On the analysis of rapidly sampled fMRI data

Abstract

Recent advances in parallel imaging and simultaneous multi-slice techniques have permitted whole-brain fMRI acquisitions at sub-second sampling intervals, without significantly sacrificing the spatial coverage and resolution. Apart from probing brain function at finer temporal scales, faster sampling rates may potentially lead to enhanced functional sensitivity, owing possibly to both cleaner neural representations (due to less aliased physiological noise) and additional statistical benefits (due to more degrees of freedom for a fixed scan duration). Accompanying these intriguing aspects of fast acquisitions, however, confusion has also arisen regarding (1) how to preprocess/analyze these fast fMRI data, and (2) what exactly is the extent of benefits with fast acquisitions, i.e., how fast is fast enough for a specific research aim? The first question is

Removal of high frequency contamination from motion estimates in single-band fMRI saves data without biasing functional connectivity

“…Sources of HF-motion in realignment parameters: factitious and true head motion Fair et al (2018) recently demonstrated the presence of high-frequency fluctuations in the motion parameters of a multiband dataset at frequencies that closely match respiration rates. This finding was corroborated by several additional reports of high frequency "motion" fluctuations in other multiband datasets (Chen et al, 2019;Etzel, 2016a;Inglis, 2016b;Power et al, 2019). At least two hypothetical mechanisms may link respiration with changes in realignment parameters.…”

“…Addressing HF-motion in fMRI analysis leads to data savings While the potential contamination of fMRI by high frequency respiration-related effects has been appreciated for nearly 20 years (Brosch et al, 2002;Durand et al, 2001;Raj et al, 2001;Van de Moortele et al, 2002), the availability of multiband sequences and the typical employment of realignment parameters to denoise fMRI data has caused renewed interest in this phenomenon (Chen et al, 2019;Fair et al, 2018;Power et al, 2019) (see also popular blog posts on this topic by Jo Etzel and Ben Inglis (Etzel, 2016a, b, c;Inglis, 2016a, b)). With the shorter TRs associated with multiband data it is possible to more clearly identify respiration-related content in realignment parameters, as Nyquist limits are higher and respiration rates do not alias to lower frequencies (Chen et al, 2019;Etzel, 2016a;Fair et al, 2018;Inglis, 2016b;Power et al, 2019). Moreover, many fMRI processing pipelines use measures of frame-toframe changes in realignment parameters as an estimate of participant head motion, and censor frames with even small amounts of head movement (e.g., 0.2 mm.)…”

“…The present findings suggest that filtering motion parameters would be beneficial in that context. Moreover, even processing strategies that do not explicitly censor high-motion frames may benefit from removing high frequency content from motion parameters if they do not track well with true head motion (Chen et al, 2019;Fair et al, 2018). As shown in Fig.…”

“…Quantification of high-frequency (HF) motion: HF-motion was quantified by determining the percent of relative power above 0.1 Hz within each motion direction. We focus on HF-motion in the y-translation (phase-encoding) direction given that factitious respiratory-related effects are largest in the phaseencoding direction (Brosch et al, 2002;Chen et al, 2019;Durand et al, 2001;Fair et al, 2018;Power et al, 2019;Raj et al, 2001;Van de Moortele et al, 2002).…”

“…However, realignment estimates may not be pure measures of head-motion and may be influenced by other factors such as respiration. Respiration may influence realignment estimates in two ways: (a) through true changes in head position caused by respiration-related movement and (b) apparent (factitious) motion in the phase-encoding direction generated by perturbations of the main (B0) magnetic field of T2* images caused by chest wall motion (Brosch et al, 2002;Chen et al, 2019;Durand et al, 2001;Fair et al, 2018;Power et al, 2019;Raj et al, 2001;Van de Moortele et al, 2002). Evidence suggests that factitious motion may be the predominant source of respiration-related motion artifacts in fMRI ((Brosch et al, 2002;Raj et al, 2001); elaborated on further in the discussion).…”