Reconstruction and modelling of dynamic contrast-enhanced MRI From k-space to K^trans

Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) is a promising technique for the non-invasive assessment of tissue perfusion and vascular permeability, with potential applications in disease diagnosis, treatment planning, and therapy monitoring. Despite its promise, quantitative DCE-MRI has not yet been widely adopted in clinical practice because of challenges related to image acquisition, reconstruction, modelling, and reproducibility.
This thesis addresses several of these challenges by developing and evaluating methods to improve the accuracy, precision, and robustness of quantitative DCE-MRI. First, a correction to the extended Tofts-Kety model is proposed to account for time-averaging effects introduced by modern non-Cartesian MRI acquisition schemes, resulting in more accurate pharmacokinetic parameter estimation. Next, a motion-corrected reconstruction framework is presented to reduce image degradation caused by patient motion during dynamic acquisitions. Building on this work, a model-based reconstruction approach is introduced that directly estimates pharmacokinetic parameters from raw k-space data, reducing the trade-off between spatial and temporal resolution. Finally, deep learning methods are investigated to accelerate quantitative parameter estimation while providing uncertainty estimates, enabling users to identify regions where model predictions are less reliable.
Beyond these technical developments, the thesis discusses the broader challenges that must be addressed for successful clinical translation of quantitative DCE-MRI, including image quality, repeatability and reproducibility, standardization of acquisition and post-processing methods, and clinical validation. Together, the work presented in this thesis contributes to a more robust methodological foundation for quantitative DCE-MRI and provides directions for future research aimed at establishing the technique as a reliable quantitative imaging biomarker in clinical practice.