Texture-based characterization of arterialization in simulated MRI of hypervascularized liver tumors

Abstract

The use of quantitative imaging for the characterization of hepatic tumors in MRI can improve the diagnosis and therefore the treatment. However, image parameters remain difficult to interpret because they result from a mixture of complex processes related to pathophysiology and to acquisition. In particular, the lesion arterialization is prominent in the resulting contrast between normal and tumoral tissues in contrast-enhanced images. In order to identify this influence, we propose a multiscale model of liver dynamic contrast-enhanced MRI, consisting of a model of the organ coupled with a model of the image acquisition. A sensitivity analysis of the model to the arterial flow has enabled us to emphasize the existence of relationships between texture parameters in simulated arterial-phase MR images, and the arterialization phenomena involved in carcinogenesis.