Functional Significance of Connectional Structures in Living Human Brain

Abstract

The functional anatomy of the brain is its connectional anatomy, that is, its structure as a network. The human brain is a network of a few hundred sharply defined components—neural organs of gray matter and white matter fiber pathways. Conventional MRI cannot detect these intricate fiber pathways. Diffusion MRI, however, offers the unique opportunity of imaging and mapping human connectional anatomy for the first time, and further, to do so in living human subjects. Here we present our works on application of diffusion spectrum imaging (DSI) to the study of functional significance, namely the relationships between structural integrity and neurological or cognitive function. Our results showed that the asymmetry of arcuate fasciculi was correlated with the functional lateralization of language despite handedness. In patients with brain tumors, the fiber stretch at the internal capsule of the corticospinal tract (CST) was associated with the muscle power weakness. In chronic subcortical stroke, we found that structural integrity at the internal capsule of CST at subacute stage could predict the motor outcome at chronic stage. In patients with obsessive compulsive disorder, the severity of clinical scores was associated with the degree of hyperconnectivity in the left cingulum bundle and left stria terminalis, supporting the neurobiological hypothesis of the deficit in the fronto-striato-thalamic circuit. In alcohol dependence, the Barratt impulsivity scale had an inverse relationship with the disruption of the callosal fibers connecting bilateral prefrontal regions. Having verified the clinical potential of DSI, future works will be focused on a standardized and automatic processing of the whole brain tractography and presentation of normalized GFA tract-specific profiles over a group of brains. This platform could greatly facilitate the process of tractography analysis and establish normative tract profiles in a large population. The normative profiles could serve as a digital resource for neuroscience research and health care guidelines.