Hemodynamics in Normal and Diseased Livers: Application of Image-Based Computational Models

This investigation demonstrates the utility of image-based computational models in portal venous hemodynamics. The long-term objective is to develop methodologies based upon noninvasive imaging and hemodynamic computational models for blood flow in major vessels of the liver that will significantly augment and improve current practices in clinical care. Magnetic resonance imaging (MRI) and computational fluid dynamics (CFD) were used to investigate liver hemodynamics. MRI data were obtained in 7 healthy subjects and 4 patients diagnosed with cirrhosis, and computational models were developed and validated for two healthy subjects and two patients. Additional simulations of post-prandial hemodynamics and portal hypertension were completed. The MRI studies identified several new parameters (portal vein V_avg/total liver volume, V_var, splenic vein flow rate per total liver volume, and % splenic flow/portal vein flow) that offer statistical differentiation between healthy subjects and patients with liver disease. Computational models were used to calculate the contribution of blood supply to the right and left lobes of the liver derived from the superior mesenteric vein (greater in healthy subjects vs. patients); and simulate post-prandial conditions and progressive portal hypertension. CFD offers a tool to test hypotheses without the acquisition of additional data and elucidate hemodynamic effects as disease progresses. In addition, several new MRI derived parameters have been identified as having promise to distinguish between healthy and patient groups and, potentially, to monitor disease progression.