MCount: Automatic Bacteria Counter for Large Biology Experiments

Jenn Hoskins
20th March, 2025

MCount: Automatic Bacteria Counter for Large Biology Experiments

Existing colony counting algorithms such as NICE, AutoCellSeg, and OpenCFU significantly underestimate colony numbers in high-throughput plating formats because they fail to accurately resolve merged Escherichia coli colonies, demonstrating the need for improved methods that incorporate both contour and regional information.

Image adapted from: Chen et al. / CC BY (Source)

Key Findings

  • Researchers at the University of Bayreuth in Germany developed MCount, a new tool to accurately count bacterial colonies in lab images
  • MCount outperforms existing methods by reducing counting errors, especially when colonies overlap or merge
  • Its user-friendly design and high precision enhance efficiency in microbiology labs, benefiting tasks like antibiotic testing
Accurate counting of microbial colonies is essential for evaluating bacterial growth, especially in high-throughput workflows where efficiency and precision are paramount. Traditional methods like manual colony counting are time-consuming and prone to errors, which has driven the development of automated solutions. However, existing automated tools often struggle with accurately counting merged colonies, a common issue in high-throughput plating. Researchers at the University of Bayreuth in Germany have introduced MCount, a novel solution designed to address the limitations of current automated colony counting methods. MCount stands out as the only known tool that effectively integrates both contour information and regional algorithms to accurately determine the number of merged colonies. This dual approach allows MCount to differentiate and count individual colonies even when they overlap, a scenario where many existing tools fall short. To develop MCount, the team optimized the pairing of contours with regional candidate circles, enabling the software to infer the correct number of merged colonies with high precision. The effectiveness of MCount was tested using a meticulously labeled dataset of Escherichia coli images provided by the University of Bayreuth. This dataset included 960 images containing a total of 15,847 segments. In comparative evaluations, MCount achieved an average error rate of just 3.99%, significantly outperforming other published solutions such as NICE, which had an error rate of 16.54%, AutoCellSeg at 33.54%, and OpenCFU, which struggled with a 50.31% error rate[1]. MCount's superior performance is largely due to its sophisticated algorithm that accurately identifies and counts colonies even in challenging conditions. The tool is also user-friendly, requiring only two hyperparameters to function effectively. This simplicity makes MCount accessible to a wide range of users, from laboratory technicians to researchers conducting high-throughput screenings. Building on previous advancements in automated cell and colony counting, MCount offers substantial improvements. Earlier studies have highlighted the challenges of manual counting and the need for reliable automated methods. For instance, a study from the Massachusetts Institute of Technology (MIT) developed an ImageJ macro called Cell Colony Edge for automated counting, which demonstrated improvements in speed and accuracy over manual methods[2]. Another MIT study introduced the Start Growth Time (SGT) method, which provided a rapid way to quantify live bacterial cells in high-throughput settings by measuring the time required for bacterial cultures to reach a growth threshold[3]. While these methods represented significant progress, they did not fully address the issue of merged colonies in high-throughput environments. MCount builds on these foundational studies by specifically targeting the problem of merged colonies, providing a more accurate and reliable solution. Additionally, to facilitate the deployment of MCount in scenarios where labeled data may be limited, the researchers developed statistical methods for selecting hyperparameters using minimal labeled or even unlabeled data points. These methods ensure that MCount maintains consistently low error rates, enhancing its practicality and adaptability in various laboratory settings. The ability to accurately count merged colonies has important implications for numerous applications, including antibiotic testing, where precise colony counts are necessary to determine the efficacy of antimicrobial agents. By providing a more accurate and efficient counting method, MCount can improve the reliability of results in such assays, leading to better-informed decisions in research and clinical settings. Furthermore, MCount's high accuracy and low error rate make it suitable for longitudinal studies where tracking bacterial growth over time is crucial. Its integration into high-throughput workflows can streamline processes, reduce the potential for human error, and increase overall productivity in microbiological research. In summary, MCount represents a significant advancement in the field of automated colony counting. By effectively addressing the challenge of merged colonies and offering a user-friendly, highly accurate solution, it enhances the capabilities of high-throughput microbial assays. This development not only builds on previous automated counting methods but also sets a new standard for accuracy and efficiency in microbial research.

BiotechBiochem

References

Main Study

1) MCount: An automated colony counting tool for high-throughput microbiology

Published 19th March, 2025

https://doi.org/10.1371/journal.pone.0311242

Related Studies

2) High-Throughput Method for Automated Colony and Cell Counting by Digital Image Analysis Based on Edge Detection.

https://doi.org/10.1371/journal.pone.0148469

3) A method for high throughput determination of viable bacteria cell counts in 96-well plates.

https://doi.org/10.1186/1471-2180-12-259


Related Articles

An unhandled error has occurred. Reload 🗙