Automatic Classification of High Angular Resolution Diffusion Data 1
暂无分享,去创建一个
J. D. Carew, G. Wahba, C. G. Koay, Y-C. Wu, A. L. Alexander, M. E. Meyerand Statistics, University of Wisconsin, Madison, WI, United States, Biostatistics and Medical Informatics, University of Wisconsin, Madison, WI, United States, Department of Physics, University of Wisconsin, Madison, WI, United States, Medical Physics, University of Wisconsin, Madison, WI, United States INTRODUCTION High angular resolution diffusion-weighted imaging (HARD) data can provide information about diffusion in voxels that intersect one or more white matter fibers. The diffusion measurements can be interpreted as noisy samples from the fiber orientation distribution function (ODF), which is defined on the surface of a sphere. Since the ODF can be complicated, it is difficult to visualize on a large scale. For many voxels a single tensor model may be adequate, but it is difficult to automatically identify these voxels. It is also difficult to characterize how the HARD measurements differ between populations of subjects (e.g., patients vs. controls). To address these problems we propose a method to automatically classify HARD data based on the shape of the ODF.
[1] Eric R. Ziegel,et al. The Elements of Statistical Learning , 2003, Technometrics.
[2] D L Parker,et al. Comparison of gradient encoding schemes for diffusion‐tensor MRI , 2001, Journal of magnetic resonance imaging : JMRI.
[3] Robert Tibshirani,et al. Estimating the number of clusters in a data set via the gap statistic , 2000 .