In search of a counter you can count on: relative efficacy of human visual and automated colony counting

Evaluation of an Automated System for the Counting of Microbial Colonies

Counting of microbial colonies is a common technique employed in research and diagnostics. To simplify this tedious and time-consuming process, automated systems have been proposed. This study aimed to elucidate the reliability of automated colony counting. We evaluated a commercially available instrument (UVP ColonyDoc-It Imaging Station) in regard to its accuracy and potential time savings. Suspensions of Staphylococcus aureus, Escherichia coli, Pseudomonas aeruginosa, Klebsiella pneumoniae, Enterococcus faecium, and Candida albicans (n = 20 each) were adjusted to achieve growth of approximately 1,000, 100, 10, and 1 colony per plate, respectively, after overnight incubation on different solid media. Compared with manual counting, each plate was automatically counted by the UVP ColonyDoc-It with and without visual adjustment on a computer display. For all bacterial species and concentrations automatically counted without visual correction, an overall mean difference from manual counts of 59.7%, a proportion of isolates with overestimation/underestimation of colony numbers of 29%/45%, respectively, and only a moderate relationship (R2 = 0.77) with the manual counting were shown. Applying visual correction, the overall mean difference from manual counts was 1.8%, the proportion of isolates with overestimation/underestimation of colony numbers amounted to 2%/42%, respectively, and a strong relationship (R2 = 0.99) with the manual counting was observed. The mean time needed for manual counting compared with automated counting without and with visual correction was 70 s, 30 s, and 104 s, respectively, for bacterial colonies through all concentrations tested. Generally, similar performance regarding accuracy and counting time was observed with C. albicans. In conclusion, fully automatic counting showed low accuracy, especially for plates with very high or very low colony numbers. After visual correction of the automatically generated results, the concordance with manual counts was high; however, there was no advantage in reading time. IMPORTANCE Colony counting is a widely utilized technique in the field of microbiology. The accuracy and convenience of automated colony counters are essential for research and diagnostics. However, there is only sparse evidence on performance and usefulness of such instruments. This study examined the current state of reliability and practicality of the automated colony counting with an advanced modern system. For this, we thoroughly evaluated a commercially available instrument in terms of its accuracy and counting time required. Our findings indicate that fully automatic counting resulted in low accuracy, particularly for plates with very high or very low colony numbers. Visual correction of the automated results on a computer screen improved concordance with manual counts, but there was no benefit in counting time.

Performance of four bacterial cell counting apps for smartphones

Quantifying bacterial colony forming units is important in microbiological diagnostics. Recent progress in imaging technology allows automation of this tedious and error-prone task. We compared the accuracy of four smartphone colony counter applications conducting standardized measurements, using a self-built apparatus. One app showed high accuracy at lower colony counts.

Scan4CFU: Low-cost, open-source bacterial colony tracking over large areas and extended incubation times

A hallmark of bacterial populations cultured in vitro is their homogeneity of growth, where the majority of cells display identical growth rate, cell size and content. Recent insights, however, have revealed that even cells growing in exponential growth phase can be heterogeneous with respect to variables typically used to measure cell growth. Bacterial heterogeneity has important implications for how bacteria respond to environmental stresses, such as antibiotics. The phenomenon of antimicrobial persistence, for example, has been linked to a small subpopulation of cells that have entered into a state of dormancy where antibiotics are no longer effective. While methods have been developed for identifying individual non-growing cells in bacterial cultures, there has been less attention paid to how these cells may influence growth in colonies on a solid surface. In response, we have developed a low-cost, open-source platform to perform automated image capture and image analysis of bacterial colony growth on multiple nutrient agar plates simultaneously. The descriptions of the hardware and software are included, along with details about the temperature-controlled growth chamber, high-resolution scanner, and graphical interface to extract and plot the colony lag time and growth kinetics. Experiments were conducted using a wild type strain of Escherichia coli K12 to demonstrate the feasibility and operation of our setup. By automated tracking of bacterial growth kinetics in colonies, the system holds the potential to reveal new insights into understanding the impact of microbial heterogeneity on antibiotic resistance and persistence.

Experimental setup and image processing method for automatic enumeration of bacterial colonies on agar plates

Automated colony counting methods have long been known in Microbiology. Numerous methods for automated image analysis have been described and a wide range of commercial products exists. Known advantages are saving cost by reducing enumeration time, automatic documentation, reproducibility, and operator independence. Still, even today the realization of all advantages of automated image analysis makes it necessary to either invest in an expensive, high performance commercial system, or to acquire expert knowledge in image processing. This is a considerable obstacle for many laboratories, and the reason why manual colony counting is still done frequently. This article describes an easy to apply automatic colony counting system–including suggestions for sample preparation–that can be put into operation with basic knowledge of image processing and low budget.