Installation

Requirements

dependencies:

  • numpy

  • scipy

  • numba

  • mkl

  • matplotlib

  • scikit-image

  • scikit-learn

  • glob

  • pynrrd

  • pydicom

  • h5py

  • nibabel

  • temp

  • sys

  • os

  • subprocess

Installation

Before proceeding, make sure you have all the requirements listed above.

Fiji can be installed here: https://imagej.net/software/fiji/downloads. BoneJ must be added as a plugin within the Fiji installation.

  1. Launch Fiji

  2. From the menu select Help › Update…

  3. Select manage update sites

  4. Select BoneJ

  5. Close manage update sites

  6. Select Apply changes

Clone the repository

git clone https://github.com/BoneJ_Headless

pip install -r requirements.txt

Install the required python libraries.

Usage

First try BoneJ_Module.py located in Examples to launch individual metrics on a single image. Each example requires an input and output directory to set by the user, along with the voxel size of the image, and the Fiji directory path.

All ROIs are acompanied by an .nrrd. The .nrrd file can be opened in Fiji/ImageJ as well. Any file type can be used as long as they are 3D binary 8 bit files. Files are read as numpy arrays by the plugins. :

BoneJ_Module.py

  • This example allows a user to load an ROI, after defining voxel size of the image, and the location of Fiji installation. Any of the primary microstructure metrics can be loaded and computed for an individual image. Users can also set specified microstructure metrics.

BoneJ_Module_Secondary.py

  • This example allows a user to load an ROI, after defining voxel size of the image, and the location of Fiji installation. Any of the secondary microstructure metrics can be loaded and computed for an individual image. Users can also set specified microstructure metrics.

Anisotropy_Parameter_Convergence.py

  • This example allows a user to load an ROI, after defining voxel size of the image, and the location of Fiji installation. Anisotropy is varied across user’s specified values for the input parameters. Results can be plotted for a stable result.

Ellipsoid_Factor_Parameter_Convergence.py

  • This example allows a user to load an ROI, after defining voxel size of the image, and the location of Fiji installation. Ellipsoid is varied across user’s specified values for the input parameters. Results can be plotted for a stable result, which is defined as greater then 90% filling percentage.

This code is currently in development, use with caution.