https://scholars.lib.ntu.edu.tw/handle/123456789/150267
標題: | 高夾角解析度擴散磁振造影神經徑路追蹤之發展與應用 The Development of High Angular Resolution Diffusion Tractography and its Applications |
作者: | 趙一平 Chao, Yi-Ping |
關鍵字: | 神經連結性;擴散磁振造影;高夾角解析度擴散磁振造影;神經徑路追蹤;Q球磁振造影;機率分割拓樸圖;細胞結構圖譜;胼肢體;擴散張量磁振造影;非等向擴散指標;neural connectivity;diffusion MRI;high angular resolution diffusion imaging;tractography;q-ball imaging;probabilistic topography;corpus callosum;fractional anisotropy;diffusion tensor imaging | 公開日期: | 2009 | 摘要: | 近幾年來,運用磁振造影技術來探討不同大腦功能區之間的複雜神經連結性,成為神經科學研究以及臨床神經醫學或是探討精神性疾病成因中相當重要的一個方法。擴散磁振造影由於具有非侵入式偵測神經纖維方向的優點,更是廣泛地被應用於解析複雜的神經連結。相較於傳統的擴散磁振造影技術,高夾角解析度擴散磁振造影技術能夠更精準地解析複雜的交叉神經結構,且其與解剖結構的相關性以及臨床應用的可行性也已驗證。然而,相對應的神經徑路追蹤卻仍有其限制及缺點。此,本論文的主要目的在於開發及應用適用於高夾角角解析度擴散磁振造影技術之神經徑路追蹤演算法,並可以分為以下兩部分。第一、開發多神經束追蹤演算法及驗證其應用於人腦Q球磁振造影之正確性。第二為透過所開發的演算法來建構出大腦胼胝體與大腦皮質功能區之相關連結,並透過分群連結來建立胼肢體的機率分割拓撲圖。最後我們利用擴散磁振影像中所提供的非等向性指標來估計胼肢體內各區域的神經纖維組成,並與解剖學所得之結果進行評估。第一部份的研究中,我們透過所建立的演算法來建構人腦中的複雜神經連結,相較於傳統擴散張量影像為基礎的神經徑路追蹤演算法,除了能夠正確解析穿越神經交錯處的連結之外,且其擷取之神經束也與大腦解剖神經連結圖譜相符。此外,我們更加以改進此演算法為以機率方式呈現神經連結的狀況,藉此提供腦內可能的神經網路訊息,提供更為有意義的資訊。二部分的研究則是透過比尼安?布洛德曼所提出的細胞結構圖譜(Brodmann’s cytoarchitectonic areas template)來定義出28個大腦皮質解剖區域,並應用前述的演算法來解構12個正常人大腦胼肢體至此28個區域的神經連結,並透過機率模式來顯示出胼肢體連結至各相對應皮質的區域。此外,我們更提出了一個以神經束為主體的轉換流程,可將各個體的神經連結由個體座標系轉換至標準的大腦圖譜上來進行分析,藉以進行群組的比較及產生機率模型的呈現方式。最後,我們利用第二部份所完成的胼肢體機率分佈拓譜圖來分析20位正常人的擴散張量磁振造影上的非等向擴散指標,並與已知胼肢體神經纖維組成的解剖證據進行評估,結果顯示出其分區的依據確實有其相關性,且能提供未來在神經發展、神經退化及神經疾病上的資訊。結而言,我們發展了新的高夾角解析度擴散磁振造影神經徑路追蹤演算法,也應用此技術於解析人腦複雜神經連結。此外,我們也建立了一個以連接至大腦皮質細胞結構區域的胼肢體連結機率分割拓樸圖譜,並應用此圖譜於分析胼肢體神經纖維結構組成。透過此技術發展及整合應用,相信對於未來神經科學的基礎研究或臨床診斷都會有相當的助益。 Since last few years, magnetic resonance imaging (MRI) has become an important approach for non-invasive map of functional organizations for neuroscience research and clinical applications. With the ability of mapping tissue anisotropy, diffusion MRI provides a novel aspect of depicting the structural connectivity between functional brain regions. Compared with conventional diffusion MRI techniques and diffusion tensor imaging (DTI), high angular resolution diffusion imaging (HARDI) provides more information for complex neural fiber architectures. Despite fruitful results have been acquired from the HARDI derived complex neural connectivity, some intrinsic problems still exist in the current multiple fiber tracking (MFT) algorithms. herefore, the objective of this dissertation is twofold. First, we has developed a modified MFT method, namely multiple fiber assignment by continuous tracking (MFACT), to trace complex neural trajectories within brain. The MFACT method was applied to map the reticular neural connections of human brain from HARDI derived fiber orientations. With appropriate tracking criteria, this method not only has the potential to map the sophisticated trajectories which are passing through the regions with fiber heterogeneous, but also can be adopted to extract the known fiber tract dispersions in a shorter time. Second, an extension of MFACT approach was proposed to estimate the probabilistic fiber trajectories. It offers a high sensitivity in mapping multi-fiber connections and has the advance to assess neural connections between subtle functional regions. sing these new developed methods, we have developed a probabilistic topography of the corpus callosum (CC) by connecting Brodmann’s cytoarchitectonic template and the CC. A topological relationship between 28 cerebral regions and the mid-sagittal CC was accordingly revealed. In addition, the probabilistic topography allows an assignment of quantitative distribution in the subdivision of the CC. Fractional anisotropy derived from DTI data of 20 healthy subjects was evaluated to study the correlation with neural composition in distinct CC regions which has explored by Aboitiz in 20 postmortem human brainsn conclusion, we have successfully developed a novel HARDI-based tractography algorithm and applied it to map probabilistic neural connectivity. From various cytoarchitectonic regions, sophisticated probabilistic population topography of the CC from 12 healthy subjects was accordingly identified. We also demonstrated that this topography may serve as a brain landmark to reveal the neural integrity among regional microstructure of the CC. These methods and applications may be potentially useful to facilitate the further neuroscience researches and clinical applications. |
URI: | http://ntur.lib.ntu.edu.tw//handle/246246/188005 |
顯示於: | 電機工程學系 |
檔案 | 描述 | 大小 | 格式 | |
---|---|---|---|---|
ntu-98-D91921025-1.pdf | 23.32 kB | Adobe PDF | 檢視/開啟 |
在 IR 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。