How to count colonies with ImageJ (and Fiji)

Getting started with automated colony counting

Colony counting remains a vital tool in microbiology for estimating the concentration of viable microbes in a sample. But despite it being tedious and error prone, most labs still practice manual colony counting. This is largely due to a lack of reliable and cost effective alternatives.

While free online tools exist, they take a little practice to perfect before being able to generate genuinely reproducible results. That can be a major barrier if you need to prove to your supervisor the superiority of your approach, before they’ll let off the hook from all that manual counting.

In this step-by-step tutorial we’ll show you how Image J’s Cell Counter Plugin can help you automate counting microbial colonies from agar plate images simply, and largely for free!

The Cell Counter Plugin from ImageJ

For those new to scientific image analysis, ImageJ is a powerful open-source image processing program and already a staple in many biology labs. Its Cell Counter plugin is a game-changer for colony counting, allowing you to achieve detailed analysis based on user-defined criteria, like colony size and morphology, providing richer insights into microbial populations.

The latest Cell Counter plugin is hosted on Fiji, an extension of ImageJ where you can find a whole host of pre-installed plugins and open-source goodies. It is particularly useful because it allows for both manual colony selection at click of the mouse and automated counting, providing a hands off solution for those with appropriate morphologies willing to spend a little more time optimising with filters.

For the purposes of this tutorial, we’ll be specifically focusing on Image J, but we also encourage you to experiment with some of the other available solutions. OpenCFU, Free Agar Plate Counter, countPHICS and BactLAB are just a few of the options out there, so have a play and see what works best for you.

Some key things to consider

Before you rush off to implement your automated colony counting workflow, here are a few things you need to consider to ensure the integrity of your results.

• Image quality is paramount. Ensure consistent and high-resolution imaging with uniform lighting to facilitate accurate object recognition by the plugin. Your pixels are your data after all and not all cameras handle images with the same degree of scientific integrity. You can read more about the impact of image quality on your science here.

• Optimising plugin parameters is also critical. Experimentation with thresholding, size filters, and circularity settings is necessary to accurately identify colonies while excluding artifacts. Over- or under-counting due to poorly defined parameters can negate the benefits of automation.

• Always validate against manual counts. This is especially important when establishing a new automated protocol to ensure the plugin’s accuracy for your specific experimental conditions and colony types. Consider potential challenges such as clustered colonies or debris, which may require manual intervention or advanced image processing techniques within ImageJ, before employing the Cell Counter.

One final word on image quality

In this tutorial, all our images have been taken with the high-resolution imager, ColonyCam. To get the best results and make your life a whole lot easier when counting colonies, ensure your plate is evenly lit and placed on a contrasting background before snapping any photographs. Otherwise you won’t get the uniform results you need for that all important statistical significance.