Rotman Research Institute at Baycrest Centre, Canada; Department of Medical Biophysics, University of Toronto, Canada.
Rotman Research Institute at Baycrest Centre, Canada.
Neuroimage. 2018 Jun;173:72-87. doi: 10.1016/j.neuroimage.2018.02.004. Epub 2018 Feb 13.
The blood-oxygenation level dependent (BOLD) functional magnetic resonance imaging (fMRI) signal is commonly used to assess functional connectivity across brain regions, particularly in the resting state (rs-fMRI). However, the BOLD fMRI signal is not merely a representation of neural activity, but a combination of neural activity and vascular response. These aspects of the BOLD signal are easily influenced by systemic physiology, potentially biasing BOLD-based functional connectivity measurements. In this work, we focus on the following physiological modulators of the BOLD signal: cerebral blood flow (CBF), venous blood oxygenation, and cerebrovascular reactivity (CVR). We use simulations and experiments to examine the relationship between the physiological parameters and rs-fMRI functional connectivity measurements in three resting-state networks: default mode network, somatosensory network and visual network. By using the general linear model, we demonstrate that physiological modulators significantly impact functional connectivity measurements in these regions, but in a manner that depends on the interplay between signal- and noise-driven correlations. Moreover, we find that the physiological effects vary by brain region and depend on the range of physiological conditions probed; the associations are more complex than previously reported. The results confirm that it is important to account for the effect of physiological modulators when comparing resting-state fMRI metrics. We note that such modulatory effects may be amplified by disease conditions, which will warrant future investigations.
血氧水平依赖 (BOLD) 功能磁共振成像 (fMRI) 信号通常用于评估脑区之间的功能连接，特别是在静息状态 (rs-fMRI) 下。然而，BOLD fMRI 信号不仅仅是神经活动的代表，而是神经活动和血管反应的组合。BOLD 信号的这些方面很容易受到全身生理学的影响，可能会对基于 BOLD 的功能连接测量产生偏差。在这项工作中，我们专注于以下 BOLD 信号的生理调节剂：脑血流量 (CBF)、静脉血氧饱和度和脑血管反应性 (CVR)。我们使用模拟和实验来检查三个静息态网络（默认模式网络、躯体感觉网络和视觉网络）中生理参数与 rs-fMRI 功能连接测量之间的关系。通过使用广义线性模型，我们证明生理调节剂显著影响这些区域的功能连接测量，但影响方式取决于信号和噪声驱动相关性之间的相互作用。此外，我们发现生理效应因脑区而异，并且取决于所探测的生理条件范围；关联比以前报道的更复杂。结果证实，在比较静息态 fMRI 指标时，考虑生理调节剂的影响很重要。我们注意到，在疾病状态下，这种调节效应可能会被放大，这将需要进一步的研究。
