Benjamin Christopher F A, Singh Jolene M, Prabhu Sanjay P, Warfield Simon K
Harvard Medical School, Boston, Massachusetts; Department of Radiology, Boston Children's Hospital, Boston, Massachusetts; Semel institute, UCLA, Los Angeles, California.
Hum Brain Mapp. 2014 Feb;35(2):683-97. doi: 10.1002/hbm.22204. Epub 2012 Dec 8.
Imaging and delineation of the optic radiations (OpRs) remains challenging, despite repeated attempts to achieve reliable validated tractography of this complex structure. Previous studies have used varying methods to generate representations of the OpR which differ markedly from one another and, frequently, from the OpR's known structure. We systematically examined the influence of a key variable that has differed across previous studies, the tractography seed region, in 13 adult participants (nine male; mean age 31 years; SD 8.7 years; range 16-47). First, we compared six seed regions at the lateral geniculate nucleus (LGN) and sagittal stratum based on the literature and known OpR anatomy. Three of the LGN regions seeded streamlines consistent with the OpR's three "bundles," whereas a fourth seeded streamlines consistent with each of the three bundles. The remaining two generated OpR streamlines unreliably and inconsistently. Two stratum regions seeded the radiations. This analysis identified a set of optimal regions of interest (ROI) for seeding OpR tractography and important inclusion and exclusion ROI. An optimized approach was then used to seed LGN regions to the stratum. The radiations, including streamlines consistent with Meyer's Loop, were streamlined in all cases. Streamlines extended 0.2 ± 2.4 mm anterior to the tip of the anterior horn of the lateral ventricle. These data suggest some existing approaches likely seed representations of the OpR that are visually plausible but do not capture all OpR components, and that using an optimized combination of regions seeded previously allows optimal mapping of this complex structure.
尽管人们多次尝试对这一复杂结构进行可靠且经过验证的神经束成像，但对视辐射（OpRs）的成像和描绘仍然具有挑战性。以往的研究采用了不同的方法来生成视辐射的表征，这些表征彼此之间存在显著差异，并且常常与视辐射的已知结构不同。我们系统地研究了一个在以往研究中存在差异的关键变量——神经束成像种子区域——对13名成年参与者（9名男性；平均年龄31岁；标准差8.7岁；年龄范围16 - 47岁）的影响。首先，我们根据文献和已知的视辐射解剖结构，比较了外侧膝状体（LGN）和矢状层的六个种子区域。LGN的三个区域植入的流线与视辐射的三个“束”一致，而第四个区域植入的流线与这三个束中的每一个都一致。其余两个区域生成的视辐射流线不可靠且不一致。矢状层的两个区域植入了视辐射。该分析确定了一组用于视辐射神经束成像植入的最佳感兴趣区域（ROI）以及重要的包含和排除ROI。然后使用一种优化方法从LGN区域向矢状层植入种子。在所有情况下，包括与迈耶袢一致的流线在内的视辐射都被描绘成了流线。流线延伸到侧脑室前角尖端前方0.2±2.4毫米处。这些数据表明，一些现有的方法可能植入了在视觉上看似合理但未捕捉到所有视辐射成分的视辐射表征，并且使用先前植入区域的优化组合能够实现对这一复杂结构的最佳映射。
