5.1. Log of progress 2020

• Author: Felipe Orihuela-Espina
• Created: January 16, 2020
• Revised: February 18, 2020
• Copyright (c) 2018-20 INAOE

Please note that advances indicated at a particular date, may actually refer to advances in the previous days/weeks.

• Log of progress

5.1.1. Advances 16-Mar-2020

• Version: v0.3
• Responsible: FOE, BGP

Summary of changes:

• Testing of the algorihtm in raster mode. Measurements of pixel distances.
• Improved raster mode for faster execution.
• Testing the code for non-squared matrices.

Bug fixing

• The implementation of the raster mode was incorrectly indexing the output matrix.

5.1.2. Advances 9-Mar-2020

• Version: v0.3
• Responsible: FOE, BGP

Summary of changes:

• Working on the implementation of the structural element, including the raster and anti-raster modes on a given matrix.
• Analysis of the algorithm version that uses the general distance.

5.1.3. Advances 25-Feb-2020

• Version: v0.3
• Responsible: FOE, BGP

Summary of changes:

• Working over structural elements and morphological operations (erosion and dilation, opening and closure).
• Implementation of the endpoint matrix (one of the steps of the segmentation algorithm).
• A bit of testing with the colour palette.
• Working of the gradient weighting

5.1.4. Advances 18-Feb-2020

• Version: v0.3
• Responsible: FOE, BGP

Summary of changes:

• Code analysis (Compilation path directions). Benjamin is looking to an automatic configuration that might not need resetting after every restart.
• Distance calculation under rastering and anti-rastering for supporting minimal path (one of the steps of the segmentation algorithm in Chapter 3 of [1]). The minimal path is found with Dijkstra algorithm:

  • [1] Villanueva Coello, B. (2017). Análisis de imágenes oftalmológicas

    de Tomografía por Coherencia Óptica OCT.

• A bit of manual testing with Dijkstra algorithm.

5.1.5. Advances 21-Jan-2020

• Version: v0.3
• Responsible: FOE, BGP

Summary of changes:

• Commit executed : “Anisotropic diffusion filter” with the following description:

  • Added anisotropic diffusion filter
  • Better integration of the stitch and flattening operations with the segmentation. These operations are now also applied to the segmentation.
  • Class RetinalLayers now encodes fluid as a new layer.
  • Layers colors can be configured in configuration Files.
  • Perfilometer axis now correctly resizes with main scan axis.
  • Detection of the background mask now working properly.

• Bug fixing. Several classes of the data package were having

  issues of cyclic importing octant. In particular, import line:

    import octant.data as octant
  
  was causing error:
  
  .. code-block:: Python
  
     AttributeError: module 'octant' has no attribute 'data'
  
  It has now been updated to:
  
  .. code-block:: Python
  
     from octant import data as octant
  
  The following files were affected:
  
    * `data.Document.py`
    * `data.OCTscanSegmentation.py`
    * `data.OCTvolume.py`
    * `data.OCTvolumeSegmentation.py`
  

Documentation

• The log of progress for 2020 was initiated. Files affected include logOfProgress_2020 and logOfProgress.
• Revision date of logOfProgress_2019 corrected

BGP adaptation

• Installation of Python and its different libraries references to the OCTan project.

• Configuration for Opencv implementation in anaconda for Python 3.7 (https://www.youtube.com/watch?v=vePJ19ZesZk)

• Configuration of the redirection routes of the OCTan project.

• Documentation of segmentation methods for OCT images:

  • Long, J., Shelhamer, E., & Darrell, T. (2015). Fully convolutional networks for semantic segmentation. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 3431-3440).
  • Markus A. Mayer, Joachim Hornegger, Christian Y. Mardin, and Ralf P. Tornow.

  Retinal nerve fiber layer segmentation on fd-oct scans of normal subjects and glaucoma patients. Biomed. Opt. Express, 1(5):1358–1383, Dec 2010.

• Familiarization with the OSF platform and documentation.

• Familiarization with the opScanSegment.py code of the OCTan project.

5.1.6. Advances 13-Jan-2020

• Version: v0.3
• Responsible: FOE

Summary of changes:

• Benjamin Israel Guillén Paz (BGP) joins the project. He will be working in the segmentation algorithm. We have been preparing his project document for admin purposes. Right before Christmas break we completed his registration in OSF and GitHub.