Brain Deface from T1/T2 image (BRAINS)

Overview

This program: 1) will deface images from a set of images. Inputs must be ACPC aligned, and AC, PC, LE, RE provided.

Panels and their use

Input Images: First Image, Second Image and Mask Image

  • Landmarks File (inputLandmarks): Input Landmark File with LE, and RE points defined in physical locations

  • Source Reference Image (inputVolume): Input images, all images must be in the exact same physical space, ACPC aligned and consistent with landmarks.

  • Source Passive Images Image (passiveVolume): Input images not used in generating masks, all images must be in the exact same physical space as inputVolumes, ACPC aligned and consistent with landmarks.

  • Optional Mask (inputMask): Optional pre-generated mask to use.

Output Files: Outputs from both MUSH generation and brain volume mask creation

  • brain volume mask (outputMask): The brain volume mask generated from the MUSH image

  • OutputDirectory (outputDirectory): The output directory to writing the defaced input files

Run Mode: Modify the program to only generate a mask

  • No Mask Application (noMaskApplication): Do not apply the mask to the input images used to generate the mask

Output Image Intensity Normalization: Parameters for normalizing the output images.

  • Upper Percentile (upperPercentile): Upper Intensity Percentile (0.99 default)

  • Lower Percentile (lowerPercentile): Lower Intensity Percentile (0.01 default)

  • Upper Output Intensity (upperOutputIntensity): Upper Output Intensity

  • Lower Output Intensity (lowerOutputIntensity): Lower Output Intensity

  • Upper Output Intensity (no_clip): Do not clip Values outside of this range to be the “Outside Value”

  • Relative Scaling (no_relative): Do not scale to the relative percentiles of the output scale

  • Debug Level (debugLevel): Level of Debugging (0=None)

Contributors

This tool is created by Hans J. Johnson.

Acknowledgements

This work was developed by the University of Iowa Department of Electrical and Computer Engineering.