Development and implementation of techniques for dynamic contrast enhancement MRI

Abstract

Magnetic Resonance Imaging (MRI) has been recognized as a non-ionizing imaging newline newlinemodality used in clinical diagnosis and prognosis. Usage of MR Contrast Agent (CA) in MRI has proven its ability in improvisation of image visualization. The most commonly used CAs are Gadolinium (Gd) - based to interpret microcirculatory properties of the tissues. Dynamic Contrast Enhancement Magnetic Resonance Imaging (DCE-MRI) is a well-established technique for non-invasive diagnosis and therapeutic monitoring of pathologies by administering intravenous CA. DCE-MRI devises a prospective to create images of physiological quantities such as blood vessel volume fraction, blood flow and blood vessel permeability which are extensively used in areas such as oncology. newline newline newlineQuantifying PharmacoKinetic (PK) maps can be accomplished by applying compartmental newline newlinemodels appropriate to the pathophysiology of tissue to be examined. This method is time-consuming because of the post-processing of a large amount of data. Therefore there is a need to accelerate the determination of PK maps accurately compared to the conventional methods. newline newline newlineFirstly, this work aims at a new frequency domain-based approach applied to the newline newlineconventional Tofts Model in Time-domain (TM-TD) to accelerate the determination of PK maps. This method has been demonstrated in silico and in vivo DCE-MRI data. Determining PK parameters consist of fitting time concentration data to these models. Tofts Model in Frequency Domain (TM-FD) is applied to an uncertainly vascularized tissue such as the breast. It results as a convolution-free model from a conventional TM-TD, which in turn reduces the complexity of the curve fitting problem from 2 dimensions (surface fit) to 1 dimension (curve fit). This technique not only reduces computation time but also increases robustness due to the reduced complexity of the fitting process. newline