Immunohistochemistry Image Analysis – an Image Paints a Thousand Words

blog / Immunology, Pathology March 11 2020

In our last article we looked at the workflow for a typical immunohistochemistry (IHC) experiment. Regardless of the tissue type under investigation, the target biomolecules, and the detection method chosen, it is necessary to apply a method to extract and analyse the information held on the stained slides. The detection method used will influence image analysis to some extent i.e. light microscopy or fluorescence microscopy is used to visualise slides after chromogenic or fluorescence detection, respectively. With the right equipment, it is possible to scan entire slides for easier viewing of entire tissue sections, further analysis and to get high-quality images in digital format for use in publications and presentations.

One may also choose to perform confocal microscopy for greater resolution and enhanced multiplexing capabilities. In efforts to screen new drug candidates and develop personalised medicines, samples can be analysed by high content screening with automated image analysis for rapid and high-throughput quantitation of small molecules such as compounds, peptides, or small interfering RNAs, and to evaluate how these may affect cellular phenotypes.

In modern histopathology labs, image analysis is increasingly performed digitally on whole slides and in high throughput to accommodate the need for rapid and reproducible histology-based diagnostics, prognostics and evaluation of therapeutic responses.

Image Analysis – Then and Now

Microscopy has it roots in the 17th century, dating back to when Antonie van Leeuwenhoek (who would later become known as the ‘Father of Microbiology’) developed the world’s first microscope lens from a soda lime glass to study thread quality in his clothing shop in Holland. He later applied his knowledge in glass processing and lens-making to microbiology, marking what would be seen as one of the greatest technological breakthroughs ever in science.

Advances in microscopy throughout the following two centuries improved lens power and function allowing users to see microscopic objects with greater detail, but until about 50 years ago, microscopic analysis was restricted to qualitative visualisation by human eyes. The emergence of video cameras and software tools in and since the 1970s transformed histological analysis from a qualitative to a quantitative discipline where it became possible to view, measure and count cellular components of interest. More recent technological developments have allowed scanning and analysis of whole slides, allowing one to analysis multiple targets simultaneously (multiplex) in a quantitative manner, with fewer concerns about inter and intra-observer reliability, producing high-quality digital images that can be further analysed as well as stored, shared and used in reports and publications.

Is Seeing the Image Not Enough?

During any IHC experiment, tissue samples are stained using antibodies that detect target biomolecules such as: proteins, organelles, cellular components, and others. The staining reactions yield visual signals (chromogenic or fluorescent) that can be viewed using a microscope and ultimately imaged. Simply viewing the stained slides with the naked eye under a light or fluorescence microscope will often provide qualitative data about the target – for instance, is the target protein expressed in the sample or not?, does it appear abnormally (e.g., abundance, location) in a diseased sample?, is the overall gross tissue morphology intact? However, while this approach provides valuable information about the samples, visual analysis on samples is time consuming and prone to incomplete sampling of data.

Quantitative image analysis offers analysis methods not feasible by visual measurements/analysis and produces results comparable between samples with less inter-observer variation.

Image analysis using dedicated computer software that helps to identify regions of interest by colour, size or shape allows the user to extract meaningful data from the stained tissues to address questions such as:

  • Are the expression levels of the target antigen clinically relevant? (i.e. too high or low compared to ‘normal’ or wild-type levels).
  • How does target abundance compare between samples e.g., non-treated vs. treated tissue?
  • What cell types are present within a tissue?
  • Where exactly is the target protein located in the cell?*
  • What proteins or organelles does the target interact with? (this of course requires that you use antibodies capable of detecting suspected interaction partners).*

*requires confocal microscope

So How Are Images Analysed?

During automated digital analysis, a light or fluorescence microscope is connected to a dedicated computer that aids the user in focusing and capturing a good image, as guided by the computer’s image analysis software. Depending on the level of automation and the algorithms available, the software can then make various calculations concerning cells or cellular components, for example:

  • The number of stained cells in a given area. This is usually expressed as no. of cells/cm2. An example could be the number of cells expressing a validated cancer marker in a diseased tissue sample.
  • The ratio of stained: unstained cells for a particular marker.
  • The proportional area of stained components versus unstained components within the sample.

Alternatively, previously scanned slides can be analysed manually or in a semi-automated manner in a dedicated scientific image-processing program.

It is possible to analyse any stained sample (histological or immunohistochemical) in which it is possible to visually separate the region of interest (i.e. the region that is believed to harbour the target) based on strong contrast or colour intensity.

The quality of analyses carried out on immunohistochemically stained slides is dependent upon efficient staining reactions (see previous articles about primary and secondary antibody selection). In general, chromogenic methods are used to detect structural features such as the nucleus, cytoplasm and cell membrane, while detection of smaller components, e.g., sub-cellular compartments, generally calls for fluorescence detection methods. Quantitative information about the target such as expression by cell type or organelle of interest is then expressed as no. of target-positive cells per cell density or given area, allowing the user to easily make comparisons between samples.

IHC Image Analysis with BioSiteHisto

At our GLP-certified laboratory, BioSiteHisto, we offer an imaging service for histological slides using our slide-scanning microscope to yield high-resolution whole slide images (WSI). These scanned images encompass the entire sampled sections in high detail, and particular regions of interest can be chosen during analysis. Image analysis at BioSiteHisto relies primarily on colour-based separation and thresholding to quantify cellular components of interest.

BioSiteHisto performs basic image analysis that focuses on structural differences or expression levels of target antigens. We strive to align our methods with our customer’s exact needs and sample types to provide data is relevant for the experimental question, and we always welcome requests for custom projects. Examples of previous custom projects include histomorphometry of bone to evaluate bone formation/growth, and proliferation indexing of tumour samples to evaluate tumour growth and progression. Based on the results we obtain,  customers can then opt for further analyses at BioSiteHisto or more in-depth analyses elsewhere, or progress directly with their studies.

Digital Images for Convenience and Flexibility

Our image analysis services are largely performed with open source tools such as QuPath and ImageJ, and we supply our customers with digital copies of their stained slides once analysis is complete. This conveniently allows our customers to expand on our analyses later using the same or other tools of their choosing if some other aspect of the samples becomes relevant. An additional major benefit to receiving digital scans of stained slides is the possibility for further visual evaluation of the samples at any time in the future – this is not a given with physical slides that have a tendency to desaturate and lose stain over time.

Get in Touch!

If you have questions about image analysis or any other aspect of IHC, or you want to know more about the services offered by BioSiteHisto please don’t hesitate to contact us for an informal chat by emailing

Related Blog Posts:

Read more about Pathology here.