| This thesis describes a body of work undertaken to improve the power of surface-based analysis for anatomical MRI of the human brain. Chapters one and two introduce the topic and present a thorough background to the research. Chapters three through six describe the four main contributions of the thesis. The first contribution was to increase the accuracy of mesh-based representations of the cortical grey/white and pial surfaces. The second contribution was to enhance the quality of between-subject cortical alignment using iterative strategies to build an unbiased group alignment template. The third was to develop a new technique for whole brain analysis of cortical position and surface area. The final contribution was to use multivariate statistical techniques to identify and parcellate sources of artefactual and anatomical variance. Taken together, these contributions constitute a significant improvement to the power of surface-based analysis of inter-subject differences in anatomical MRI. |
