Data Acquisition Basics
Saskia Bollmann1,2,3

1Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Boston, MA, United States, 2Department of Radiology, Harvard Medical School, Boston, MA, United States, 3Centre for Advanced Imaging, University of Queensland, Brisbane, Australia

Synopsis

This course gives an introduction to data acquisition for fMRI using echo planar imaging (EPI). Key sequence parameters (voxel size, repetition time, echo time, echo train length, flip angle, parallel imaging, simultaneous multislice) and typical artifacts (ghosting, distortions, signal loss) and their impact on contrast, geometry and speed of EPI time series will be discussed.

Target Audience

This course is designed for basic and clinical researchers who would like to get an elementary introduction to data acquisition for fMRI, as well as current users who would like to learn more about key parameters and common artefacts. A basic understanding of the principles of MRI such as the k-space formalism and relaxation times is assumed.

Outcome/Objectives

The aim of this course is to introduce the most prominent parameters (voxel size, repetition time, echo time, echo train length, flip angle, parallel imaging, simultaneous multislice), which define the properties of an fMRI sequence, and how they impact the spatial and temporal characteristics of an fMRI time series (tSNR, BOLD sensitivity). Additionally, common artefacts (ghosting, distortions, signal loss), their origins and how to mitigate them will be discussed.

Introduction

The fundamental requirements for fMRI sequences using the BOLD effect are sensitivity to susceptibility changes and fast acquisition speed. Echo-planar imaging (EPI) (Mansfield, 1984) fulfils these criteria, and builds the basis for most fMRI experiments.

Echo Planar Imaging (EPI)

Echo planar imaging (EPI) describes a k-space trajectory that samples one Cartesian 2D k-space plane in one shot, i.e. after the RF excitation pulse read- and phase-encoding gradients are continuously switched to acquire all gradient echoes within a few tens of milliseconds. To acquire a whole volume, this process is then repeated for each slice, leading to a repetition time (TR), i.e. volume acquisition time, of a few seconds.

Contrast

The T2*-weighting of EPI lends its high sensitivity to the BOLD effect (Bandettini et al., 1992). Importantly, the echo time (TE), i.e. the time after the RF excitation at which the k-space origin is sampled, should be chosen to maximize this sensitivity, with TE ≈ T2* assuming only the presence thermal noise (Deichmann et al., 2002; Posse et al., 1999).

Geometry

The Cartesian EPI trajectory leads, in the presence of slight system imperfections (e.g. eddy currents), to an aliasing artefact termed an N/2- or Nyquist-ghost. An N/2-ghost is caused by a phase error, and thus, slight misalignment, between the odd and even lines of the k-space trajectory. Ghosting is reduced by measuring the phase error at the beginning of the readout and then including a correction term in the image reconstruction[1].

The choice of voxel size determines the available image signal-to-noise ratio (SNR) and the echo train length (ETL). A smaller voxel size reduces the image SNR (Edelstein et al., 1986; Pohmann et al., 2016). Thereby, the thermal noise in the EPI time series increases, and the sensitivity to the small signal changes in fMRI is reduced. Additionally, the prolonged ETL increases the distortions in areas with B0 inhomogeneities, e.g. at air/tissue interfaces. This is because phase errors accumulate over a longer time, especially in the phase encoding direction, leading to larger voxel displacements (Jezzard and Clare, 1999). Further, the continuing T2* decay during the readout also introduces blurring along the phase encoding direction.

The introduction of parallel imaging allows to shorten the ETL, by undersampling the k-space and only acquiring every Rth line of k-space (with R being the acceleration factor) (Griswold et al., 2002; Pruessmann et al., 1999). Thus, the image distortions can be significantly reduced. However, the image SNR decreases proportional by $$$√R$$$, and additional spatially varying noise amplification occurs due to the ill-conditioning of the image reconstruction problem (Pruessmann et al., 1999). Last, EPI is also susceptible to signal loss induced by B0 inhomogeneities within a voxel causing signal dephasing (i.e. an apparent reduction in T2*). In contrast to distortions, however, the signal is lost and cannot be recovered through postprocessing. A number of parameters, such as TE, slice thickness and slice tilt, can be optimized to reduce the loss in BOLD sensitivity (Weiskopf et al., 2006).

[1] Most commercial vendors implement this ghost-correction online.

Speed

Ultimately, fMRI analyses are time series analyses, and temporal stability of the acquisition are crucial to detect the small changes in the BOLD signal. The temporal SNR (tSNR), i.e. the mean value of voxel intensities over time divided by the standard deviation, gives a first approximation of the temporal stability, and thus, sensitivity, of an EPI protocol. The tSNR is a composite measure and depends not only on the voxel size, but also the TE, participant motion, and thermal and physiological noise processes (Krüger and Glover, 2001).

The successive excitation of each slice at each TR introduces a T1-weighting into the EPI time series. To minimize these effects and maximize the available signal at each excitation, the flip angle is commonly set to the Ernst angle $$$arcos(exp(-T_R/T_1))$$$ (Brown et al., 2014).

