定量评价弥散张量成像预处理联合双张量无损卡尔曼滤波算法优化锥体束显像用于胶质瘤手术导航的可靠性研究

The pyramidal tracts are the most important motor pathways in the human brain. Diffusion tensor imaging (DTI) and single-tensor Fiber Assignment by Continuous Tracking (FACT) algorithm were widely used by neurosurgeons to visualize the pyramidal tracts, which gave an insight of the spatial relationship between tracts and tumor and guided the surgical resection as well. However, due to issues of DTI artifacts, crossing fiber and peritumoral edema, FACT algorithm cannot completely visualize the whole structure of pyramidal tracts and the consequent false negative results are posing huge potential threats to the patients. Our preliminary study suggests that the combination of comprehensive DTI pre-processing and newly developed two-tensor unscented Kalman filter (UKF) algorithm, can significantly improve the visualization of pyramidal tracts in the pathology of gliomas. In this prospective controlled study, we commence with utilizing the DTI pre-processing procedures combined with two-tensor UKF algorithm and FACT algorithm to track the pyramidal tracts, respectively. Then the 3D Slicer software is used to integrate the results into intraoperative navigation. Finally according to the “gold standard”- intraoperative electrical stimulation, the accuracy and reliability of tractography methods can be in vivo quantified and compared. The result of this project may provide solid scientific evidence regarding whether the new tractography method can be used to optimize pyramidal tracts delineation and used for glioma surgical navigation.

锥体束是人类大脑中最重要的运动传导束。临床上广泛应用弥散张量成像（DTI）和单张量连续追踪纤维分配（FACT）算法对其进行显像，协助神经外科医生在术前了解锥体束与肿瘤的空间关系并指导手术。但由于DTI伪影、交叉纤维和瘤周水肿等原因，FACT算法不能完整地可视化锥体束结构，假阴性结果对患者造成巨大的潜在手术风险。我们的前期研究提示，完善的DTI预处理流程联合最新开发的双张量无损卡尔曼滤波（UKF）算法可以显著提高胶质瘤病理情况下锥体束的可视化程度。为此，本项目拟采用前瞻性对比实验，分别使用DTI预处理联合双张量UKF算法和FACT算法对锥体束进行显像，之后利用3D Slicer软件将结果整合入术中导航，最终根据“金标准”术中电刺激结果在活体上定量验证和比较锥体束显像方法的准确性和可靠性，为回答能否借助新方法优化锥体束显像并将其用于胶质瘤手术计划和导航的科学问题提供可靠的参考依据。

锥体束是人类大脑中最重要的运动传导束。临床上广泛应用弥散张量成像(DTI)和单张量连续追踪纤维分配(FACT)算法对其进行显像，协助神经外科医生在术前了解锥体束与肿瘤的空间关系并指导手术。但由于DTI伪影、交叉纤维和瘤周水肿等原因，FACT算法不能完整地可视化锥体束结构，假阴性结果对患者造成巨大的潜在手术风险。我们的前期研究提示，完善的DTI预处理流程联合最新开发的双张量无损卡尔曼滤波(UKF)算法可以显著提高胶质瘤病理情况下锥体束的可视化程度。为此，本项目采用前瞻性对比实验，分别使用DTI预处理联合双张量UKF算法和FACT算法对锥体束进行显像，之后利用3D Slicer软件将结果整合入术中导航，最终根据“金标准”术中电刺激结果在活体上定量验证和比较锥体束显像方法的准确性和可靠性，为回答能否借助新方法优化锥体束显像并将其用于胶质瘤手术计划和导航的科学问题提供可靠的参考依据。我们的研究结果证实，新方法进行的锥体束追踪与电刺激的结果具有明显相关性，电刺激量与相距锥体束的距离正相关。降低GA值能够协助在交叉纤维区域获得更多的纤维追踪结果，对于PTs的扇形展开结构的描绘大有帮助，但同时，会带来更多的假阳性纤维，需要仔细甄别和删除。之前的基于纤维束追踪的定量相关性研究结果之间存在争议，且部分结果和本研究不同，一个可能的原因是选择的DWI预处理方法和纤维束追踪算法尚有缺陷。本研究得益于TORTOISE软件包和UKF算法，两者提供了更加严谨和全面的DWI预处理和纤维束描绘，所以能够发现相关性结果。

内容获取失败，请点击重试