DEP. OF EXERCISE > Publications > PhD thesis > 2011 > Henrik Lundell
Methods for mapping spinal cord connectivity using diffusion, functional and conventional MRI
PhD thesis by Henrik Lundell
2011, 131 pages, 100,- d.kr.
ISBN: 978 87 917 7132 3
The aim of this work was to develop MRI strategies for assessing white matter connectivity in the spinal cord. Five papers make up the basis of this thesis.
In the first paper, a morphological method based on standard MRI was introduced. It was shown that regional atrophy following spinal cord injury correlates to sensory and motor function after spinal cord injury. The regions affected corresponded well with the main locations of sensory and motor pathways. Those results suggest that spinal cord morphology can be used to assess connectivity of specific spinal pathways.
From PhD defence March 18 2011.
In the second paper, the cortical activation during foot dorsiflexion was measured with fMRI and related to regional atrophy. Increased activity in M1, PMC and S1 was correlated to atrophy in the lateral parts of the spinal cord. This increase indicates a compensatory mechanism to account for lost connectivity after lesion. Activation was mainly increased in the ipsilateral hemisphere and related to atrophy on the same side suggesting that the crossed corticospinal tract, descending on the contralateral side, can mediate voluntary drive to the ipsilateral foot.
In the third paper, cortical stimulation of the corticospinal tract with transcranial magnetic stimulation and coherent supraspinal synaptic input during gait were compared to regional atrophy following injury. Both related to lateral atrophy but the common synaptic input was related to atrophy in larger and more dorsal parts of the lateral column. Those results suggest that the cortical stimulation is mediated by the more ventral monosynaptic pathways and that the supraspinal drive during gait is more distributed over the whole corticospinal tract and possibly other pathways as well.
In the forth paper, the difficult conditions for diffusion-weighted imaging of the human cervical spinal cord were addressed. Instead of reducing the effect of artifacts in the scanning stage, the point spread function of the voxel intensity was measured and used for correction after acquisition. Datasets with high isotropic resolution and large volume coverage were corrected and enabled consistent diffusion tensor imaging and tractography over the full length off the cervical spinal cord in healthy people. In an individual with spinal cord injury, pathways through a lesion were visualized.
In the fifth and last paper, a high quality diffusion-weighted dataset of the monkey spinal cord was acquired with a high field preclinical magnet. With multi-fiber reconstruction techniques, white matter collaterals were detected as an independent fiber compartment and the distribution in different columns and levels was mapped. An increased number of voxels with detected collaterals was found in the lateral columns around the cervical enlargement. Those regions correspond to areas with increased corticospinal input for the control of hand and arm muscles.
From a clinical perspective, this work provides tools for tracing axonal loss in different spinal pathways and for early detection of lesions. This could have an impact on the planning of rehabilitation and treatment for people with spinal cord lesions in the future.
From a basic research perspective, the methods developed and the initial results obtained show the potential for assessing details of the human spinal cord earlier out of reach with noninvasive methods.
From PhD defence March 18 2011.
Content (pdf, 65 kb)
Summary (pdf, 42 kb)
Preface (pdf, 38 kb)
List of original papers (pdf, 66 kb)


