Optimizaiton of Quantitative Susceptibility Mapping of Human Brain using Magnetic Resonance Imaging

論文翻譯標題: 利用核磁共振影像達成大腦組織磁化率定量之最佳化
  • 王 俊坤

學生論文: Master's Thesis


Recent studies have shown that Quantitative Susceptibility Mapping (QSM) of brain tissues has great potential for neuronal diseases Researchers have proposed various algorithms to reduce the influence of the streaking artifact which is generated during dipole model deconvolution calculation One of the main purposes of these algorithms is to reduce the impact of the streaking artifact while simultaneously keeping anatomical structures on susceptibility maps as intact as possible On the other hand in order to make QSM widely applicable for clinical diagnosis and research we use Generalized Autocalibraing Partially Parallel Acquisition (GRAPPA) to accelerate the image acquisition However whether acceleration using GRAPPA will have any influence on QSM remains to be studied This study proposed automatic streaking artifact detection to achieve optimal QSM reconstruction results We also study the influence of GRAPPA acceleration on QSM Results show that streaking pattern detection not only can be used for parameter optimization in QSM algorithms but also provide an objective method for comparing residual streaking artifacts in different QSM algorithms In addition results from QSM accelerated with GRAPPA show that there are visually perceptible aliasing artifacts on susceptibility maps with R=3 or higher even though in ROI analysis the measured susceptibility values seem to be less affected Therefore it is concluded that GRAPPA with R=2 results in accelerated image acquisition for reliable QSM reconstruction for clinical applications
獎項日期2015 二月 5
監督員Ming-Long Wu (Supervisor)