Outlier Rejection for Diffusion Weigthed Imaging

M. Niethammer, S. Bouix, Santiago Aja-Fernandez, C.-F. Westin, M. E. Shenton
Tenth International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI '07)
Volume 4791, Pages 161-168
November, 2007

Download full paper

Abstract

This paper introduces an outlier rejection and signal reconstruction method for high angular resolution diffusion weighted imaging. The approach is based on the thresholding of Laplacian measurements over the sphere of the apparent diffusion coefficient profiles defined for a given set of gradient directions. Exemplary results are presented.

Color by orientation from linear least squares tensor estimation with and without outlier rejection. A signal intensity drop can cause major misestimations of diffusion direction (b). After reconstruction, the artifactual blue coloring due to outliers in (b) is replaced by sensible values in (e) through interpolation of ADC on the sphere. (a)

Reference

Niethammer M, Bouix S, Aja-Fernandez S, Westin CF, Shenton ME. Outlier rejection for diffusion weigthed imaging. In Tenth International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI '07), volume 4791 of Lecture Notes in Computer Science. Brisbane, Australia, 2007;161-168.

Bibtex entry

@InProceedings{niethammerMICCAI07,
  author         = "Marc Niethammer and Sylvain Bouix and Santiago             
                   Aja-Fernandez and Carl-Fredrik Westin and Martha E.         
                   Shenton",                                                   
  title          = "Outlier Rejection for Diffusion Weigthed Imaging",         
  booktitle      = "Tenth International Conference on Medical Image Computing  
                   and Computer-Assisted  Intervention (MICCAI '07)",          
  month          = "November",                                                 
  year           = "2007",                                                     
  volume         = {4791},                                                     
  series         = "Lecture Notes in Computer Science",                        
  pages          = "161--168",                                                 
  address        = "Brisbane, Australia"}

Grants

NIH R01-MH074794