Tissue concentration time curves of the MCA [panel (a)] and the SCA [panel (b)] in the tibialis anterior muscle of the ischemic (gray, dashed) and control limb (black, solid). Dots represent measured raw DCE-MRI data and the curves represent the fits with the optimal pharmacokinetic model. For the MCA, the Patlak model was used, whereas the SCA curve was fitted with the generalized kinetic model.
Averaged elements of the covariance matrix for fits with the extended generalized kinetic model (upper panels) and the optimal two-parameter models [MCA: Patlak, panel (c); SCA: GKM, panel (d)] for the control tibialis anterior muscle. Diagonal elements represent the percentage parameter estimation uncertainties and off-diagonal elements represent parameter interdependencies. A positive interdependency between two parameters indicates that underestimation of one parameter is likely to be accompanied by underestimation in the other, while a negative interdependency tells that underestimation in one parameter is often accompanied by overestimation of the other (and vice versa). For a high interdependency, the chance of the occurrence of this effect is high.
Pharmacokinetic parameters of the different muscle groups as estimated with the optimal models. With the MCA, pharmacokinetic modeling gave estimations for [panel (a)] and [panel (c)], and with the SCA [panel (b)] and [panel (d)] were estimated. Error bars indicate the standard error of the mean. Significant differences between muscle groups were indicated with .
Pharmacokinetic parameters [panels (a) and (b)], [panels (c) and (d)], and [panels (e) and (f)] of the different muscle groups as estimated with the extended generalized kinetic model. Error bars indicate the standard error of the mean. Significant differences between muscle groups were indicated with .
AUC values in the different muscle groups for the MCA [panel (a)] and the SCA [panel (b)]. For each contrast agent, the AUC that corresponded best with was chosen (SCA: ; MCA: ). Error bars indicate the standard error of the mean. Significant differences are indicated with .
Standardized regression coefficients in multivariate linear regression between AUC at , 3, 6, and 12 min after contrast injection, and pharmacokinetic parameters estimated with the optimal model. For the MCA, the contribution of increased with increasing , while the influence of decreased. Alternatively, for the SCA the contribution of decreased with increasing , whereas increasingly contributed. Error bars indicate the standard error of the mean. Parameters that contributed significantly most are marked with .
MR images obtained with the DCE-MRI protocol for the MCA [panel (a)] and SCA [panel (b)]. Voxel size was for the MCA and for the SCA. ROIs delineated in black indicate the soleus. As a reference a high resolution -weighted image with voxel size is shown [panel (c)].
Mathematical description of GKM (Ref. 23), Patlak model (Ref. 25), and extended GKM (Ref. 24) used for the analysis. [: Volume transfer constant; : Volume fraction of the extracellular extravascular space; : Plasma volume fraction; : Tissue concentration time curve; and : Plasma concentration time curve.]
Acquisition parameters for DCE-MRI with medium-sized contrast agent (MCA) and small contrast agent (SCA).
Pharmacokinetic parameters estimated with various pharmacokinetic models and the corresponding fit errors ( error) for the tibialis muscle in the control limb. indicate the best fit error for the two-parameter models. Expansion to the three-parameter extended generalized kinetic model (extended GKM) gave a significant improvement of the fit error for the SCA.
Article metrics loading...
Full text loading...