CT-FIRE for Individual Fiber Extraction(Current Version: V2.0 Beta )

The purpose of CT-FIRE is to provide researchers with a tool to automatically extract individual collagen fibers from images for a quantitative assessment of fiber metrics including fiber angle, fiber length, fiber straightness, and fiber width.  CT-FIRE reads image files supported by MATLAB and extracts individual collagen fibers via a combined method we call “CT-FIRE” (ctFIRE in the software implementation).  The approach of CT-FIRE is described in a publication1, which combines the advantages of the fast discrete curvelet transform2 for denoising images, enhancement of the fiber edge features, and the fiber extraction (FIRE) algorithm3 for extracting individual fibers.  

The output of CT-FIRE may be displayed on the screen and written into .CSV, .XLSX, .MAT data files, or .TIF/.TIFF image files for further analysis, including statistical tests or image classification. The CT-FIRE results can also imported into another LOCI collagen quantification tool named CurveAlign for additional fiber-based feature extraction including bulk assessment of fiber density and alignment. The difference between CT-FIRE and CurveAlign is explained in the “Frequently Asked Questions”.

Key Features include:

  • Collagen fiber analysis is a key emphasis, and is flexible for analysis of any fibrous structure.
  • Analysis of many types of microscopy data including Second Harmonic Generation (SHG), Multiphoton, PolScope and histology stains.
  • Ability to measure individual fiber parameters including angle, length, straightness, and width.
  • Ability to combine and threshold results from large dataset containing e.g. hundreds of images.
  • Fiber properties can be imported to CurveAlign for fiber-based bulk assessment.
  • The region of interest (ROI) analysis module provides localized fiber feature extraction.
  • Free and open source software.

If you find CT-FIRE useful in your work, please reference it with the following citation:

      J. S. Bredfeldt, Y. Liu, C. A. Pehlke, M. W. Conklin, J. M. Szulczewski, D. R. Inman, P. J. Keely, R. D. Nowak, T. R. Mackie, and K. W. Eliceiri, "Computational segmentation of collagen fibers from second-harmonic generation images of breast cancer," J. Biomed. Opt. 19, 016007–016007 (2014).


Standalone Application Package for Windows 64-bit

Standalone Application Package for Mac OS

Standalone Application Package for Linux

Matlab Source Code

Testing images

User Manual of full operating instructions

Documentation of C++ MEX functions for fast fiber extraction

Installation Instructions can be seen here.

Frequently Asked Questions about CT-FIRE can be seen here.

Main Developers:
Yuming Liu (primary contact and lead developer, Feb 2012-)

Adib Keikhosravi (current graduate student developer, Aug 2014-)

Michael Pinkert (current graduate student developer, May 2017-)

Andrew Leicht (current student developer, Aug 2017-)

Guneet Singh Mehta (former graduate student developer, Jun 2014-May 2017)

Jeremy Bredfeldt (former LOCI PhD student, Feb 2012- Jul 2014)

Carolyn Pehlke (former LOCI PhD student, Feb 2012- May 2012)

Prashant Mittal, former undergraduate student from IITJ (India), had contribution on testing and debugging, Aug 2014-May 2015

Release notes:

CT-FIRE V2.0 Beta (November 2016 first test version, December 2017 formal version):

  1. Add parallel computing for the fiber extraction of multiple images or image stack(s).
  2. Add the ROI manager for defining the region of interest in an image.
  3. Add ROI analysis for post-processing CT-FIRE results of individual image.
  4. Add ROI analysis for running CT-FIRE on the defined ROIs of individual image.
  5. Add ROI analysis and ROI post-processing for multiple images.
  6. Add a colormap for visualizing the fiber properties in the advanced output module.
  7. Optimize the graphical user interface to make it more intuitive.
  8. Add the multiple stacks analysis in the advanced output module.

CT-FIRE V1.3 Beta2 (November 2014):

Primary changes were the addition of the advanced output control module including the following functions:

  1. Manually look up the fiber properties and remove the over-segmented fibers.
  2. Automatically select fibers by use of one threshold condition or combined conditions including the absolute or relative limits of fiber width, length, angle and straightness.
  3. Output a variety of statistic measures of the selected fibers obtained from (2) as well as the option to output raw data and overlaid images of them.
  4. Analyze the selected fibers in a single image/stack or in multiple images and visualize the associated statistics.

Other changes included:

  1. Batch-mode fiber extraction for stacks.
  2. Improved width calculation and the option to output the maximum width of each fiber.
  3. Option to estimate the number of bins for the histogram based on the number of fibers.
  4. Control the output image resolution and improve .MAT file loading function.

Fixed bug to load and update the existing fiber extraction parameters; and save the fiber properties of selected fibers into different sheets of one excel file.

Version 1.2.1 beta (December, 2013):

  1. Add batch mode fiber extraction for both multiple image files and .mat files, image stack fiber extraction, post-processing for both single image output, single fiber extraction results ".mat" file and multiple fiber extraction results ".mat" files.
  2.  Add the features of outputting non-overlaid extracted fiber image, fiber straightness, and fiber width besides fiber length and fiber angle.
  3.  In addition, add the button to load parameters in .csv files as well as to update and save the updated parameters for ctFIRE in .csv format.

Version 1.0 (December, 2012)

        Fiber extraction for a single image

More Information

Lastest source code on Github


1.         Bredfeldt, J. S. et al. Computational segmentation of collagen fibers from second-harmonic generation images of breast cancer. J. Biomed. Opt. 19, 016007–016007 (2014).

2.         Candes, E., Demanet, L., Donoho, D. & Ying, L. Fast discrete curvelet transforms. Multiscale Model. Simul. 5, 861–899 (2006).

3.         Stein, A. M., Vader, D. A., Jawerth, L. M., Weitz, D. A. & Sander, L. M. An algorithm for extracting the network geometry of three-dimensional collagen gels. J. Microsc. 232, 463–475 (2008).

4.         Pavone, F. S. & Campagnola, P. J. Second Harmonic Generation Imaging. (Taylor & Francis, 2013).

5.         Bredfeldt, J. S. et al. Automated quantification of aligned collagen for human breast carcinoma prognosis. J. Pathol. Inform. 5, (2014).