Technical NoteCombined T2*-weighted measurements of the human brain and cervical spinal cord with a dynamic shim update☆
Graphical abstract
Introduction
Numerous functions of the central nervous system such as processing of sensory signals from the periphery and from the viscera, motor execution, reflexes or autonomic functions, involve the spinal cord. To investigate its functional role in vivo, functional magnetic resonance imaging (fMRI) experiments based on the blood-oxygenation-level-dependent (BOLD) contrast (Ogawa et al., 1990, Ogawa et al., 1993) represent a valuable non-invasive approach (e.g. Brooks et al., 2012, Eippert et al., 2009, Maieron et al., 2007, Sprenger et al., 2012, Stroman et al., 1999, Summers et al., 2010a, Yoshizawa et al., 1996; for reviews see e.g. Giove et al., 2004, Stroman, 2005, Summers et al., 2010b). fMRI of the spinal cord is technically more challenging than in the brain due to the small size of the gray matter structure and the inhomogeneity of the magnetic flux density in the spinal canal, which is caused by differences of the magnetic susceptibilities, in particular of the vertebrae and the vertebral disks (e.g., Cooke et al., 2004, Maieron et al., 2007). Nevertheless, fMRI of the human spinal cord has provided important insights into the human spinal cord physiology (Eippert et al., 2009, Mainero et al., 2007, Sprenger et al., 2012). For instance, it could be shown that the modulation of pain perception by higher cognitive functions, e.g. attention or in the context of placebo analgesia, can already be detected at the spinal level (Eippert et al., 2009, Sprenger et al., 2012).
With the recent progress in spinal cord fMRI, interest has risen to also target its functional interaction with the brain. Studies investigating functional connectivity between the brain and the spinal cord systems would be informative with regard to many pertinent research questions such as top-down and bottom-up modulation of neuronal activity during physiological signaling in health and its alteration in pathological conditions including its interference with therapeutic approaches.
Performing separate fMRI experiments of the brain and the spinal cord successively is straightforward because individual fMRI protocols are well established for each region. However, such measurements are not suitable for a reliable estimation of the functional or effective connectivity between the brain and the spinal cord which requires to sample the response to each, individual stimulus in all regions considered (Friston et al., 1997, Friston et al., 2003), i.e. the brain and the spinal cord must be covered within the same measurement.
Such combined fMRI measurements are challenging because the desired setups differ considerably between brain and spinal cord measurements. First, typical acquisition parameters like resolution, slice thickness, and field-of-view, differ by a factor of two or more. For instance, most brain fMRI acquisitions employ isotropic voxel sizes between 2.0 × 2.0 × 2.0 mm3 and 4.0 × 4.0 × 4.0 mm3, protocols for the spinal cord usually have an in-plane resolution of 1.0 × 1.0 mm2 or even below and a slice thickness of 4 mm or more (see, e.g., Giove et al., 2004). However, standard MRI sequences can only handle a single parameter set for all slices yielding a non-optimal or insufficient image quality in at least one of the regions, which could hamper the reliable detection or localization of activations. Second, different receive coils are relevant for the brain and the spinal cord. Acquiring data for the brain and the spinal cord with all available coil elements rather than those optimal for each part, leads to a reduced signal-to-noise ratio. Third, and most importantly, the optimum shim adjustment usually differs considerably between the brain and the spinal cord. BOLD fMRI experiments are based on T2*-weighted acquisitions that in the presence of a non-optimal shim setup suffer from signal dropouts. Furthermore, echo-planar imaging (Mansfield, 1977) which is commonly used in T2* weighted fMRI, can exhibit geometric distortions that can degrade the image quality significantly.
In this work, we provide a practical approach for combined T2*-weighted measurements of the human brain and the cervical spinal cord with an accordingly extended echo-planar imaging pulse sequence. First, specific parameters (e.g., in-plane resolution, slice thickness, field-of-view) with an individually optimized timing (i.e., receiver bandwidth, echo spacing, echo time) were used for the brain and the spinal cord subvolumes. Second, the receive coil elements were dynamically selected such that for each slice only elements with significant signal contributions are included. Third, the frequency and the linear shims were dynamically updated during the measurement (Blamire et al., 1996, Morrell and Spielman, 1997) to use different values for the brain and the spinal cord subvolume. The feasibility of this approach to perform combined T2*-weighted measurements of the human brain and cervical spinal cord is demonstrated in healthy volunteers at 3 T.
Section snippets
Material and methods
Measurements were performed on a 3 T whole-body MR system (TIM Trio, Siemens Healthcare, Erlangen, Germany) using a 12-element head and a 4-element neck coil (both receive-only). The MR system is equipped with a 40-mT-m− 1 gradient coil and five second-order shim coils. Linear shims were realized by current offsets to the gradient coils. For testing purposes, water phantoms were investigated. In vivo data were obtained from healthy volunteers that declared their informed consent prior to the
Results
The dynamic coil selection applied improved the SNR, in particular for the spinal cord. For the slices in the brain, the signal amplitude was slightly reduced (by about 3%) but the noise level was decreased by more than 10% which yielded an overall SNR increase (data no shown). In the spinal cord, the signal amplitude was about 2% lower when using only the neck coil elements but the noise level was reduced by more than 45%, i.e. the SNR gain was about 80% (data no shown). The noise reduction
Discussion
In the present work, we have presented an implementation of combined T2*-weighted measurements of the human brain and cervical spinal cord in vivo. The chosen setup involved voxel sizes and fields-of-view that were adapted to each region. The timing was optimized accordingly to provide a shorter echo time and echo train length for the slices in the brain that were acquired with a lower spatial resolution, which reduced geometric distortions and signal dropouts. To achieve the best
Conclusion
The feasibility of combined T2*-weighted measurements of the human brain and cervical spinal cord has been demonstrated in vivo. Using (i) geometry parameters including voxel size and field-of-view adapted to each region and a correspondingly optimized sequence timing, (ii) a dynamic coil element selection, and (iii) a dynamic update of the frequency and linear shim values, an image quality that can be expected to be sufficient for functional MRI experiments, can be achieved in both regions.
