Computational neuroanatomy is one of the main research themes at RCCM. This field combines computer science, mathematics and neuroscience to provide insight that will enable us to quantitatively analyze the variability in neuroanatomical structures. Currently we are primarily concerned with MR images and Diffusion Tensor Images(DTI) analysis.

MR images analysis mainly focus on brain tissue segmentation of MR images, intra- and inter-modality image registration, shape and morphometric analysis of the cerebral cortex, and cortical network analysis. For brain tissue segmentation, we have developed multi-context based brain image segmentation and J-divergence based variational segmentation methods. In addition, a parallel genetic algorithm-based active model has been proposed and applied to segment the lateral ventricles. For image registration, both a viscoelastic based model and a non-uniform B-spline based non-rigid registration approach have been proposed and widely applied to the alignment of individual brains to an average template. As to shape and morphometric analysis, we investigate the cerebral cortex at a different levels from global fractal dimension to local thickness. Various metrics have been used to search for significant differences among different populations. Finally, we build cortical networks by connecting anatomical regions of the cortex using links, the strength of which are determined by correlation coefficients of various metrics (e.g. cortical thickness, voxel-based morphometry) between two regions. Then we apply graph theory, as well as small-world network analysis methods, to the networks, investigating how anatomical changes in a few regions affect the performance of the cortex as a whole.

Diffusion tensor imaging (DTI) is a recently developed MRI technique that can non-invasively measure macroscopic axonal organization in nervous system tissues. With this technique, white matter integrity and fiber connectivity can be evaluated in vivo. Two of the major uses of DTI in the central nervous system are in fiber tracing and quantitative white matter analysis. Our group focuses on developing new methods for analyzing diffusion imaging data and clinical applications of DTI. We have developed a scale-invariant parameterization method by arc-angle for quantitative analysis along the cingulum as an analytical method.This technique can yield a continuous and exact description of any segment of the cingulum and can establish the correspondence in the cinculum between subjects. Through this method, a significant left-greater-than-right asymmetry pattern was obtained in most segments of the cingulum bundle in normal right-handed subjects. A similar method for the quantitative analysis of the pyramidal tract has also been proposed. Clinical applications of DTI in brain development and to neuropsychiatric disorders, such as schizophrenia, Alzheimer disease, depression, multiple scleroses, brain tumors and so on, are also an important part of our work. We attempt to investigate normal and pathologic changes in white matter and brain connectivity over the life span, the white matter abnormalities specific to different brain disorders and the correlation between psychiatric symptoms and white matter lesions. A hot issue in our research recently has been the use of DTI to obtain the anatomical network of the cerebral cortex. This has allowed us to characterize the global architecture of the anatomical connection pattern in the human brain.