Figure 6
Multiple linear regression models (\({\beta }_{0}+{\beta }_{1}{x}_{1}+{\beta }_{2}{x}_{2}\)) shown in each panel were fitted to predict drug Cmax, drug AUC, and drug dose from pairs of quantitative MRI parameters. Each panel is an added variable plot of a drug pharmacokinetic parameter on the x-axis versus pairs of quantitative MRI parameters. Each point represents a linear combination of pairs of parameter changes on the y-axis (\(\frac{{\beta }_{1}}{\Vert \beta \Vert }{x}_{1}+\frac{{\beta }_{2}}{\Vert \beta \Vert }{x}_{2}\), where \({x}_{1}\) and \({x}_{2}\) are the selected parameter changes, \({\beta }_{1}\) and \({\beta }_{2}\) their respective weight in the linear regression, and \(\Vert \beta \Vert =\sqrt{{{\beta }_{1}}^{2}+{{\beta }_{2}}^{2}}\)) versus crolibulin Cmax, crolibulin AUC, and crolibulin dose on the x-axis. 95% confidence bounds are shown in green and the adjusted R2 and p values for both parameters are shown above each panel. The multiple linear regression model described by the equation on the left upper corner is shown by a red line.
