3D Canny Edge Detection with Gaussian Smoothing and Thresholding

This Python script applies 3D Canny edge detection to a NIfTI image with optional Gaussian smoothing, thresholding, and the ability to combine edges from all axes. It also supports auto-completion for command-line arguments.

Requirements

• Python 3
• nibabel
• scikit-image
• argparse
• argcomplete

You can install the required packages using pip:

pip install nibabel scikit-image argparse argcomplete

Usage

Command-Line Arguments

• input_filepath: Path to the input NIfTI file.
• output_filepath: Path to save the output NIfTI file.
• --sigma: Sigma value for Canny edge detection (default: 1.0).
• --gaussian_sigma: Sigma value for Gaussian filter (default: 1.0).
• --threshold: Threshold value to remove data below this value (default: 0.0).
• --view: Open the OrthoSlicer3D viewer to visualize the result.

Example Command

python mri_canny_filter.py path/to/your/input_image.nii.gz path/to/your/output_image.nii.gz --sigma 2.0 --gaussian_sigma 1.5 --threshold 0.1 --view

Enabling Auto-Completion

To enable tab completion for the script's arguments, follow these steps:

1. Install argcomplete if you haven't already:

   pip install argcomplete

2. Add the following line to your .bashrc or .bash_profile:

   eval "$(register-python-argcomplete mri_canny_filter.py)"

   Replace script_name.py with the actual name of your script.

3. Source your shell configuration file to apply the changes:

   source ~/.bashrc
   # or
   source ~/.bash_profile

Alternatively, you can enable argcomplete for the current shell session with:

activate-global-python-argcomplete --dest=- | source /dev/stdin

Script Explanation

1. Thresholding: The script applies a threshold to the input data, setting all values below the threshold to zero.
2. Gaussian Smoothing: A Gaussian filter is applied to smooth the data and reduce noise.
3. Canny Edge Detection: Canny edge detection is performed along each axis (sagittal, coronal, axial) of the 3D image.
4. Combining Edges: The edges detected along each axis are combined using np.maximum to form a single edge-detected volume.
5. Saving the Result: The combined edges are saved as a new NIfTI image.
6. OrthoSlicer3D Viewer: If the --view flag is set, the OrthoSlicer3D viewer is opened to visualize the result.