Donders Institute for Brain, Cognition and Behaviour
Zoek in de site...

Thesis defense Daniel Sharoh (Donders series 426)

24 January 2020

Promotors: prof. dr. P. Hagoort, prof. dr. David Norris

Co-promotor: dr. K. Weber (MPI)

Advances in layer specific fMRI for the study of language, cognition and directed brain networks

This thesis describes research into laminar speci1c functional magnetic resonance imaging (lfMRI), and details the development and adaptation of analysis tools which use lfMRI measurements to extract previously inaccessible information from the laminar resolution blood oxygen level dependent (BOLD) signal.
This information relates to distinctive functional properties of speci1c depthcompartments as well as to directed connectivity patterns between speci1c depth-compartments spanning the multiple distal brain regions.The mammalian isocortex is comprised of at most six cortical cell layers which have been shown to relate to brain function in systematic ways. It is thought that functional MR contrasts, such as give rise to the BOLD signal, are sensitive to layer speci1c signal differences, in turn encoding a signature trace of the synaptic activity speci1c to these cell layers. This is of interest because brain activity in different layers is thought to arise from distinctive layer-dependent connectivity patterns with distal brain regions. One of the principal motivations for the work in this thesis was to assess whether layerdependent measurements might be used to identify layer-dependent networks during language processing, and demonstrate the directionality of information 2ow through these networks based on these layer-dependent patterns.
This undertaking was not without its share of hurdles. It was necessary to overcome a multitude of challenges related to image registration and validation, and to create the tools needed to perform task-based connectivity analysis at the laminar level. Numerous brain regions – described within – are known to be involved in reading. The exact con1guration of these regions is contentious, however. The description of how linguistic information is represented and processed throughout these regions is controversial, and a complete neurobiological theory of reading hinges on understanding how the visual information encoding the words we read relates to our knowledge of these words. In this thesis, it is speci1cally shown that information related to word meaning is communicated backward to a lower order brain region which is not critical for knowledge of word meaning. This was shown by examining the connectivity patterns uniquely related to different cortical depths and relied wholly on noninvasive methods. This 1nding is evidence for the importance of feedback connections to earlier visual areas during word reading, and also evidence that unique information is contained in the depth-dependent BOLD signal.

As a result of this work, there is now a foundation to perform similar investigations in the future, and it is clear that research based on laminar-speci1c connectivity can be informative as to the nature of brain networks.