基于扩散光谱成像的脑白质纤维束走向跟踪  

Tracking of Crossing White Matter Fiber Bundles Based on Diffusion Spectrum Imaging

作　　者：吴占雄[1] 李珣[2] 何志伟[1] 

机构地区：[1]杭州电子科技大学电子信息学院,浙江杭州310018 [2]杭州电子科技大学计算机学院,浙江杭州310018

出　　处：《杭州电子科技大学学报（自然科学版）》2014年第6期43-46,共4页Journal of Hangzhou Dianzi University：Natural Sciences

基　　金：国家自然科学基金资助项目(51207038;51107024);浙江省自然科学基金资助项目(Q12F010017)

摘　　要：扩散张量成像对于每一体元只体现水分子自扩散宏观各向异性,不能实现神经纤维束走向分叉跟踪。为了实现分叉跟踪,更加真实地揭示脑神经网络连接情况。该文基于扩散光谱成像技术,提出一种白质纤维束走向分叉跟踪新方法。首先对扩散加权数据施加傅立叶变换得到扩散光谱图像,再计算每一体元扩散概率函数并得到其所有局部最大值,最后设置跟踪步长、走向角度阈值等参数,通过在相邻体元间寻找平滑连接向量以实现纤维束分叉走向跟踪。最后采用实际数据进行计算验证,扩散光谱成像方法跟踪结果更加符合脑组织生理结构特点。从实际计算结果可以看出,扩散光谱成像在白质交叉纤维束跟踪上比扩散张量成像有效合理,但是计算过程比后者复杂很多,不利于临床应用。The tracking of crossing WM fiber bundles cannot be accomplished based on diffusion tensor imaging because it just only reflects the macroscopic self-diffusion characteristics of water molecules. In order to reveal brain-network connections, this paper put forward with a new method to track crossing fiber bundles based on diffusion spectrum imaging. Firstly DSI data was get from DWI data through Fourier transformation, and then all local maximums of diffusion probability function of each voxel. Finally smooth connecting vectors of adjacent voxels were found to finish tracking under the constraint of tracking step and angel threshold. This method was tested on a volunteer, and more realistic results were got. The DSI method is better than DTI in tracking of crossing fiber bundles, but it is too complicated to clinical application at present.

关 键 词：扩散张量 扩散光谱成像 脑白质 纤维束 

分 类 号：Q64[生物学—生物物理学]