WHITE MATTER ARCHITECTURE FROM DT-MRI

点击翻译按钮获取中文摘要

This subproject is one of many research subprojects utilizing the resources provided by a Center grant funded by NIH/NCRR. The subproject and investigator (PI) may have received primary funding from another NIH source, and thus could be represented in other CRISP entries. The institution listed is for the Center, which is not necessarily the institution for the investigator. The purpose of the DT-MRI subproject is to develop novel filtering, segmentation, and tractography methods for DT-MRI data. The specific clinical objectives are to detect and localize white matter brain abnormalities in schizophrenia using techniques in DT-MRI and post-processing. To resolve these issues, we are developing new technology for white matter tractography based on simulation of the basic PDE describing anisotropic diffusion and also developing tools for visualization and interactive exploration of 3D DT-MRI data. In addition to supporting the specific needs in this core, these tools will be directly applicable to early detection of white matter damage in the developing infant brain, to exploration and characterization of white matter disruptions caused by MS lesions, and to the visualization of important white matter tracts during neurosurgical procedures (e.g., the cortical spinal tract connecting to the primary and secondary motor cortex). During the past year the DTI core has focused on developing new methods for DTI tractography using stochastic methods. Using a Bayesian approach, we have developed a principled way to describe the inherent uncertainty in fiber tract location by generating a distribution of tracts from a single point, rather than one single tract as is customary. From this work we have derived new measures of connectivity. We have further continued our work on clustering fibers into anatomically meaningful bundles, and have new exciting results on how to generate a high-dimensional atlas for describing the neuroanatomy based on the clustering space. The same technology has also proved useful for group analysis of DTI data, where clustering is used to find correspondences between regions across different brains. Another important line of work is to extend the widely used tensor model of local water diffusion profile. During the past year, we have also continued to develop the two-tensor model. The two-tensor model reduces the uncertainty generated by complex neighborhoods. In general, higher order models are apt to improve both tracking and quantification of diffusion anisotropy.