Effect of trial-to-trial variability on optimal event-related fMRI design: Implications for Beta-series correlation and multi-voxel pattern analysis.
Effect of trial-to-trial variability on optimal event-related fMRI design: Implications for Beta-series correlation and multi-voxel pattern analysis.
复制标题
DOI：
10.1016/j.neuroimage.2015.11.009
复制
发表时间：
2016-01-15
期刊：
影响因子：
5.7
通讯作者：
Henson RN
中科院分区：
文献类型：
--
作者：
Abdulrahman H;Henson RN
功能性磁共振成像（fMRI）研究通常采用快速的，事件相关的设计行为的原因，并与统计效率的原因。根据一般线性模型（GLM）估计的参数（Beta）的精度计算效率，其中试验开始与血流动力学反应函数（HRF）进行卷积。然而，以前的效率计算忽略了试验之间神经反应的可能变化，例如由于注意力波动或试验之间的不同刺激。在这里，我们比较了三种GLM在估计试验中的平均和个体Beta的效率，作为试验变异性，扫描噪声和刺激发作不同步（SOA）的函数：“最小二乘全部”（LSA），“最小二乘分离”（LSS）和“最小二乘酉”（LSU）。对个体试验的响应的估计对于使用“β系列相关性”和“多体素模式分析”（MVPA）的功能连接性特别重要。我们的模拟表明，试验间变异性与扫描噪声的比值影响最佳SOA和最佳GLM，特别是对于< 5 s的短SOA：当该比值较高时，LSA更好，而当该比值较低时，LSS和LSU更好。对于MVPA，试验变异性和扫描噪声的体素之间的一致性也至关重要。这些发现不仅对使用Beta系列回归和MVPA的实验设计具有重要影响，而且对仅寻求有效估计试验平均响应的统计参数映射研究也具有重要影响。尝试式连接和多体素模式分析可以使用不同的GLM。试验间变异性影响最优GLM和最优SOA。最优GLM和SOA依赖于总体或样本均值的估计。用于单体素分析的最佳GLM取决于试验与扫描变异性的比率。用于模式分析的最佳GLM取决于跨体素的可变性的一致性。
Functional magnetic resonance imaging (fMRI) studies typically employ rapid, event-related designs for behavioral reasons and for reasons associated with statistical efficiency. Efficiency is calculated from the precision of the parameters (Betas) estimated from a General Linear Model (GLM) in which trial onsets are convolved with a Hemodynamic Response Function (HRF). However, previous calculations of efficiency have ignored likely variability in the neural response from trial to trial, for example due to attentional fluctuations, or different stimuli across trials. Here we compare three GLMs in their efficiency for estimating average and individual Betas across trials as a function of trial variability, scan noise and Stimulus Onset Asynchrony (SOA): “Least Squares All” (LSA), “Least Squares Separate” (LSS) and “Least Squares Unitary” (LSU). Estimation of responses to individual trials in particular is important for both functional connectivity using “Beta-series correlation” and “multi-voxel pattern analysis” (MVPA). Our simulations show that the ratio of trial-to-trial variability to scan noise impacts both the optimal SOA and optimal GLM, especially for short SOAs < 5 s: LSA is better when this ratio is high, whereas LSS and LSU are better when the ratio is low. For MVPA, the consistency across voxels of trial variability and of scan noise is also critical. These findings not only have important implications for design of experiments using Beta-series regression and MVPA, but also statistical parametric mapping studies that seek only efficient estimation of the mean response across trials. Trial-wise connectivity and multi-voxel pattern analyses can use different GLMs. Trial-to-trial variability affects both the optimal GLM and optimal SOA. Optimal GLM and SOA depend on estimation of population or sample mean. Optimal GLM for single voxel analysis depends on ratio of trial- to scan-variability. Optimal GLM for pattern analysis depends on coherency of variability across voxels.