Center for Brain Health, Department of Psychiatry, New York University School of Medicine, 145 East 32nd Street, New York, NY 10016, USA; Comprehensive Epilepsy Center, Department of Neurology, New York University School of Medicine, 223 East 34th Street, New York, NY 10016, USA.
Center for Molecular and Behavioral Neuroscience, Rutgers, The State University of New Jersey, 197 University Avenue, Newark, NJ 07102, USA.
Neuroimage. 2014 Aug 15;97:245-51. doi: 10.1016/j.neuroimage.2014.04.026. Epub 2014 Apr 13.
Septal nuclei, located in basal forebrain, are strongly connected with hippocampi and important in learning and memory, but have received limited research attention in human MRI studies. While probabilistic maps for estimating septal volume on MRI are now available, they have not been independently validated against manual tracing of MRI, typically considered the gold standard for delineating brain structures. We developed a protocol for manual tracing of the human septal region on MRI based on examination of neuroanatomical specimens. We applied this tracing protocol to T1 MRI scans (n=86) from subjects with temporal epilepsy and healthy controls to measure septal volume. To assess the inter-rater reliability of the protocol, a second tracer used the same protocol on 20 scans that were randomly selected from the 72 healthy controls. In addition to measuring septal volume, maximum septal thickness between the ventricles was measured and recorded. The same scans (n=86) were also analyzed using septal probabilistic maps and DARTEL toolbox in SPM. Results show that our manual tracing algorithm is reliable, and that septal volume measurements obtained via manual and automated methods correlate significantly with each other (p<.001). Both manual and automated methods detected significantly enlarged septal nuclei in patients with temporal lobe epilepsy in accord with a proposed compensatory neuroplastic process related to the strong connections between septal nuclei and hippocampi. Septal thickness, which was simple to measure with excellent inter-rater reliability, correlated well with both manual and automated septal volume, suggesting it could serve as an easy-to-measure surrogate for septal volume in future studies. Our results call attention to the important though understudied human septal region, confirm its enlargement in temporal lobe epilepsy, and provide a reliable new manual delineation protocol that will facilitate continued study of this critical region.
隔核位于基底前脑,与海马体强烈相连,在学习和记忆中起着重要作用,但在人类 MRI 研究中受到的关注有限。虽然现在已经有了用于估计 MRI 上隔核体积的概率图,但它们尚未独立于 MRI 的手动追踪进行验证,而手动追踪通常被认为是勾画脑结构的金标准。我们根据神经解剖标本的检查,制定了一种在 MRI 上手动追踪人类隔核区域的方案。我们将该追踪方案应用于颞叶癫痫患者和健康对照者的 T1 MRI 扫描(n=86),以测量隔核体积。为了评估方案的组内可靠性,第二名追踪者使用相同的方案对 72 名健康对照者中随机选择的 20 个扫描进行了追踪。除了测量隔核体积外,还测量并记录了脑室之间的最大隔核厚度。同样的扫描(n=86)也使用隔核概率图和 SPM 中的 DARTEL 工具箱进行了分析。结果表明,我们的手动追踪算法是可靠的,并且通过手动和自动方法获得的隔核体积测量值彼此之间具有显著相关性(p<.001)。手动和自动方法都检测到颞叶癫痫患者的隔核明显增大,这与隔核和海马体之间的强烈连接相关的拟补偿性神经可塑性过程一致。隔核厚度很容易测量,组内可靠性也很好,与手动和自动隔核体积都有很好的相关性,这表明它可以作为未来研究中隔核体积的一种易于测量的替代指标。我们的研究结果引起了对重要但研究较少的人类隔核区域的关注,证实了其在颞叶癫痫中的增大,并提供了一种可靠的新手动勾画方案,将有助于对这一关键区域的进一步研究。