Non-invasive Mapping of Connections Between Human Thalamus and Cortex Using Diffusion Imaging

MRI techniques have become a critical component of perioperative planning for neurosurgical procedures. Functional MRI is widely used to locate areas of activation of motor, sensory, language, and even memory functions prior to resective surgeries in the epilepsy, tumor, and neurovascular arenas, whereas high resolution MRI protocols are a standard component of surgical localization of subcortical gray matter structures during stereotactic movement disorder surgery. However, these techniques have limited ability to trace white matter pathways from the cortex to subcortical gray matter structures. The identification of areas of cortical motor activation relative to a Rolandic lesion is readily accomplished by fMRI, but localizing the displacement of subcortical white matter pathways by such a lesion requires intraoperative white matter fiber stimulation mapping. In addition, while MRI can readily identify subcortical gray matter structures such as the subthalamic nucleus, functional divisions cannot be delineated using currently available anatomic images.Figure: DTI Tracing of connections from the LGN (a, b), MD (c), and VL thalamic nuclei (d).Recent work published in Nature Neuroscience (6:750–757, 2003) from the Centre for Functional Magnetic Resonance Imaging of the Brain at the University of Oxford and the Institute of Neurology at University College London/Queen Square provides the first evidence that corticothalamic connections can be noninvasively mapped in the human brain. Behrens and colleagues developed a new probabilistic tractography algorithm to improve on existing diffusion tensor imaging (DTI) techniques. Individual thalamic voxels were interrogated to determine their probability of connectivity to cortical regions and subregions. Whereas other DTI techniques have been limited in their ability to define white matter pathways as they near the cortex or subcortical gray matter targets, the authors' algorithms allowed them to document corticothalamic connections including nuclear subdivisions within the thalamus (see photo), the organization of thalamocortical pathways within the internal capsule, and even the internal medullary lamina of the thalamus. This new methodology has constraints. It does not determine tract polarity (e.g corticothalamic versus thalamocortical), and is more sensitive to larger pathways. Its reproducibility has not been systematically quantified, and it has not been applied on non-control subjects. Nevertheless, this work raises the possibility that we may soon be able to use DTI to trace cortical-subcortical gray matter connections for preoperative surgical planning of cortical resective surgeries in eloquent areas and for more precise movement disorder surgery stereotactic targeting. Probabilistic tractography also will have widespread research applicability in the investigation of cortical connectivity, organization and reorganization, and processing. GUY M. MCKHANN II, M.D.