Book Description
MRI methods to evaluate diffuse hepatic diseases have advanced substantially in recent years, with a particular focus on chemical shift encoded (CSE)-MRI. However, the systematic evaluation of these methods requires the development of highly controlled quantitative imaging test objects ("phantoms"). In addition, CSE-MRI methods face major limitations, particularly their high sensitivity to physiological motion, and challenges in measuring parameters such as magnetic susceptibility (an emerging biomarker for tissue iron deposition). In this thesis, technical developments are introduced to address these challenges in both quantitative imaging phantoms and quantitative liver MRI methods. Novel quantitative MRI phantoms were designed and developed to mimic the simultaneous presence of fat, iron, and fibrosis in the liver by controlling different MR biomarkers while mimicking in vivo liver signals. An MRI and CT compatible fat phantom was developed and evaluated in a round-robin multi-center multi-vendor study. Motion-robust free breathing 2D sequential CSE-MRI method with flip angle modulation (FAM) and centric encoding approach was developed to achieve high SNR performance and minimal T1 bias simultaneously for liver fat and iron quantification. To further boost SNR performance, motion-corrected averaging method was combined with FAM CSE-MRI. Finally, a recently proposed liver quantitative susceptibility mapping (QSM) method was evaluated across iron levels and acquisition protocols at 1.5T and 3.0T, using superconducting quantum interference device (SQUID) liver susceptometry as the reference. Overall, these developments in this thesis have advanced MRI of the liver towards becoming a truly quantitative imaging modality.