2. What is the importance of acquiring physiological data both for noise and signals of interest?

Acquiring physiological data during neuroimaging experiments is crucial for enhancing the quality and interpretability of neuroimaging data. The main motivations are:

1. To model & remove the effects of general physiological fluctuations on neural fluctuations of interest in blood oxygen level dependent (BOLD) fMRI signal
2. To model the relationship between this physiological fluctuation and a fluctuation of interest in the neural signal (e.g. to see general vascular effects or vigilance)
3. To monitor the state of the participant in real time

This section provides a succinct introduction to which physiological data are typically recorded during an fMRI experiment, how these signals are recorded, and how these signals can improve our modeling of fMRI time series data. This is an active field of research, and we encourage all researchers to get the latest recommendations prior to initiating a new study. For a more in-depth article on similar subject matter, please see Bulte and Wartolowska (2017) Neuroimage 154:81-91.

Physiological monitoring is key to understanding the physiological sources of signal variance in fMRI data. Monitoring physiology during scanning is critical to enable the characterization of a given subject’s physiologic state at the time of the scan and to track variations in physiology throughout the scan. With these data, we can more accurately model how these factors manifest in the fMRI signal time series.

Depending on the research question of the imaging experiment, physiological fluctuations can be identified as either “noise” or as “signals of interest”. For most fMRI experiments, the goal is to isolate signal fluctuations in the brain that are associated with a neural stimulus and the resulting hemodynamic response (Caballero-Gaudes and Reynolds, 2017). In these data, it is important to model and remove signals with a remote, non-neural origin, such as breathing or cardiac related signal variance. Removing these confounds will improve the sensitivity and confidence of the fMRI analysis. In some fMRI experiments, the goal is to characterize a physiologic effect. For example, several studies that rely on gas concentrations to map cerebrovascular reactivity aim to quantify the dilation of blood vessels during certain non-neural stimuli (Liu et al., 2019; Pinto et al., 2021). Furthermore, for neuroimaging studies that aim to study arousal/vigilance or autonomic system features, physiological signals may also be of interest. In these studies, it is essential that the relevant physiologic parameters are recorded so that the analysis produces robust, quantitative physiological parameter maps.

Another benefit of collecting physiological data is that it provides a method to monitor the subject and/or patient during the scan in real-time. Any sudden changes in the different aspects being monitored can help those in the control room identify if the person is experiencing discomfort or complying with the scan protocol. Looking out for these changes is particularly helpful during an individual’s first MRI scan, when they may react poorly to the scan environment. In some protocols, tracking physiology in real-time can ensure that values stay within safe limits, in accordance with what is approved by your Institutional Review Board (IRB) or ethics committee. Real-time monitoring can also provide an opportunity to rescan the participant if problems are noted.

Although current modeling of physiology is imperfect, and fMRI signal processing techniques do not yet accurately factor in all physiologic signals, the field continues to develop, and our modeling continues to improve. We encourage all fMRI researchers to collect these data to more fully capture the variable human physiology inherent to imaging experiments.