Application of Brodmann's area templates for ROI selection in white matter tractography studies

Abstract

Brodmann's areas are part of the common vernacular used by neuroscientists to indicate specific location of brain activity in functional brain imaging studies. Here, we have employed a template based on the Brodmann's areas as a means of compartmentalizing underlying white matter pathways. White matter tractography was performed on the diffusion tensor data of sixteen subjects using a streamline tracking technique with Runge–Kutta integration. After co-registration, the Brodmann template was utilized for ROI selection. Tracts were segmented based on their termination in a particular area of the template. Binary masks were generated based on the tractography segmentation for a given Brodmann's area in each individual subject. Following registration to a normalized coordinate space, the binary masks were averaged, generating a map that estimates the probability of tractography connectivity for particular white matter pathways to a specific Brodmann's area. The probability maps were color-coded and overlaid on anatomical images to provide perspective. In this study, particular attention was given to the areas of the frontal cortex. A composite map of these areas was generated by assigning each voxel to the Brodmann's area with the highest probability of connectivity, based on the average results. The average maps generated with this method reveal consistent patterns of connectivity across subjects. The use of a normalized template for ROI selection automates the process of segmenting tractography data, making it particularly useful for multi-subject studies. In the future, this method could be used to help elucidate relationships between function and anatomical structure.

Introduction

The study of brain function can lend insight into the processes underlying neurological diseases and disorders. Beneath the abstraction of human brain function is the physical connectivity that allows for communication and interaction between the various structures and regions of the brain. The determination of the relationships that exist between function and structure will lead to a more comprehensive understanding of the human brain. Of particular interest is the connectivity of cortical and subcortical regions to underlying gray matter structures. Studies using functional magnetic resonance imaging (fMRI) have been successful in imaging and localizing brain activations on the cortex in response to an event. Such studies provide a basis for understanding differences between the function of normal and pathological brains. The activations recorded on the cortex using fMRI are indicative of a particular brain process or function.

In the early 20th century, German physician and scientist Korbinian Brodmann subdivided the human cortex into forty-six areas based on cytoarchitectural features (Brodmann, 1909). Brodmann theorized that the areas were not only anatomically distinct but functionally as well. Brodmann's theory of the localization of brain anatomy and function is now widely accepted. The maps pertaining to Brodmann's parcellation of the human cortex are extensively used as a reference for indicating the specific location of brain activity in functional imaging studies. Scientists who utilize functional imaging to conduct their research are able to communicate their results through the Brodmann nomenclature, which provides a common framework for comparing data between studies.

In this study, a method for estimating the white matter connectivity of each of the Brodmann's areas was developed. The method is aimed at generating maps of the most likely white matter pathways that terminate in or originate from the specified areas of the cortex. An increased understanding of the underlying physical connectivity of the cortex will be valuable for understanding functional deficits caused by brain injury and disease.

The physical connectivity of the cortex, and hence Brodmann's areas, is governed by the organization of white matter fiber bundles that terminate within the specific areas. In white matter, the diffusion of water is anisotropic. Generally, it is assumed that the apparent diffusivity is greater in the direction parallel to axon fiber bundless than in the perpendicular direction. This dependence of water diffusivity on tissue microstructure is exploited by diffusion tensor imaging (DTI) (Basser and Pierpaoli, 1996). In white matter, the magnitude, anisotropy, and orientation of water diffusivity can be characterized using the diffusion tensor. The information obtained from a DT image is representative of the local physical properties of tissue and may be used to estimate the location and orientation of white matter within the human brain. Once the white matter has been distinguished, the continuity of individual white matter fibers within three-dimensional space can be estimated using white matter tractography (WMT) (Conturo et al., 1999, Mori et al., 1999, Basser et al., 2000). WMT is a method that uses the directional information intrinsic in the diffusion tensor to predict the path of major white matter fiber tracts. This may be used to estimate patterns of anatomical white matter connectivity within the brain. The 3D reconstruction of probable trajectories with WMT provides potentially important perspectives regarding the relationships between specific white matter tracts and cortical areas. Recent studies have demonstrated that WMT can provide anatomically plausible estimates of white matter pathways (Stieltjes et al., 2001, Mori et al., 2002, Catani et al., 2002, Jellison et al., 2004).

In this study, a cortical template representing the frontal Brodmann's areas (BAs) was employed to segment WMT data in a group of healthy adults (Fig. 1). Probabilistic maps of white matter pathways were constructed for specific BAs. The method for applying a Brodmann's area template to ROI selection for WMT is described. The approach is relatively simple to implement and adaptable to other brain and cortical templates. The overall objective of this project is to construct a white matter connectivity template that corresponds to specific BAs.