VZPSTAT

VisionX V4
NAME

vzpstat − computes slice-by-slice image statistics

SYNOPSIS

vzpstat [if=<inputimage>] [im=<inputmask>] [os=<outputslice>] [og=<outputglobal>] [-m] [-s] [-h] [-t] [-w] [-v]

DESCRIPTION

vzpstat computes mean and standard deviation for each slice and globally for the image region specified by the mask. Saves computed statistics to the specified comma-separated values (csv) text file. If mask is not given as the parameter, the statistics are computed from all pixels of the input image. [-m] or [-s] force the output of only mean or standard deviations respectively. [-h] enables the output in Hounsfield units as opposed to default raw pixel units of the image. [-t] adds the row with the column names: "sliceno", "smean", "sstd" for slice number, slice-by-slice mean and standard deviation respectively; "gmean", "gsdev" for global mean and standard deviation. For the slices where the mask is uniformly blank, both "smean" and "sstd" are set to zero. For example, the first few rows of the resultant ’csv’ file for the slice-by-slice statistics may look as follows:

#UU.VISX slice statistics csv
# DICOM
# vzpstat if=img.vx og=global.csv os=slice.csv -t
sliceno,smean,sstd
0,-212.634697,997.761249
1,-209.187424,999.982985
2,-203.990341,1003.829118
3,-196.213089,1009.559676

The global statistics file has a similar structure but, in this case has a single data row containing the following:
gmean: the global mean value
gsdev: the global standard deviation
gnum: the number of pixels evaluated.
smean: the mean of the individial slice means
smsdev: the stadard deviation of the slice means
mssdev: the mean slice standard deviation
snum: the number of slices evaluated.
Use the -w option to obtain statistics in a wish shell script format.

The global standard deviation may be used as an esitmate of noise when homogenious regions are selected. The mean slice standard deviation may be a more suitable noise estimator when there is a significant change of mean value between slices.

OPTIONS

if=

input image file (short)

im=

input region mask (byte), default: entire image is used

os=

output slice-by-slice image statistics (csv)

og=

output global image statistics (csv)

-m

output means only

-s

output standard deviations only

-t

display column titles in first row

-h

convert to Hounsfield units from slope/intercept info

-w

wish output

-v

verbose output

SEE ALSO

VPS(1)

AUTHOR

Sergei Fotin, August 2013