binarize

Purpose:

Converts 8 or 16 bit image files to 1 bit 'binary' files

where the following definitions apply:

input: the name of the input file
output: the name of the 1 bit 'binary' file to be created
overwrite?(y/n): 'y' grants permission to overwrite output
min: pixels with intensities from this value to max inclusive are converted to 1's; if max is not listed, any pixels with intensities greater than or equal to this value are converted to 1's
max: pixels with intensities from minto this value inclusive are converted to 1's

binarize editedmri bineditedmri y

The file editedmri.img will be converted to a binary file bineditedmri.img. All voxels with non-black values in editedmri.img will be assiged a value of 1 in the binary output

binarize statistical_file significant_file y 16384

The file statistical_file.img will be converted to a binary file significant_file.img. Any value greater than or equal to 16384 will be converted to a 1 in the output, all other values will be 0.

binarize mri_file white_matter_file y 100 170

The file mri_file.img will be converted to a binary file white_matter_file.img. Any value from 100 to 170 inclusive will be converted to a 1 in the output, all other values will be 0.

Binary files represent an efficient way to store editing, regions-of-interest, traced landmarks, etc.
The values of min and max should be specified with respect to the actual values in the input file. You should not attempt to compensate for how AIR will map these values to their internal representation or to the output file values. This program will automatically map the values that you specify to the corresponding internal values.
When AIR is compiled in 8 bit mode, the mapping of min and max depends on the corresponding header maximum, so values greater than a given file's header maximum value will be illegal, even if the same value is otherwise within the acceptable range.
If max is less than min, all output voxels will have a value of zero
If your ultimate goal is just to superimpose the binary file back onto a structural image, overlay_mask may allow you to achieve that purpose more directly.