請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/27491
標題: | 螺旋槳式磁振造影的進階發展與應用 PROPELLER keeps spinning: Advanced technique development and applications of PROPELLER MRI |
作者: | Tzu-Chao Chuang 莊子肇 |
指導教授: | 鍾孝文(Hsaio-Wen Chung) |
關鍵字: | 螺旋槳式磁振造影,面迴訊影像,平行影像技術,影像模糊,擴散影像,資料共用, PROPELLER,EPI,parallel imaging,image blurring,diffusion imaging,,view-sharing, |
出版年 : | 2007 |
學位: | 博士 |
摘要: | 在這篇論文中,我們提出了兩種方式來有效減少磁化率不均在螺旋槳式面迴訊磁振影像中所產生的影像模糊。首先在影像重建方面,k空間權重函數強調失真較少的資料在合併時的重要性,使得影像對比更為銳利。另外在收取資料方面,我們搭配圓形對稱的多通道線圈使用平行影像技術,來增加每張槳片在相位方向的頻寬,以大幅改善螺旋槳式面迴訊影像中可能的模糊效果。這兩種方法可以同時運用,以獲得更佳的影像品質。在這個研究中,模擬結果和實驗數據證實了所提方法的有效性,並將其應用在高強度磁場下,獲取高空間解析度的擴散影像。本論文第二個部分的重點是「資料共用形式的螺旋槳磁振造影」的研發,以及其於快速動態對比影像的應用。在維持原本的取樣方式下,利用k空間的中心不斷地被每個槳片所取樣的特性,資料共用形式的螺旋槳磁振造影可以重組出與原本完整取樣之影像組相較起來,更高時間解析度且維持空間解析度和影像對比的動態影像。研究結果顯示該方法成功地增快影像擷取速度達五倍之多,並且取樣不完全所造成的假影也有效地被我們所提出的「槳片組合最佳化」所消除。 Two approaches of methods were proposed in this thesis to reduce the image blurring induce by field inhomogeneities in PROPELLER EPI. For image reconstruction, the k-space weighting scheme emphasizes data with less off-resonance errors to exhibit better sharpness. In addition, parallel imaging was applied in data acquisition to increase the bandwidth along phase encoding direction in each blade, leading to dramatic reduction of blurring effect in PROPELLER echo planar images. Numerical simulations and in-vivo experiments were demonstrated in this study that PROPELLER EPI with the help of both methods, which are compatible to each other, is suited for high resolution diffusion imaging at high fields. The second part of this thesis is view-sharing PROPELLER (VS-Prop) technique aiming at high frame-rate dynamic contrast-enhanced studies. With the central k-space being updated by every blade, VS-Prop reconstructs a series of images at increased temporal resolution, and meanwhile, the outer k-space was refreshed at a faster rate compared with traditional “keyhole” techniques. Results revealed that dynamic images obtained by VS-Prop at 5-fold acceleration demonstrate high consistency with full-k-space images, and the under sampling artifacts can be substantially reduced by pixel-based optimal blade selection (POBS) algorithm. |
URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/27491 |
全文授權: | 有償授權 |
顯示於系所單位: | 電機工程學系 |
文件中的檔案:
檔案 | 大小 | 格式 | |
---|---|---|---|
ntu-96-1.pdf 目前未授權公開取用 | 5.09 MB | Adobe PDF |
系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。