Katherine Bottenhorn1, Taylor Salo1, Michael Riedel1, Erica Musser1, Jennifer Robinson2, Matthew Sutherland1, Angela Laird1
1Florida International University, Miami, FL, 2Auburn University, Auburn, AL
The utility of physiological data collected during functional magnetic resonance imaging (fMRI) is twofold, both in removing non-neural physiological noise from fMRI data and in studying psychophysiological phenomena. However, MR sequences impart significant artifacts on simultaneously-collected physiological data, obscuring phenomena of interest and necessitating effective data cleaning approaches. While single-echo (SE) MRI sequences that measure the blood-oxygenation level dependent (BOLD) signal are the norm in fMRI research, advances in MR technology and denoising are prompting researchers to turn to multi-echo (ME) sequences, offering better differentiation between BOLD signal and non-neural noise for improved estimates of brain activation.1-3 Despite their benefits, ME sequences limit temporal resolution, as adding echoes extends repetition times (TRs), and introduce added artifacts to concurrent physiological data. The former can be mitigated by using multi-band (MB) excitation, reducing the time required to acquire a single volume by acquiring several slices simultaneously. The latter complicates the use of physiological data in fMRI studies using ME sequences, which we address here by comparing strategies for MR-artifact removal in physiological data across SE and ME BOLD fMRI scans.
Physiological data were collected at 2000Hz using MRI-compatible equipment from BIOPAC Systems. Data were acquired using a BIOPAC MP150 system with MRI-RFIF filters: electrocardiogram (ECG) data with a 3-lead bipolar configuration and an ECG100C-MRI amplifier; electrodermal activity (EDA) recordings, from the nondominant hand, with an EDA100C-MRI amplifier.
Physiological data were acquired in the bore of a 3-Tesla Siemens MAGNETOM Prisma during both a MBSE BOLD echo planar imaging (EPI) sequence and a MBME BOLD EPI sequence.
The MBSE sequence is by the Adolescent Brain Cognitive Development (ABCD) Study,4 which acquires 60 slices with TE=30ms, TR=800ms, and MB factor of 6.
The MBME sequence is from the distribution of MB GRE-EPI sequences developed by the Center for Magnetic Resonance Research at the University of Minnesota, and acquires 48 slices in 4 echoes (TE1=11.80ms, TE2=28.04ms, TE3=44.28ms, TE4=60.52ms), TR=1500ms, MB factor of 3.5
Code used to create and apply these filters is available on GitHub (github.com/62442katieb/mbme-physio-denoising).
Following Fourier transforms of ECG and EDA data before and during both sequences, we identified MR-related noise in frequencies corresponding to the TR and slice acquisition. The effects of this noise were greater during ME than SE sequences, which is visually apparent in (1) the difference between recordings collected before and during the two BOLD EPI sequences and (2) their frequency spectra (MBSE-ECG in Fig. 1A-B vs. 1C-D; MBME-ECG in Fig. 1I-J vs. 1K-L and MBSE-ECG in Fig. 2A-B vs. 2C-D; MBME-ECG in Fig. 2I-J vs. 2K-L). By adapting recommendations from industry leaders and prior research, we developed and applied digital infinite impulse-response (IIR) notch filters, greatly reducing the impact of this noise on physiological data.
·Figure 1. Electrocardiogram recordings (and Fourier transforms thereof) and the denoising options
·Figure 2. Electrodermal activity recordings (and Fourier transforms thereof) and the denoising options
MB and ME pulse sequences introduce more complicated artifacts concurrent physiological data than can be addressed with current industry recommendations. Our research finds that these artifacts are predictable and their effects can be greatly mitigated with notch filters centered at their fundamental frequencies and harmonics. By targeting the slice acquisition frequency, adjusted for MB factor, and TR frequency, especially with ME sequences, researchers should be able to remove significant MR-related artifacts from ECG and EDA data collected during fMRI scans using straightforward digital filters. Such recommendations allow researchers to capitalize on the improved SNR afforded by MBSE and MBME BOLD sequences, while including rich physiological data for both cleaning fMRI data and investigating psychophysiological phenomena.
Modeling and Analysis Methods:
Methods Development 1
Physiology, Metabolism and Neurotransmission :
Physiology, Metabolism and Neurotransmission Other 2
1|2Indicates the priority used for review
My abstract is being submitted as a Software Demonstration.
Please indicate below if your study was a "resting state" or "task-activation” study.
Healthy subjects only or patients (note that patient studies may also involve healthy subjects):
Are you Internal Review Board (IRB) certified?
Please note: Failure to have IRB, if applicable will lead to automatic rejection of abstract.
Was any human subjects research approved by the relevant Institutional Review Board or ethics panel?
NOTE: Any human subjects studies without IRB approval will be automatically rejected.
Was any animal research approved by the relevant IACUC or other animal research panel?
NOTE: Any animal studies without IACUC approval will be automatically rejected.
Please indicate which methods were used in your research:
Other, Please specify
For human MRI, what field strength scanner do you use?
Provide references using author date format
Boyacioğlu R, Schulz J, Koopmans PJ, Barth M, Norris DG. (2015) Improved sensitivity and specificity for resting state and task fMRI with multiband multi-echo EPI compared to multi-echo EPI at 7 T. NeuroImage. 2015;119:352-361. doi:10.1016/j.neuroimage.2015.06.089
2. Kundu P, Voon V, Balchandani P, Lombardo MV, Poser BA, Bandettini PA. (2017) Multi-echo fMRI: A review of applications in fMRI denoising and analysis of BOLD signals. NeuroImage. 2017;154:59-80. doi:10.1016/j.neuroimage.2017.03.033
3. Olafsson V, Kundu P, Wong EC, Bandettini PA, Liu TT. (2015) Enhanced Identification of BOLD-like Components with Multi-Echo Simultaneous Multi-Slice (MESMS) fMRI and Multi-Echo ICA. NeuroImage. 2015;112:43-51. doi:10.1016/j.neuroimage.2015.02.052
4. Casey BJ, Cannonier T, Conley MI, et al. (2018) The Adolescent Brain Cognitive Development (ABCD) study: Imaging acquisition across 21 sites. Dev Cogn Neurosci. 2018;32:43-54. doi:10.1016/J.DCN.2018.03.001
5. Salo T, Bottenhorn KL, Riedel MC, et al. (2020) Dense Investigation of Variability of Affect (DIVA). OpenNeuro [Dataset]. doi:10.18112/openneuro.ds002278.v1.0.1