016/j.cmpb.2013.05.010 Copyright © 2013 The Authors Terms and Conditions
Fig. 1 System diagram of the general compartmental model Computer Methods and Programs in Biomedicine 2013 112, 1-15DOI: (10.1016/j.cmpb.2013.05.010) Copyright © 2013 The Authors Terms and Conditions
Fig. 2 Example of observed ARS and DHA plasma concentrations (nmol/l) for three patients. Error bars represent ±15% of the observations and are representative of the assay error (for reasons discussed in Section 4.2). Computer Methods and Programs in Biomedicine 2013 112, 1-15DOI: (10.1016/j.cmpb.2013.05.010) Copyright © 2013 The Authors Terms and Conditions
Fig. 3 Example of model predicted ARS and DHA quantities/concentrations in each compartment for patient A, with table of parameter estimates and their uncertainties. As previously, error bars are representative of assay error. Computer Methods and Programs in Biomedicine 2013 112, 1-15DOI: (10.1016/j.cmpb.2013.05.010) Copyright © 2013 The Authors Terms and Conditions
Fig. 4 Example of model predicted ARS and DHA quantities/concentrations in each compartment for patient B, with table of parameter estimates and their uncertainties. As previously, error bars are representative of assay error. Computer Methods and Programs in Biomedicine 2013 112, 1-15DOI: (10.1016/j.cmpb.2013.05.010) Copyright © 2013 The Authors Terms and Conditions
Fig. 5 Example of model predicted ARS and DHA quantities/concentrations in each compartment for patient C, with table of parameter estimates and their uncertainties. As previously, error bars are representative of assay error. Computer Methods and Programs in Biomedicine 2013 112, 1-15DOI: (10.1016/j.cmpb.2013.05.010) Copyright © 2013 The Authors Terms and Conditions
Fig. 6 Distribution of the coefficient of determination (%) over the dataset; marks in red correspond to patients with unexpected profiles. Recall that the objective was not to maximise the coefficient of determination, but this statistic allows easier comparison between subjects than the actual objective function values. Computer Methods and Programs in Biomedicine 2013 112, 1-15DOI: (10.1016/j.cmpb.2013.05.010) Copyright © 2013 The Authors Terms and Conditions
Fig. 7 Normalised sensitivity plots for the model about the estimated parameters for patient A, first with respect to the observations and then the unobserved compartments Computer Methods and Programs in Biomedicine 2013 112, 1-15DOI: (10.1016/j.cmpb.2013.05.010) Copyright © 2013 The Authors Terms and Conditions