Clone wiki

AnaMorf / Home

AnaMorf: Morphological Quantification of Filamentous Microbes

Dave Barry, Image Analyst, Advanced Light Microscopy, Francis Crick Institute, London, UK.

Introduction

AnaMorf is a plug-in developed for the ImageJ platform (rsb.info.nih.gov/ij) to analyse the microscopic morphology of filamentous microbes. The program returns average data on a population of mycelial elements, using the descriptors projected area, circularity, total hyphal length, number of hyphal tips, hyphal growth unit, lacunarity and fractal dimension. The plug-in accepts as input a user-specified directory of images, analysing each and outputing tabulated results.

For the full algorithmic details of AnaMorf, please consult the following publications:

  1. Barry DJ, Chan C, Williams GA. 2009. Morphological quantification of filamentous fungal development using membrane immobilization and automatic image analysis. J Ind Microbiol Biotechnol 36:787–800.
  2. Barry DJ. 2013. Quantifying the branching frequency of virtual filamentous microbes using fractal analysis. Biotechnol Bioeng 110:437–447.

Installation

AnaMorf may be installed by downloading the latest Ana_Morf_v#.###.zip (available through the Downloads icon on the left), unpacking the zip archive, copying all the jar files to the ImageJ plugins folder (.\ImageJ\plugins) and restarting ImageJ. The command AnaMorf > Batch Analyser should then be visible under the Plugins menu. Sample data is also included in the zip archive, together with javadocs (for developers).

User Interface

Upon selecting AnaMorf > Batch Analyser from ImageJ's Plugins menu, the user is presented with an interface, through which various parameters (described below) may be set:

BasicInterface.png AdvancedInterface.png

When the OK button is clicked, a dialog opens for the user to specify the directory in which the images to be analysed are contained. Once a valid directory is provided, the analysis proceeds and results are output to ImageJ's results table.

Basic Settings

Analysis Parameters

  • Image Format: the file extension (BMP, GIF, JPG, PNG, TIF) of the images to be analysed.
  • Image Resolution: the side length of one pixel (in microns/pixel) in the images to be analysed.
  • Minimum Branch Length: the minimum line length (in microns) required to recognize a branch as a 'true' mycelial branch, rather than an artifact of skeletonization, for example.
  • Maximum Circularity: a morphological threshold used to distinguish between mycelial objects and unwanted artifacts. Objects with a circularity greater than this threshold value are not analysed.
  • Minimum Area: the minimum area (in microns) required for an object to be analysed and included in the results.

Output Data

  • Projected Area (Ap): the projected area of an object in square microns.
  • Circularity (C): defined as C = Ap(4 * P^2), where P is the length of the perimeter.
  • Total Length (Lh): the total length (in microns) of a mycelial structure. This length is estimated from skeletonised, binary representations of each mycelium.
  • Number of End-Points (N): the number of hyphal tips on a mycelial structure, estimated as the number of 'end-points' on a skeletonised, binary representation.
  • Number of Branch-Points (Nb): the number of hyphal branches on a mycelial structure, estimated as the number of crossing points on a skeletonised, binary representation.
  • Hyphal Growth Unit (Lhgu): the number of hyphal tips per unit length, Lhgu = Lh / N.
  • Box-Counting Fractal Dimension (Db): a measure of the 'self-similarity' of an object, or the degree to which a mycelium fills space.
  • Fourier Fractal Dimension (Df): An alternative form of the fractal dimension, evaluated based on an analysis of the spectral periodicity of the mycelial boundary, which has been shown to correlate strongly with branching frequency.
  • Lacunarity (L): defined as L = |<sigma>^2 / x^2 - 1.0|, where where x is the average luminance value of pixels within an object boundary (so internal 'holes' are included) and <sigma>^2 is the variance of the luminance values. This represents an estimation of how a mycelium fills space.

Advanced Settings

  • Light/Dark Background: specifies whether the images consist of dark objects against a white background (typical of conventional light microscopy) or light objects against a dark background (typical of fluorescence microscopy).
  • Remove Uneven Background: if ticked, background correction will be performed using ImageJ's implementation of the 'Rolling Ball' algorithm. Generally, background correction is desirable, but can have unexpected effects if large, dense mycelia are being analysed.
  • Maximum Object Size: estimated size of the largest object in the image. This parameter is used to determine the radius of the rolling ball filter for background subtraction.
  • Separate Touching Objects: perform watershed segmentation to break apart touching objects. This can be useful for when analysing images of fungal spores or pellets, for example.
  • Create Mask Images: if ticked, a folder (as a subdirectory within the folder containing the analysed images) will be created containing a binary image of each object analysed, together with a skeletonised version, with branch-points and end-points highlighted:

Input.png Output.png

  • Analyse Whole Image: if selected, AnaMorf will not attempt to isolate individual objects in each image. Instead, the entire field of view will be treated a single object to be analysed.
  • Exclude Edge Objects: exclude from the analysis any object in contact with an edge of the image.
  • Auto/Manual Threshold & Manual Grey Level Threshold: a manual value can be entered for grey-level image segmentation, or segmentation can be performed automatically using one of ImageJ's thresholding algorithms.

Updated