Acknowledgments
We would like to thank all volunteers that participated in this study.
Conflict of interest statement
The authors have no actual or potential conflict of interest including any financial, personal or other relationships with other people or organizations within three (3) years of beginning the work submitted that could inappropriately influence (bias) their work.
References (36)
- et al.
Unified segmentation
Neuroimage
(2005) - et al.
Mapping of neural activity produced by thermal pain in the healthy human spinal cord and brain stem: a functional magnetic resonance imaging study
Magn. Reson. Imaging
(2011) - et al.
BOLD signal responses to controlled hypercapnia in human spinal cord
Neuroimage
(2010) - et al.
Single, slice-specific z-shim gradient pulses improve T2*-weighted imaging of the spinal cord
Neuroimage
(2012) - et al.
Psychophysio-logical and modulatory interactions in neuroimaging
Neuroimage
(1997) - et al.
Dynamic causal modelling
Neuroimage
(2003) - et al.
Issues about the fMRI of the human spinal cord
Magn. Reson. Imaging
(2004) - et al.
Dynamic shim updating on the human brain
J. Magn. Reson.
(2006) - et al.
Mapping the spinal and supraspinal pathways of dynamic mechanical allodynia in the human trigeminal system using cardiac-gated fMRI
Neuroimage
(2007) - et al.
Functional brain mapping by blood oxygenation level-dependent contrast magnetic resonance imaging. A comparison of signal characteristics with a biophysical model
Biophys. J.
(1993)
Attention modulates spinal cord responses to pain
Curr. Biol.
A quantitative comparison of BOLD fMRI responses to noxious and innocuous stimuli in the human spinal cord
Neuroimage
Functional exploration of the human spinal cord during voluntary movement and somatosensory stimulation
Magn. Reson. Imaging
Functional magnetic resonance imaging of motor activation in the human cervical spinal cord
Neuroimage
Dynamic shim updating: a new approach towards optimized whole brain shimming
Magn. Reson. Med.
Stimulus site and modality dependence of functional activity within the human spinal cord
J. Neurosci.
Functional MRI using gradient echo EPI in the presence of large field inhomogeneities
J. Magn. Reson. Imaging
Quantitative proton magnetic resonance spectroscopy of the cervical spinal cord
Magn. Reson. Med.
Cited by (44)
Connectomic mapping of brain-spinal cord neural networks: Future directions in assessing spinal cord injury at rest
2022, Neuroscience ResearchCitation Excerpt :Detection of the SCI-related resting state BOLD responses in the two sites within a single measurement could reveal the functional coupling between the brain and the spinal cord, and enable us to gain deeper insights into the human neurophysiology following traumatic SCI (Sprenger et al., 2015). More recently, combined measurement in the brain and the spinal cord have been developed that allows to unveil their mutual interplay (Finsterbusch et al., 2013; Harita et al., 2019) (Table 2). The thought-provoking discoveries of Vahdat et al. opened new avenues for investigating the underlying neurophysiological mechanisms of the plasticity and the dynamic interaction between the brain and spinal cord in normal control subjects, which could have major implications for understanding the effect of injuries in SCI patients (Vahdat et al., 2015).
Advanced spinal cord MRI in multiple sclerosis: Current techniques and future directions
2022, NeuroImage: ClinicalCitation Excerpt :Other existing methods developed for functional imaging of the brain may be applied in the future, including but not limited to exploring the features of the hemodynamic response function in the cord, fMRI of the WM which could reveal information on the functional properties of affected tracts, and dynamic connectivity approaches (Kinany et al., 2020). The development of simultaneous brain and CSC fMRI protocols (Cohen-Adad et al., 2010; Finsterbusch et al., 2013; Islam et al., 2019) may also enable investigations of cortico-spinal activity at rest and during motor or sensory tasks, which may be of value when studying impairments in sensorimotor function (Vahdat et al., 2020), nociception and neurogenic pain mechanisms (Tinnermann et al., 2021), and functional plasticity across the neuraxis. Finally, a few investigations have reported moderate-to-good within-session reproducibility of ventral and dorsal networks functional connectivity (Barry et al., 2016; Hu et al., 2018), as well as consistency between acquisition methods (Kinany et al., 2022) and across studies, which lends credibility to SC fMRI as a research tool, although more test–retest and longitudinal studies are required.
An optimized MP2RAGE sequence for studying both brain and cervical spinal cord in a single acquisition at 3T
2021, Magnetic Resonance ImagingCitation Excerpt :Concurrently, there has been a growing interest in studying the whole central nervous system (CNS) in recent years. Indeed, investigating both brain and SC holds great potential in further understanding pathophysiological relationships occurring in degenerative or traumatic pathologies [23–27]. However, the large FOV required to capture the whole brain and cSC and the high sub-millimetric resolution needed to investigate small structures of SC, combined with B0 inhomogeneities and various artifacts and potential distortions, turn simultaneous brain and cSC imaging into a challenge.
- ☆
Grant sponsor: European Research Council (ERC-2010-AdG 20100407).