Cellprofiler imge processing software#

Individual grid tiles from a whole-slide image may be analyzed with a standard CellProfiler pipeline. Image credit: Meryem Aloulou, Liston Lab, Babraham Institute, Cambridge, UK. Also shown outlined in yellow is an enlargement of tile 22. Delineated in cyan are the TZ and BZ that together form the WP. For cell type classification, anti-CD3 (white) and anti-B220 (red) were included to identify T cells and B cells respectively. The section was labelled with a DAPI nuclear counter stain (blue). It should however also be possible to automatically segment them in the WhISCR pipeline by identifying areas of densely clustered B and T cells.įig.1: Whole-slide immunolabelled tissue section from mouse spleen constructed from 40 image tiles. In this example WP, BZ and TZ areas have been manually segmented on the image (segmentations delineated in cyan on fig. In mouse spleen these regions are often juxtaposed, and together form the white pulp (WP). B cells cluster together in the B cell zone (BZ), T cells in the T cell zone (TZ). In figure 1, the typical pattern that B and T cells organize themselves around the spleen can clearly be seen: They are scattered randomly across some regions of the spleen, the red pulp (RP) areas, but densely cluster in other subregions. Together this panel allows for the identification of B cells, T cells, and cells that are neither of these cell types. On multiplex imaging systems however, such as the MACSima imaging platform, cyclic labelling and imaging means that theoretically the number of markers would be unlimited, in reality limited only by the availability of validated antibody panels.įor simplicity, the whole-slide image used to describe the WhISCR pipeline, and shown in figure 1, contains only three fluorescent labels: a pan nuclear stain (DAPI in blue), a T cell specific marker (CD3 in white) and a B cell specific marker (B220 in red). The number of markers that can be assessed within a tissue section will depend on the experimental setup: using simultaneously labelled markers, on a conventional fluorescence imaging system the size of the panel would typically be limited to a maximum of 7 or 8 markers, mainly dependent on the available number of excitation lines. what cell type they are, as well as the identity of their close neighbours. Sections are labelled for a range of cell type-specific markers, which then allows for the extraction of information on the location of individual cells within the tissue, their identity, i.e. Whole-slide immunofluorescence images are large grids of overlapping image tiles that stitched together capture a whole tissue section. I have therefore used it as the focus of this article. As the number of reports that describe the use of CellProfiler for this type of application is limited, I thought the WhISCR pipeline might be of interest to a wider audience. The resulting pipeline, WhISCR (Whole-slide Immunofluorescence Scripts in CellProfiler and R), has been deposited on GitHub. More recently however, I was asked to analyze a number of whole-slide immunofluorescence imaging experiments, and I decided to use CellProfiler to do this. Projects analyzed to date range in size from a few wells on a 96‑well plate, up to tens of 384-well plates. Personally, when I first started using CellProfiler in 2015, it was exactly for this purpose: I had joined the Babraham Institute’s Imaging Facility to take responsibility for its InCell6000 High Content Imaging platform, and started using CellProfiler for analysis of the steady stream of image data produced by this system. Since its launch in 2006, CellProfiler’s popularity across the scientific community has steadily increased, currently gathering around 2000 citations per year.Ī common application of CellProfiler is for analysis of image-based screens, such as genome-wide RNAi and CRISPR screens, as well as morphological profiling by cell painting assays.

Cellprofiler imge processing software#

CellProfiler is an open-source image analysis software package that can be used to extract high content data from microscopy images.