How automation can enhance your colony counting workflows

Introduction

Colony counting is widely used in food and drug safety tests, biomedical tests, and environmental monitoring as a method for accurately determining organism abundance and cell survival. But colony counting can quickly become a tedious and time-consuming task, especially when faced with hundreds or even thousands of colonies per plate.

The challenges of manual colony counting

Manual counting remains the standard approach in many labs, but the process is tedious, lacks consistency and is inadequate for high-throughput experiments (Hallas and Monis, 2015). This is especially apparent when multiple scientists are collecting data, due to heuristic bias and associated subjectivity over what is a colony versus an artefact. Manual counting is also painstakingly slow and even more inaccurate when researchers become fatigued and bored.

Analyses relying on manual colony counting can struggle to achieve statistical significance due to variation in human perception and other inconsistencies driving the need for a larger dataset. This in turn increases variability in the timescale to publication of colony counting results, adding cost and time to the protocol.

Microbiology’s problem with colony counting is longstanding, as stated by Archambault et al. way back in 1937 (Archambault et al., 1937). Although researchers have moved on since the first semi-automated colony counter by the Spencer Lens company from 1943, similar designs are still frequently used today as beginner options. See below for images of Spencer Lens Quebec Colony Counter (left) and a contemporary design (right, image source: Qingdao Youtuo Medical Technology Co. Ltd.). However, these tools do not automate colony measurements, which still require manual work. Furthermore, manual counting itself presents significant challenges. It is a tedious and time-consuming process, prone to human error due to factors such as fatigue, boredom, and eyestrain. This subjectivity and inconsistency in manual counting can lead to unreliable data, hindering reproducibility and accurate comparisons across experiments and laboratories.

Computer-aided analysis

Using open source software tools is often the next step in a lab’s automation journey. The ImageJ Cell Colony Edge macro and the MATLAB based AutoCellSeg have both proved to be more reliable than the commonly used Cell Profiler and OpenCFU. All of which have been tested on a variety of bacterial and human seeded agar plates (Choudhry, 2016; Khan et al., 2018).

Both pieces of software are freely available and easily accessible methods of introducing automation into your colony counting workflows. However, these still are unsuitable for high-throughput experiments, often requiring constant changes to parameters for accurate colony identification. Generic algorithms also struggle greatly with poor quality Petri dish images.

Automation

You could spend years perfecting your home camera and lighting setup to capture high-quality images for editing. However, this can be time-consuming and costly. If you’re not willing to invest significant effort, it might be worth considering specialised automation designed for colony counting.

Automated colony counters are designed to capture high-resolution, colour-accurate images of colonies without any reflections. These instruments often come equipped with specialised software that accurately identifies colonies. The most significant advantage is their ability to repeatedly capture images with consistent lighting and with minimal human intervention, providing quick and precise count and measurement results for your important samples. 

Ready to let the robots do the counting?

Discover why automated colony counting solutions are here to stay.

Fiona Kemm MRes | Scientist

Fiona is a vital member of our Research team, rigorously testing our robots to ensure scientists don’t break them. With no prior robotics experience, she was the ideal guinea pig for our world-class user experience and support. Holding a BSc in Biochemistry and an MRes in Molecular Microbiology, Fiona brings extensive hands-on expertise she applies across departments, supporting both users and internal teams. From writing insightful web articles to specialising in SQWERTY, Fiona ensures our innovations perform flawlessly, helping customers focus on the creative and interpretive aspects of science that can’t be automated.