Thesis defense Michiel Kleinnijenhuis (Donders Series 159)
19 May, 2014
Promotors: Prof.em.dr. D.J. Ruiter, Prof.dr. D.G. Norris
Copromotors: dr. A. van Cappellen van Walsum, dr. M. Barth
Imaging fibres in the brain
In these times where connectionist accounts of brain function are gaining in popularity, there is a need for reliable tools for determining anatomical connectivity in the living human brain. The technique of choice is diffusion MRI, but it is debatable whether this tool is suitable for mapping all but the major pathways. The thesis describes my contribution to the development and validation of tools to map the connections in the human brain.
To honour the giants whose shoulders we stand on, and to provide neuroanatomical background, the thesis starts with a historical essay on connectional neuroanatomy. MRI techniques are introduced, focusing on the two modalities most relevant to the topic: diffusion MRI and susceptibility MRI.
The thesis starts with proposing a novel tractography method: Structure Tensor Informed Fibre Tractography (STIFT). With STIFT, the strengths of diffusion MRI (angular resolution) and susceptibility MRI (spatial resolution) are harnessed in one technique. It provides improved spatial specificity of the resulting tracts. Furthermore, in regions with complex fibre configurations, STIFT is able to distinguish between crossing and kissing fibres. Although the method might not be applicable to all tracts in the brain, STIFT is expected to be a useful addition to the tractographer’s toolkit.
The focus then shifts to the cortex. Cortical diffusion imaging becomes increasingly relevant now that high resolutions can be achieved in vivo, which perhaps allows fibres to be tracked into the cortex. By imaging human tissue samples of the primary visual cortex ex vivo on preclinical MR systems, it was demonstrated that cortical diffusion properties are layer-specific. While infra- and supragranular layers show anisotropic diffusion tensors oriented radially to the cortical sheet, the stria of Gennari has low anisotropy. Additionally, the thesis has shown that cortical layers could be better distinguished with the biophysical model NODDI than with conventional diffusion models. In that investigation, diffusion MRI and histology both suggested that fibre dispersion patterns at the grey-white matter boundary vary over the folding cortical sheet. The gyral fibre configurations were investigated further by high resolution diffusion tensor imaging at 7T in vivo. A characteristic pattern of fibre anatomy of the gyrus was derived, in which we observed variations of tensor anisotropy and radiality with cortical curvature, not only in the white matter, but also within the cortex. This set of experiments has considerable implications for tractography, suggesting that (artefactual) biases towards particular locations on the cortical sheet might exist; that models should be designed to capture a variety of dispersion and crossing patterns for tracking fibres in the gyrus; and that intracortical tractography might one day be feasible.
The neuroanatomical teaching tools that are described in the final part of the thesis were created by combining white matter dissection, plastination and tractography. The plastinated prosections have considerable advantages over formalin-fixed specimens because they are durable, non-toxic and easy to handle. These tools might inspire new generations of students to take up research in connectional neuroanatomy.