Note that the TR also determines the number of samples acquired per unit time, and thereby, the statistical power to detect an effect of interest. The introduction of simultaneous multislice (SMS) imaging to fMRI (Feinberg and Yacoub, 2012; Setsompop et al., 2012) allows the acquisition of multiple slices at once, at the cost of additional image SNR through the worsening of the conditioning of the image reconstruction problem and shortened longitudinal signal recovery. In combination with changes in the noise correlation structure, careful protocol and analysis optimisation is required to fully harness the advantages of SMS imaging (Chen et al., 2019).

Other Data

This overview was limited to the acquisition of MRI data; however, most fMRI experiments utilize additional data sources, such as behavioural response measurements, cardiac and respiratory activity, electrodermal activity, EEG, prospective motion correction, field probes, eye tracking, etc. Also, more advanced k-space readout (spiral EPI, segmented EPI, multi-echo EPI, EVI, 3D EPI) or contrast mechanism (spin echo, GRASE, VASO) are available for specific applications; building upon the fundamental principles of EPI.

Summary

Data acquisition in fMRI utilizes the speed and BOLD sensitivity provided by EPI. However, EPI is also susceptible to a number of artefacts, and a balance needs to be found between voxel size, distortions and image SNR, tSNR and TR, and BOLD sensitivity and signal loss. Thus, when designing an fMRI protocol, the region of interest and the planned analysis need to be considered to acquire data that are most suited to answer the research question.

Acknowledgements

No acknowledgement found.

References

Bandettini, P.A., Wong, E.C., Hinks, R.S., Tikofsky, R.S., Hyde, J.S., 1992. Time course EPI of human brain function during task activation. Magn. Reson. Med. 25, 390–397.

Brown, R.W., Cheng, Y.-C.N., Haacke, E.M., Thompson, M.R., Venkatesan, R., 2014. Chapter 18 - Fast Imaging in the Steady State, in: Magnetic Resonance Imaging. Wiley, pp. 447–510. https://doi.org/10.1002/9781118633953.ch18

Chen, J.E., Polimeni, J.R., Bollmann, S., Glover, G.H., 2019. On the analysis of rapidly sampled fMRI data. NeuroImage 188, 807–820. https://doi.org/10.1016/j.neuroimage.2019.02.008

Deichmann, R., Josephs, O., Hutton, C., Corfield, D.R., Turner, R., 2002. Compensation of Susceptibility-Induced BOLD Sensitivity Losses in Echo-Planar fMRI Imaging. NeuroImage 15, 120–135. https://doi.org/10.1006/nimg.2001.0985

Edelstein, W.A., Glover, G.H., Hardy, C.J., Redington, R.W., 1986. The intrinsic signal-to-noise ratio in NMR imaging. Magn. Reson. Med. 3, 604–618. https://doi.org/10.1002/mrm.1910030413

Feinberg, D.A., Yacoub, E., 2012. The rapid development of high speed, resolution and precision in fMRI. NeuroImage 62, 720–725. https://doi.org/10.1016/j.neuroimage.2012.01.049

Griswold, M.A., Jakob, P.M., Heidemann, R.M., Nittka, M., Jellus, V., Wang, J., Kiefer, B., Haase, A., 2002. Generalized autocalibrating partially parallel acquisitions (GRAPPA). Magn. Reson. Med. 47, 1202–1210. https://doi.org/10.1002/mrm.10171

Jezzard, P., Clare, S., 1999. Sources of distortion in functional MRI data. Hum. Brain Mapp. 8, 80–85. Krüger, G., Glover, G.H., 2001. Physiological noise in oxygenation-sensitive magnetic resonance imaging. Magn. Reson. Med. 46, 631–637.

Mansfield, P., 1984. Real-time echo-planar imaging by NMR. Br. Med. Bull. 40, 187–190.

Pohmann, R., Speck, O., Scheffler, K., 2016. Signal-to-noise ratio and MR tissue parameters in human brain imaging at 3, 7, and 9.4 tesla using current receive coil arrays. Magn. Reson. Med. 75, 801–809. https://doi.org/10.1002/mrm.25677

Posse, S., Wiese, S., Gembris, D., Mathiak, K., Kessler, C., Grosse-Ruyken, M.-L., Elghahwagi, B., Richards, T., Dager, S.R., Kiselev, V.G., 1999. Enhancement of BOLD-contrast sensitivity by single-shot multi-echo functional MR imaging. Magn. Reson. Med. 42, 87–97.

Pruessmann, K.P., Weiger, M., Scheidegger, M.B., Boesiger, P., 1999. SENSE: Sensitivity encoding for fast MRI. Magn. Reson. Med. 42, 952–962. https://doi.org/10.1002/(SICI)1522-2594(199911)42:5<952::AID-MRM16>3.0.CO;2-S

Setsompop, K., Gagoski, B.A., Polimeni, J.R., Witzel, T., Wedeen, V.J., Wald, L.L., 2012. Blipped-controlled aliasing in parallel imaging for simultaneous multislice echo planar imaging with reduced g-factor penalty. Magn. Reson. Med. 67, 1210–1224. https://doi.org/10.1002/mrm.23097

Weiskopf, N., Hutton, C., Josephs, O., Deichmann, R., 2006. Optimal EPI parameters for reduction of susceptibility-induced BOLD sensitivity losses: A whole-brain analysis at 3 T and 1.5 T. NeuroImage 33, 493–504. https://doi.org/10.1016/j.neuroimage.2006.07.029

Proc. Intl. Soc. Mag. Reson. Med. 27 (2019)