Bayesian modeling of Dynamic Contrast Enhanced MRI data in cerebral glioma patients improves the diagnostic quality of hemodynamic parameter maps - Fig 2

Representative K1 (permeability rate), Ktrans (leakage-flow), Ve (interstitial volume), vp (plasma volume), and Fp (plasma flow) images of a patient with glioblastoma multiforme (A-C) and astrocytoma grade II (D-F). Dynamic Contrast Enhanced MRI data are fitted to the 2-compartment-exchange model (2CXM; B and E) and the extended Toft’s model (ETM; C and F) using either the Levenberg-Marquardt (LM) method or Bayesian modeling (BM). Conventional images (contrast enhanced T1, CE T1, and T2 FLAIR) are demonstrated in (A) and (D).