Fekonja Lucius S, Wang Ziqian, Aydogan Dogu B, Roine Timo, Engelhardt Melina, Dreyer Felix R, Vajkoczy Peter, Picht Thomas
Department of Neurosurgery, Charité-Universitätsmedizin Berlin, Berlin, Germany.
Cluster of Excellence: "Matters of Activity. Image Space Material", Humboldt-Universität zu Berlin, Berlin, Germany.
Front Oncol. 2021 Jan 27;10:622358. doi: 10.3389/fonc.2020.622358. eCollection 2020.
Tumors infiltrating the motor system lead to significant disability, often caused by corticospinal tract injury. The delineation of the healthy-pathological white matter (WM) interface area, for which diffusion magnetic resonance imaging (dMRI) has shown promising potential, may improve treatment outcome. However, up to 90% of white matter (WM) voxels include multiple fiber populations, which cannot be correctly described with traditional metrics such as fractional anisotropy (FA) or apparent diffusion coefficient (ADC). Here, we used a novel fixel-based along-tract analysis consisting of constrained spherical deconvolution (CSD)-based probabilistic tractography and fixel-based apparent fiber density (FD), capable of identifying fiber orientation specific microstructural metrics. We addressed this novel methodology's capability to detect corticospinal tract impairment. We measured and compared tractogram-related FD and traditional microstructural metrics bihemispherically in 65 patients with WHO grade III and IV gliomas infiltrating the motor system. The cortical tractogram seeds were based on motor maps derived by transcranial magnetic stimulation. We extracted 100 equally distributed cross-sections along each streamline of corticospinal tract (CST) for along-tract statistical analysis. Cross-sections were then analyzed to detect differences between healthy and pathological hemispheres. All metrics showed significant differences between healthy and pathologic hemispheres over the entire tract and between peritumoral segments. Peritumoral values were lower for FA and FD, but higher for ADC within the entire cohort. FD was more specific to tumor-induced changes in CST than ADC or FA, whereas ADC and FA showed higher sensitivity. The bihemispheric along-tract analysis provides an approach to detect subject-specific structural changes in healthy and pathological WM. In the current clinical dataset, the more complex FD metrics did not outperform FA and ADC in terms of describing corticospinal tract impairment.
浸润运动系统的肿瘤会导致严重残疾，通常是由皮质脊髓束损伤引起的。扩散磁共振成像（dMRI）已显示出在描绘健康与病变白质（WM）界面区域方面具有良好潜力，这可能会改善治疗效果。然而，高达90%的白质（WM）体素包含多种纤维群，使用传统指标如分数各向异性（FA）或表观扩散系数（ADC）无法正确描述这些纤维群。在此，我们采用了一种基于体素的新型沿束分析方法，该方法由基于约束球形反卷积（CSD）的概率性纤维束成像和基于体素的表观纤维密度（FD）组成，能够识别纤维方向特异性微观结构指标。我们探讨了这种新方法检测皮质脊髓束损伤的能力。我们对65例世界卫生组织III级和IV级浸润运动系统的神经胶质瘤患者进行了双侧测量和比较，测量并比较了与纤维束成像相关的FD和传统微观结构指标。皮质纤维束成像种子基于经颅磁刺激得出的运动图谱。我们沿着皮质脊髓束（CST）的每条流线提取100个均匀分布的横截面用于沿束统计分析。然后对横截面进行分析，以检测健康半球和病变半球之间的差异。所有指标在整个束以及瘤周节段的健康半球和病变半球之间均显示出显著差异。在整个队列中，瘤周FA和FD值较低，但ADC值较高。与ADC或FA相比，FD对CST中肿瘤诱导的变化更具特异性，而ADC和FA显示出更高的敏感性。双侧沿束分析提供了一种检测健康和病变白质中个体特异性结构变化的方法。在当前临床数据集中，就描述皮质脊髓束损伤而言，更复杂的FD指标并不优于FA和ADC。
