Topographical functional connections between the brain and the cervical spinal cord the brain and the cervical spinal cord
Résumé
Introduction:
The sensorimotor system, throughout its neural axis, is characterized by a topographical organization that
reflects body space representation. While the somatotopic organization of the brain (Penfield & Boldrey, 1937),
along with the dermatomal and myotomal spinal cord arrangement (Keegan & Garrett 1948, Schirmer et al.,
2011), are well described, little is known about the spontaneous functional communication existing between
these topographic networks in humans. To address this knowledge gap, this study uses a unique scanning
protocol to acquire functional images of the brain and the whole cervical spinal cord (CSC) simultaneously
(Vahdat et al. 2015, Landelle et al. 2021, Khatibi et al. 2022). To capture brain-spinal cord functional
connectivity (FC) and quantify information transfer, we propose to use a metric derived from information theory
(mutual information, MI).
Methods:
36 right-handed healthy controls underwent an MRI session, consisting in simultaneous brain and cervical
spinal cord acquisition of an anatomical (T1-weighted) and functional images at rest (resolution=1.6x1.6x4mm3,
TR=1550 ms, 260 volumes, 3T Siemens Prisma). The brain and CSC section were preprocessed separately
using an in-house pipeline built with the Spinal Cord Toolbox (SCT), FSL, SPM12 and Nilearn; including motioncorrection,
denoising (motion parameters, RETROICOR, CompCor, discrete cosine transform), band-pass
filtering [0.01-0.17] Hz, normalization to PAM50 (spinal cord) or MNI (brain) templates and spatial smoothing. To
investigate brain-CSC FC, we used pairwise MI; a metric that measures both linear and non-linear relationships
between time series. Specifically, we used the mean time series extracted from different spinal regions of
interest (ROIs) and computed their MI with time series extracted from all brain voxels (spinal seed to whole
brain analysis). First, we investigated the FC between the different quadrants of the CSC (right ventral, left
ventral, right dorsal and left dorsal quadrants) and brain. We then investigated the topographic organization of
the FC between each level of the CSC (seeds at C1, C2, C3, C4, C5, C6) and the sensorimotor cortex. The
individual MI maps were entered in a non-parametric permutation-based t-test (with N permutations = 1000).
T-maps were corrected for multiple comparisons at cluster level (p-FWE < 0.05).
Results:
First, we found significant FC between each spinal quadrant with cortical (M1, S1, SMA, pMC) and subcortical
(putamen, thalamus, cerebellum) sensorimotor regions (Fig 1A-B). Most of the significant MI clusters were
found in the contralateral sensorimotor cortex to the spinal ROIs. These results were confirmed through a
contrast analysis revealing stronger FC between the hemi-parts related to the right upper limb (dominant)
compared to the left upper limb (non-dominant) function (Fig 1C). Second, a topographical organization
emerged from the FC derived from the different levels of the cord – known to innervate distinct body parts – to
the sensorimotor cortex. Interestingly, the FC between these two structures followed the functional subdivisions
of the sensorimotor hierarchy. Notably, there was a rostro-caudal gradient of FC patterns, with different spinal
levels more strongly connected to different parts of the left sensorimotor cortex, in accordance with knowledge
on body space representation (Fig 2).
Conclusions:
Our findings reveal the existence of in vivo functional topographical communication between the CSC and
brain. They show that the brain-spine functional connectivity follows key features of the sensorimotor system,
namely lateralization and somatotopy. They constitute the first demonstration of a somatotopic connectivity
pattern in agreement with homunculi representations. They also suggest that simultaneous brain-CSC fMRI
acquisitions hold great potential to improve the characterization of the sensorimotor integration in vivo at
multiple levels of the human central nervous system.
Domaines
Neurosciences
Origine : Fichiers produits par l'(les) auteur(s)