Utah NAMIC EAB 2007
description
Transcript of Utah NAMIC EAB 2007
NA-MICNational Alliance for Medical Image Computing http://na-mic.org
Utah NAMIC EAB 2007
National Alliance for Medical Image Computing – http://na-mic.org
Rician Noise and Tensor Estimation
• Tensor bias from Rician noise
Utah
National Alliance for Medical Image Computing – http://na-mic.org
Tensor Estimation
• ML/MAP tensor estimates• Filtering on DW with
simultaneous estimates of tensors
• Complex vs magnitude averaging/estimation
• Basu et al., MICCAI 2006
Utah
National Alliance for Medical Image Computing – http://na-mic.org
Volumetric Connectivity
• Define paths between ROIs– Analyze circuits (cortex, subcortex)
• Work entirely on the grid• Quantify point-wise evidence for connection
UtahUtah
National Alliance for Medical Image Computing – http://na-mic.org
Define 3D Paths from ROIs
QuickTime™ and aTIFF (LZW) decompressor
are needed to see this picture.
QuickTime™ and aTIFF (LZW) decompressor
are needed to see this picture.
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
FA
Distance along tract
FA
UtahUtah
National Alliance for Medical Image Computing – http://na-mic.org
Particle Shape Correspondence
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
• No explicit parameterization– Wider class of shapes, no topological constraints
• Information theory regularization– Reduced free parameters
• Compares favorably against MDL in 2D– Simple examples
Utah, UNC, Harvard PNL
National Alliance for Medical Image Computing – http://na-mic.org
Particle 3D Correspondence
• Works well in 3D• Adaptive distribution of surface points• Work in progress on caudates
Surface sampling with a max entropy particle system
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
QuickTime™ and aTIFF (Uncompressed) decompressor
are needed to see this picture.
National Alliance for Medical Image Computing – http://na-mic.org
Software
• Segmentation– ITK/NAMIC sandbox of brain tissue classification
based on entropy of nonparametric statistics (Tasdizen et al., MICCAI 2005)
• DTI– MAP smoothing in ITK (Basu et al., MICCAI 2006)– Integration of module into Slicer3
• Shape– ITK implementation of particle system for shape
correspondence
National Alliance for Medical Image Computing – http://na-mic.org
Publications
• DTI– Fletcher PT, Joshi S. Riemannian Geometry for the Statistical Analysis of Diffusion
Tensor Data. Signal Processing, 2006. (To appear).– Corouge I, Fletcher PT, Sarang J, Gouttard S, Gerig G. Fiber Tract-Oriented Statistics
for Quantitative Diffusion Tensor MRI Analysis. MedIA (To appear)– Basu S, Fletcher PT, Whitaker RT. Rician Noise Removal in Diffusion Tensor MRI,
MICCAI'06.– Corouge I, Fletcher PT, Joshi S, Gilmore JH, Gerig G. Fiber Tract-Oriented Statistics
for Quantitative Diffusion Tensor MRI Analysis. MICCAI, 2005• Shape
– Fletcher PT, Whitaker RT. Riemannian Metrics on the Space of Solid Shapes. MICCAI'06 Workshop on Mathematical Foundations of Computational Anatomy (MFCA).
– Cates J, Meyer M, Fletcher PT, Whitaker R. Entropy-Based Particle Systems for Shape Correspondence. MICCAI'06 Workshop on Mathematical Foundations of Computational Anatomy (MFCA).