Home
This Title All WIREs
WIREs RSS Feed
How to cite this WIREs title:
WIREs Dev Biol
Impact Factor: 4.711

Quantitating the cell: turning images into numbers with ImageJ

Full article on Wiley Online Library:   HTML PDF

Can't access this content? Tell your librarian.

Modern biological research particularly in the fields of developmental and cell biology has been transformed by the rapid evolution of the light microscope. The light microscope, long a mainstay of the experimental biologist, is now used for a wide array of biological experimental scenarios and sample types. Much of the great developments in advanced biological imaging have been driven by the digital imaging revolution with powerful processors and algorithms. In particular, this combination of advanced imaging and computational analysis has resulted in the drive of the modern biologist to not only visually inspect dynamic phenomena, but to quantify the involved processes. This need to quantitate images has become a major thrust within the bioimaging community and requires extensible and accessible image processing routines with corresponding intuitive software packages. Novel algorithms both made specifically for light microscopy or adapted from other fields, such as astronomy, are available to biologists, but often in a form that is inaccessible for a number of reasons ranging from data input issues, usability and training concerns, and accessibility and output limitations. The biological community has responded to this need by developing open source software packages that are freely available and provide access to image processing routines. One of the most prominent is the open‐source image package ImageJ. In this review, we give an overview of prominent imaging processing approaches in ImageJ that we think are of particular interest for biological imaging and that illustrate the functionality of ImageJ and other open source image analysis software. WIREs Dev Biol 2017, 6:e260. doi: 10.1002/wdev.260

The Trainable Weka Segmentation (TWS) pipeline for pixel classification. Given a sample input image, in this example a maize stem slab acquired using a flat scanner with a resolution of 720 DPI, corresponding to 35.3 µm/pixel (courtesy of David Legland) (a), a user is dependent on the image alone in order to extract features and to properly segment those features; this process can vary greatly depending on the input image (b). Using the power of machine learning, the TWS plugin takes an input image and a set of labels defined by the user to represent feature vectors (c); a WEKA learning scheme is trained on those labels (d) to define and apply a classifier (e) to the remaining image data to properly and automatically segment (f) the image.
[ Normal View | Magnified View ]
An ImageJ plugin for colocalization, Coloc 2. HeLa cells were transfected with HIV‐1 Gag‐CFP (c) and RNA‐tagging protein (MS2‐YFP‐NLS) (d), fixed at 24‐h posttransfection, and imaged with 100× Plan Apo (NA = 1.45). After applying a region of interest (e), the Coloc 2 output (a) can be exported as PDF and is also summarized in the log window (b). Images are courtesy of Jordan Becker from the laboratory of Nathan Sherer at the University of Wisconsin–Madison.
[ Normal View | Magnified View ]
Fluorescence lifetime microscopy (FLIM) images analyzed with spectral lifetime imaging (SLIM) Curve. The figure represents a NADH lifetime of microglia cells activated with lipopolysaccharide (LPS). 740‐nm excitation wavelength was used for NADH excitation, with a 450/70 emission filter. FLIM data were acquired with a Becker and Hickl‐830 board for 60 seconds on a multiphoton microscope. The analysis was performed using the SLIM Curve plugin for ImageJ (b). This image represents FLIM analysis with a 2‐component fit and 5 × 5 binning. The fitted image represents mean lifetime, which is the proportional combination of the free and bound lifetime. The histogram and the color‐coded bar represents the distribution of mean lifetime (a). Exponential decay for a single pixel is shown in distribution (c). The Instrument Response Function (IRF) is also shown (d).
[ Normal View | Magnified View ]
Example of C. elegans lineage performed in TrackMate. H2B‐GFP C. elegans embryos were prepared and imaged on a laser‐scanning confocal microscope. TrackMate was used to segment and track C. elegans nuclei to semi‐automatically generate a full lineage (a, b). The lineage is made of four tracks: the lineage of the AB progenitor, the lineage of the P1 progenitor, and the two basal bodies tracks, followed up to their disappearance (c).
[ Normal View | Magnified View ]
ClearVolume is an open‐source multichannel volume renderer. The plugin offers an intuitive user interface with a number of configuration options, including voxel size and axes parallel cropping, as well as setting lookup tables, brightness, contrast, transparency, and render quality for each individual channel. The visualized dataset in this figure shows two Drosophila neurons that were labeled with twin‐spot mosaic analysis with a repressible cell marker (MARCM) and imaged in the lab of Tzumin Lee at HHMI Janelia Research Campus.
[ Normal View | Magnified View ]
The TrakEM2 plugin assembles 3D volumes and reconstructs, measures, and analyzes structures contained within. In this example, we show a 512 × 512 × 30 px volume at 4 × 4 × 50 nm resolution of a Drosophila larval central nervous system, which was registered using automated methods. All cytoplasmic membranes, synapses and mitochondria are segmented (a) using manual (custom brush tools) and semiautomatic methods (Level Sets ImageJ plugin by Erwin Frise). Each element is editable from the UI (a). The reconstructions are hierarchically organized and can be manipulated as groups (d). The volumes can be rendered (b) and measured (e) using ImageJ's 3D Viewer plugin and results table, which provide further means for exporting the data for further processing elsewhere. In addition to the TrakEM2 interface, individual methods can be combined with other techniques from within the script editor (c).
[ Normal View | Magnified View ]

Browse by Topic

Early Embryonic Development > Development to the Basic Body Plan
Technologies > Analysis of Cell, Tissue and Animal Phenotypes
Adult Stem Cells, Tissue Renewal, and Regeneration > Methods and Principles

Access to this WIREs title is by subscription only.

Recommend to Your
Librarian Now!

The latest WIREs articles in your inbox

Sign Up for Article Alerts

Twitter: molecular Follow us on Twitter

    RT @FEBSnews: FEBS Young Scientists' Forum (YSF): deadline is Friday! FEBS funds YSFers for #FEBS2017 too 😀 Take the opportunity! https://t…
    RT @embojournal: Recruitment and positioning determine the role of XPF‐ERCC1 endonuclease in interstrand crosslink repair @_Hubrecht https:…