DTI tractography based parcellation of white matter: Application to the mid-sagittal morphology of corpus callosum

Abstract

Morphology of the corpus callosum (CC) at the mid-sagittal level has been a target of extensive studies. However, the lack of internal structures and its polymorphism make it a challenging task to quantitatively analyze shape differences among subjects. In this paper, diffusion tensor Imaging (DTI) and tract tracing technique were applied to incorporate cortical connectivity information to the morphological study. The CC was parcellated into six major subdivisions based on trajectories to different cortical areas. This subdivision was performed for eight normal subjects and one stroke patient. The parcellated CCs of the normal subjects were normalized for morphological analysis. When comparing the stroke patient to the normal population, we detected significant atrophy in the motor and sensory areas of the patient CC, in line with the clinical deficits. This approach provides a new tool to investigate callosal morphology and functional relationships.

Introduction

Corpus callosum (CC), which interconnects the two cerebral hemispheres, contains more than 300 million axons and is by far the largest fiber bundle in the human brain. Most of these fibers interconnect homologous cortical areas in roughly mirror-image sites, but a substantial number have heterotopic connections, ending in asymmetrical areas (Nolte, 1999). The CC plays an integral role in relaying sensory, motor and cognitive information between homologous regions in the two cerebral hemispheres. Using MRI, the CC can be discretely identified at the mid-sagittal level where it crosses the midline and its morphology has been extensively studied. Such studies have shown that the shape of the CC may be related to gender (DeLacoste-Utamsing and Holloway, 1982), handedness (Witelson, 1985, Witelson, 1989), Down's syndrome (Wang et al., 1992), Alzheimer's disease (Thompson et al., 2003), Tourette's syndrome (Yazgan and Kinsbourne, 2003), dysphasia (Duara et al., 1991, Hynd et al., 1995), schizophrenia (Bookstein, 2003, Narr et al., 2000) and dyslexia (Zaidel et al., 1998). However, the establishment of such a structural-functional correlation is limited by the fact that there are no gross anatomical landmarks that clearly delimit anatomically and functionally distinct callosal regions (Jancke et al., 1997, Sowell et al., 2001). In some approaches, the mid-sagittal CC has been subdivided into vertical partitions (Duara et al., 1991, Larsen and Leonhardt, 1992), while others use equal angular sectors relative to the callosal centroid (Rajapakse et al., 1996). In all of these techniques, the landmarks for the vertical partition ratio or the number of angular sectors needed to be set without the availability of prior intra-CC anatomical information.

The purpose of this paper is to investigate whether incorporation of diffusion tensor imaging (DTI) and DTI-based 3D tract information can facilitate the study of anatomical parcellation of intra-CC structures. The CC is parcellated into six sub-regions based on trajectories into subcortical nuclei and the orbital, frontal, parietal, occipital and temporal lobes. The so-called two ROI and brute-force tractography method was used, in which tracking was initiated from all pixels within the brain and all trajectories that penetrate ROIs were retrieved (Conturo et al., 1999, Huang et al., 2004, Mori and Van Zijl, 2002, Stieltjes et al., 2001). In our protocol, the first ROI is always at mid-sagittal level of CC and the second ROIs are defined in 2D planes characterizing a specific cortical area. After parcellation, the results of healthy volunteers are grouped and statistical maps of occurrence of the different cortical areas are created. For cross-subject registration, large deformation diffeomorphic metric mapping (LDDMM) (Miller et al., 2002) is applied to normalize the callosal shapes across the subjects. Data from a stroke patient, who lost somatic motor functions due to a pontine stroke, are used as a pathological case to test a hypothesis that this approach can detect CC atrophy in the area corresponding to the motor cortex.