CSBDeep, a toolbox for Content-aware Image Restoration (CARE) in Fiji


Deep learning for fluorescence image restoration (denoising, deconvolution). Requires training on your data set but the procedure is described.


Simple Tracing DF-Tracing


We have developed a novel approach, named DF-Tracing, to tackle this challenge. This method first extracts the neurite signal (foreground) from a noisy image by using anisotropic filtering and automated thresholding. Then, DF-Tracing executes a coupled distance-field (DF) algorithm on the extracted foreground neurite signal and reconstructs the neuron morphology automatically. Two distance-transform based “force” fields are used: one for “pressure”, which is the distance transform field of foreground pixels (voxels) to the background, and another for “thrust”, which is the distance transform field of the foreground pixels to an automatically determined seed point. The coupling of these two force fields can“push” a “rolling ball” quickly along the skeleton of a neuron, reconstructing the 3D cell morphology.

Simple Tracing - DT-fields



nctuTW is a "high-throughput computer method of reconstructing the neuronal structure of the fruit fly brain. The design philosophy of the proposed method differs from those of previous methods. We propose first to compute the 2D skeletons of a neuron in each slice of the image stack. The 3D neuronal structure is then constructed from the 2D skeletons. Biologists tend to use confocal microscopes for optimal images in a slice for human visualization; and images in two consecutive slices contain overlapped information. Consequently, a spherical object becomes oval in the image stack; that is, neurons in the image stack do not reflect the true shape of the neuron. This is the main reason we chose not to work directly on the 3D volume.

The proposed method comprises two steps. The first is the image processing step, which involves computing a set of voxels that is a superset of the 3D centerlines of the neuron. The shortest path graph algorithm then computes the centerlines. The proposed method was applied to process more than 16 000 neurons. By using a large amount of reconstructions, this study also demonstrated a result derived from the reconstructed data using the clustering technique." (Extracted from reference publication: https://doi.org/10.1371/journal.pcbi.1002658)

Illustrative image shows gold standard (top) and method results (bottom). 




WormGUIDES Atlas is an interactive 4D portrayal of neural development in C. elegans. It will ultimately contain nuclear positions for every cell in the embryo, identified and tracked from the 2 cell stage until hatching. Single-cell and subcellular information, including neural outgrowth dynamics for each cell as well as cell function, gene expression, the adult neural connectome and related literature will be collated for each cell from public sources and also integrated with the atlas model. WormGUIDES Atlas integrates tools for exploratory data analyses and insight sharing. Navigation is linked between 3D and lineage tree views. In both contexts, community single cell information can be accessed with a click, creating live web queries that summarize knowledge about a cell. In many cases this information can be used to control cell color, creating customized interactive visualizations. A user's insights can be annotated directly into the embryo model with a note-taking interface that attaches each annotation to a cell or other point in space and time. These multi-dimensionally located notes can then be ordered into a (chrono)logical story sequence that explains developmental events as they unfold in the embryo. Annotations can be saved and shared with collaborators or the community.

WormGuides screenshot



GelBandFitter is a user-friendly software specific for analysis of protein gels and estimation of relative protein content. Using non-linear regression methods to fit mathematical functions to densitometry profiles, it is able to estimate content from protein bands that partially overlap. The software is available either as Matlab code (Optimization toolbox required) or a Windows executable. Reference: Mitov, M. I., Greaser, M. L., & Campbell, K. S. (2009). GelBandFitter – A computer program for analysis of closely spaced electrophoretic and immunoblotted bands. Electrophoresis, 30(5), 848–851. http://doi.org/10.1002/elps.200800583

GelBandFitter screenshot



PopulationProfiler – is light-weight cross-platform open-source tool for data analysis in image-based screening experiments. The main idea is to reduce per-cell measurements to per-well distributions, each represented by a histogram. These can be optionally further reduced to sub-type counts based on gating (setting bin ranges) of known control distributions and local adjustments to histogram shape. Such analysis is necessary in a wide variety of applications, e.g. DNA damage assessment using foci intensity distributions, assessment of cell type specific markers, and cell cycle analysis.

PopulationProfiler screenshot



ADAPT is capable of rapid, automated analysis of migration and membrane protrusions, together with associated
fluorescently labeled proteins, across multiple cells. ADAPT can detect and morphologically profile filopodia.

ADAPT (Automated Detection and Analysis of ProTrusions) is a plug-in developed for the ImageJ/Fiji platform to automatically detect and analyse cell migration and morphodynamics. The program provides whole-cell analysis of multiple cells, while also returning data on individual membrane protrusion events. The plug-in accepts as input one or two image stacks and outputs a variety of data. ADAPT may also be run in batch mode.


ADAPT logo



The ImageJFX Project aims to create a new user interface for the software ImageJ in order to ease scientific image analysis. While keeping the core components of ImageJ, ImageJFX brings scientists closer to their goal by making the interface clearer for beginners and more practical for advanced users.

ImageJFX screen Capture


need a thumbnail

Clustered cell nuclei


One of the principal challenges in counting or segmenting cells or cell nuclei is dealing with clustered objects. To help assess algorithms' performance in this regard, synthetic 3D image sets of HL60 cell line are provided consisting of four subsets with increasing degree of clustering. Each subset is also provided in two diferent levels of quality: high SNR and low SNR.Ground truth is available as well. The datasets are part of [Masaryk University Cell Image Collection (MUCIC)](http://cbia.fi.muni.cz/datasets/) as well [Broad Bioimage Benchmark Collection (BBBC)](https://data.broadinstitute.org/bbbc/) - entry BBBC024.

need a thumbnail