inormalize

intensity normalize a MINC volume following a registered model volume

inormalize [options] <infile> [<outfile>]

DESCRIPTION

inormalize normalizes the intensity of a MINC volume, optionally relative to a model volume. It supports several normalization methods including ratio of means, ratio of medians, mean of ratios, mean of log-ratios, and median of ratios. The normalization can be performed along specific axes (z/y/x-normalize) to correct for slice-to-slice intensity variation. A mask can be used to restrict the normalization to a specific region.

OPTIONS

-clobber
Overwrite an existing output file.
-noclobber
Do not overwrite an existing output file (default).
-verbose
Print progress information.
-quiet
Do not print progress information.
-print
Print the normalization parameters without writing an output file.
-compress
Compress the output file.
-nocompress
Do not compress the output file.
-nocache
Do not use volume caching.
-const <val>
Normalize to a constant value.
-const2 <val1> <val2>
Normalize using two constant values.
-range <pct>
Set the percentage of voxels to use for range estimation.
-model <model.mnc>
Specify a model volume for relative normalization.
-znormalize
Normalize along the z-axis (transverse slices).
-ynormalize
Normalize along the y-axis (coronal slices).
-xnormalize
Normalize along the x-axis (sagittal slices).
-minvoxels <n>
Set the minimum number of voxels required per slice (default: 1000).
-constrained
Use constrained normalization.
-rms
Use root mean square normalization method.
-vr
Use variance ratio normalization method.
-ratioOfMeans
Use ratio of means normalization method.
-ratioOfMedians
Use ratio of medians normalization method.
-meanOfRatios
Use mean of ratios normalization method.
-meanOfLogRatios
Use mean of log-ratios normalization method.
-medianOfRatios
Use median of ratios normalization method (default).
-mask
Specify a mask volume for restricting normalization.
-useVoxelCoord
Use voxel coordinates for mask evaluation (default).
-useWorldCoord
Use world coordinates for mask evaluation.
-threshold <min> <max>
Set intensity thresholds for voxel selection.
-nonormalize
Do not apply normalization; report parameters only.

EXAMPLES

inormalize input.mnc output_normalized.mnc

inormalize -model model.mnc -znormalize input.mnc normalized.mnc

inormalize -ratioOfMeans -mask brain_mask.mnc -model model.mnc input.mnc normalized.mnc

AUTHOR

Andrew Janke - McConnell Brain Imaging Centre, Montreal Neurological Institute, McGill University.

COPYRIGHTS

Copyright &copy; 2000 by Andrew Janke

SEE ALSO

nu_correct, classify, mincmath, volume_stats