Optimized EPI for fMRI using a slice-dependent template-based gradient compensation method to recover local susceptibility-induced signal loss
Optimized EPI for fMRI using a slice-dependent template-based gradient compensation method to recover local susceptibility-induced signal loss
复制标题
DOI：
10.1007/s10334-010-0215-x
复制
发表时间：
2010-06-01
影响因子：
2.3
通讯作者：
Zaitsev, Maxim
中科院分区：
文献类型：
--
作者：
Rick, Jochen;Speck, Oliver;Zaitsev, Maxim
大多数功能磁共振成像（fMRI）实验使用梯度回波回波平面成像（GE EPI）来检测血氧水平依赖（BOLD）效应。这种技术可能会失败的解剖结构相关的磁化率感应场梯度在人的头部的存在。在这项工作中，我们提出了一种新的3D补偿方法，结合基于模板的校正，可以在特定的感兴趣区域进行优化，以恢复可测量性引起的信号丢失，而不会造成采集时间损失。基于对8名受试者的B-0场图的评估，推导出两个受损区域中最大BOLD灵敏度的切片相关梯度补偿矩：眶额皮质和杏仁核区域改进的EPI序列在所有三个成像方向上使用这些附加梯度矩。该方法相比，非补偿，模板为基础的和受试者特定的校正梯度，也在屏气experiments.The切片相关的梯度补偿方法显着提高信号强度/BOLD的灵敏度约35/43%，在眶额皮质和杏仁核地区相比，传统的收购17/30%。基于模板的校正和特定于主题的校正同样表现良好。在屏气实验中，补偿区域的BOLD灵敏度得到了有效提高，新方法在不影响时间分辨率的前提下，解决了可伸缩性引起的信号丢失问题。它可用于与事件相关的功能实验，而无需额外的特定于受试者的校准或计算时间。
Most functional magnetic resonance imaging (fMRI) experiments use gradient-echo echo planar imaging (GE EPI) to detect the blood oxygenation level-dependent (BOLD) effect. This technique may fail in the presence of anatomy-related susceptibility-induced field gradients in the human head. In this work, we present a novel 3D compensation method in combination with a template-based correction that can be optimized over particular regions of interest to recover susceptibility-induced signal loss without acquisition time penalty.Based on an evaluation of B-0 field maps of eight subjects, slice-dependent gradient compensation moments are derived for maximal BOLD sensitivity in two compromised regions: the orbitofrontal cortex and the amygdala areas. A modified EPI sequence uses these additional gradient moments in all three imaging directions. The method is compared to non-compensated, template-based and subject-specific correction gradients and also in a breath-holding experiment.The slice-dependent gradient compensation method significantly improves signal intensity/BOLD sensitivity by about 35/43% in the orbitofrontal cortex and by 17/30% in the amygdala areas compared to a conventional acquisition. Template-based correction and subject-specific correction perform equally well. The BOLD sensitivity in the breath hold experiment is effectively increased in compensated regions.The new method addresses the problem of susceptibility-induced signal loss, without compromising temporal resolution. It can be used for event-related functional experiments without requiring additional subject-specific calibration or calculation time.