Effect of Respiration on the B 0 Field in the Human Spinal Cord at 3T Tanya Verma 1,3 and Julien Cohen-Adad 1,2 * Purpose: Respiration affects the bulk magnetic susceptibility, causing frequency shifts as shown in the brain at 7T. Due to the close proximity of the spine to the lungs, this effect is expected to be even larger in the spinal cord, resulting in det- rimental B 0 offset. The goal of this study was to quantify the effect of respiration on B 0 variation in the spinal cord. Methods: Seven healthy subjects were scanned at 3T. Field maps were acquired during inspired and expired conditions. Frequency shift was quantified in the brain, brainstem, and cervico-thoracic spinal cord. A skewed Gaussian function with linear term was fitted to the frequency shift as a function of z- location along the spine. Results: Large frequency shifts were measured along the cord, with a maximum of 74 Hz at C7 (P < 0.05), correspond- ing to 0.58 ppm. The proposed model was adequately fitted to the respiratory-induced frequency-shifts (adjusted R 2 ¼ 0.9954). The morphology of subjects (weight and height) seemed to have an impact on the amplitude of frequency shift, although correlations were not significant. Conclusions: This study provides a deeper understanding of the contribution of respiration to B 0 shift in the spinal cord. The proposed model can be useful for designing future hard- ware or software strategies to compensate for these B 0 variations dynamically. Magn Reson Med 72:1629–1636, 2014. VC 2014 Wiley Periodicals, Inc. Key words: field map; respiration; spinal cord; magnetic sus- ceptibility; shimming INTRODUCTION Susceptibility artifacts are highly prevalent in the spi- nal cord due to different susceptibilities in the verte- brae, the intervertebral disks, and the air-filled lung, giving rise to small-scale and large-scale variations in the B 0 field. These inhomogeneities hamper MRI and spectroscopy of the spinal cord by causing geometric dis- tortions in echo-planar imaging, yielding signal dropouts and echo time shifts in T 2 *-weighted acquisitions, and broadening lines in MR spectroscopy. Consequently, it is essential to perform a shim of the magnetic field to mini- mize such inhomogeneities in the acquisition volume. However, even with a careful adjustment of the standard shim procedure, the image or spectrum quality may be insufficient, because not all inhomogeneities can be com- pensated with only first- and second-order shim coils (1). This notably motivated the development of higher order shim coils for brain regions prone to high B 0 inhomogene- ities (2–4) and slice-wise dynamic shimming for the spi- nal cord (5). Another problem that limits the quality of the shim in the spine region is the variation of B 0 field with time. The respiratory cycle induces B 0 field fluctuations through motion of the chest and therefore fluctuations in local oxygen concentration, which is paramagnetic and therefore distorts the B 0 field dynamically (6). Movement of tissue in the chest and abdomen further adds to varia- tions in the B 0 field. The associated NMR frequency changes then lead to image artifacts such as ghosting and blurring and to temporal signal fluctuations in functional MRI time series (7–9). Although this effect is minor in anatomical imaging of the brain at 1.5T (10) and 3T (9,11), it was shown to become detrimental at 7T (12,13) because the absolute amplitude of the frequency effect scales with field strength. As a result, the frequency vari- ation in a given region is exacerbated, yielding problems during transmission and reception of the RF signal. For transmission, the slice location can be altered, and the excitation profile will be less sharp. For reception, the shortening of T2* can induce blurring (loss of signal dur- ing the course of k-space filling) and important signal dropout at long echo time. Another challenge of higher field strength is that the higher signal-to-noise ratio increases the sensitivity to ghosting and other artifacts. Therefore, while respiration may not appreciably affect quality of brain imaging at 1.5T, it becomes increasingly problematic at higher fields. Moreover, given that the spinal cord is closer to the lungs, we would expect higher sensitivity to respiration effects, even at lower fields. Indeed, respiration was shown to cause ghosting in spinal cord diffusion (14) and functional imaging (15). In spinal cord echo planar imaging (EPI), the typical signature of respiratory-induced variations of B 0 is a time-dependent voxel shift along the phase-encoding direction (typically anterior to posterior), which mani- fests as an apparent oscillation of the cord (16,17). The amount of voxel shift—which can be up to several millimeters—not only varies with time, but also spa- tially along the spine, because it depends on the distance to the lungs, suggesting this effect is indeed caused by dynamic B 0 variations due to respiration. Similar effects 1 Institute of Biomedical Engineering, Polytechnique Montreal, Montreal, Quebec, Canada. 2 Functional Neuroimaging Unit, CRIUGM, Universit e de Montreal, Montreal, Quebec, Canada. 3 Birla Institute of Technology and Science Pilani, Pilani, Rajasthan, India Grant sponsor: SensoriMotor Rehabilitation Research Team of the Cana- dian Institute of Health Research; Grant sponsor: National MS Society; Grant number: FG1892A1/1; Grant sponsor: Fonds de Recherche du Qu ebec-Sant e; Grant sponsor: Quebec BioImaging Network; Grant spon- sor: Natural Sciences and Engineering Research Council of Canada. *Correspondence to: Julien Cohen-Adad, Ph.D., Department of Electrical Engineering, Polytechnique Montreal, 2900 Edouard-Montpetit Bld, room L- 5610, Montreal, QC H3T1J4, Canada. E-mail: jcohen@polymtl.ca Additional Supporting Information may be found in the online version of this article. Received 6 September 2013; revised 1 November 2013; accepted 19 November 2013 DOI 10.1002/mrm.25075 Published online 3 January 2014 in Wiley Online Library (wileyonlinelibrary. com). Magnetic Resonance in Medicine 72:1629–1636 (2014) VC 2014 Wiley Periodicals, Inc. 1629