next up previous contents
Next: NMR-Analysis Tool Up: NMR-Segment Tool Previous: Multi-D Grad Calc   Contents

Image segmentation example

One common use of the segmentation routines is to identify particular tissues to be used as a mask for analysis of other data sets.

  1. construct and load an inital estimate of model parameters using the format gven above.
  2. Calibrate the tissue density model using hfit-scale, E-step, M-mean, M-cov, M-prior. Check the adequacy of the fit in the Imcalc Tool graph Tv via use of the 1D hist plotting routines. Iterate this process until convergence.
  3. Select the required probability map and compute it using prob.
  4. Threshold the resulting probability map in the Imcalc Tool as required.
  5. Manually select connected sub-regions with the mark-$> $connect mouse tool.
  6. Mulitply binary mask with other registered images of the same subject in order to identify data within the required region.

Tissue segmenatation using the rusinek process is similar but requires two hfit operations in order to determine the full set of mean tissue values.


next up previous contents
Next: NMR-Analysis Tool Up: NMR-Segment Tool Previous: Multi-D Grad Calc   Contents
root 2017-11-24