Journal of Microbiological Methods

Measuring the growth rate of filamentous fungi is an essential phenotype assay in fungal biology, enabling the comparison of nutrient-related fitness metrics across various isolates, species and genera. Conventional methods are time consuming and labor intensive, which prohibits the adaptation and implementation of high-throughput phenotyping. Here, we suggest a high-throughput methodological pipeline to study fungal growth on solid media combining the use of 24-well plates, an automated image acquisition system, and human assisted deep learning analysis of acquired images. Training a deep learning model through an iterative process - with continuous feedback and corrective annotations - enabled the development of a satisfying model that automatically segments pixels belonging to either fungus or background within a few hours. We evaluated this deep learning model by applying it to two test sets: First, a set of 336 images was used to validate the results by comparison with manual measurements. We demonstrate that the automated segmentation approach provides robust estimation of fungal growth not significantly different to manually segmented data. Second, a larger test set consisting of 2,016 images was used to illustrate the scalability of the model. After training the model for less than two hours, the deep learning model segmented the entire image data set automatically within minutes. The presented method is easily scalable and adjustable to other fungi and growth morphologies, due to the interactive training. Moreover, by combining 24-well plates and automatic image acquisition, measurements can be sped up as growth is detected across a smaller surface area than a standard six or nine cm diameter petri dish. The proposed methodological pipeline thus offers a new tool for estimating fungal growth rates, which can accelerate measurements, reduce bias, and increase throughput.

This paper reflects on the development and performance of an advanced artificial intelligence (AI) algorithm for the automated processing and classification of Gram stain images obtained from actual microbiology samples used in clinical microbiology. The aim of the project was to effectively categorize non-standardized Gram stain images into the six most common categories: Gram-negative rods, Gram-positive cocci in chains, Gram-positive cocci in clusters, Gram-positive rods, Gram-negative cocci, and yeasts. The development and testing relied on 1,077 Gram stain images of varying sizes, originating from different laboratories and captured using diverse microscopes at different points in time, resulting in differences in image quality, scaling, color balance, and the presence of artifacts. The dataset was split into 80% training and 20% testing subsets, with the split performed in a stratified manner so that each object group was proportionally represented in both the training and testing sets. Preprocessing involved computer vision techniques to improve contrast and color balance, detect contours and object borders, and implement filtering mechanisms to remove unwanted artifacts. Morphological analysis of shapes was then performed to extract parameters characterizing each contour. Next, human-like classification criteria--based on gradient, morphological features (e.g., shape, size) and spatial arrangement that mimic microbiologists visual assessment--were established, achieving around 92% accuracy in image classification without using machine learning (ML) methods. However, any further improvements turned practically impossible, prompting the use of ML methods. Building on pre-obtained features, a random forest ML algorithm was employed to further refine the criteria, with three models trained and tested successively. The first model determined the Gram stain reaction (positive or negative) of each object. The second model classified objects into one of six predefined categories. The third model aggregated individual object classifications to generate an overall classification for each slide, based on the number of objects observed in each category and their occupied area. Overall, the ML solution was significantly more accurate, reaching 99.9% accuracy in classifying the images into one of the aforementioned groups. The algorithms limitations include inability to classify mixed cultures, as it primarily focuses on the dominant category. In cases where positive and negative objects coexist, the algorithm tends to prioritize Gram-positive objects. Additionally, the current morphological assessment is insufficient for yeast classification. Addressing these limitations is a crucial avenue for future research to enhance the algorithms versatility and accuracy.

Morphological switching in response to environmental stimuli is a well-known phenomenon in fungi, leading to diverse morphotypes. Microscopic observation remains a widely used approach to study these phenotypes. However, variation in sample preparation and operators skill can limit the scale of sample processing or introduce bias. Although several image-based cell detection tools have been developed, most are tailored to specific applications or limited to a particular taxon. To address the need for a tool applicable to the polymorphic, yeast-like fungus Aureobasidium pullulans, and with potential applicability to other taxa, we developed TU_MyCo-Vision, an Ultralytics YOLO (You Only Look Once) based object detection tool for identifying 13 fungal morphotypes in bright-field microscopic images. The tool integrates a YOLOv11m-based object detector trained on a custom dataset of 1,504 annotated images and a standalone graphical user interface that enables downstream data analysis and visualization of results. The best-performing model (Zulu_s3) achieved a mean precision of 73.4%, a recall of 66.5%, a mean average precision at 50% IoU (mAP@50) of 73.5%, and a mean average precision at varying IoU thresholds between 50 to 90% IoU (mAP@50-95) of 54.5% across all 13 classes. The single-group analysis pipeline was validated on a 90-image test set, generating six quantitative summaries, including absolute counts, relative and mean relative abundance plots, stacked bar plots, and clustered heatmaps. Multi-group evaluation on previously unseen datasets comprising Candida albicans, Komagataella phaffii, and Aspergillus niger spores demonstrated the tools potential applicability to other genera. TU_MyCo-Vision is distributed as a fully packaged, cross-platform executable, eliminating the need for environment setup or manual installation of dependencies. Built entirely on open-source frameworks, it provides a foundational and potentially extensible solution for automated fungal morphology detection and analysis. Author SummaryWe developed TU_MyCo-Vision to address challenges in fungal microscopic imaging. Fungi, such as Aureobasidium pullulans, display a remarkable ability to switch cell shapes (up to thirteen in this species alone) depending on their environment. While microscopy remains a popular method for observing these changes, manual analysis is limited by individual expertise and the number of images that can be processed, often making results subjective and difficult to scale. To overcome these challenges, we built an Ultralytics YOLOv11-based cell detector that can automatically detect and categorize thirteen fungal cell shapes from brightfield microscopic images. We designed TU_MyCo-Vision to be accessible, with a simple graphical user interface, integrated data analysis suite, and distribution as a standalone application for both Windows and macOS, so it can be used even by those with limited computational skills. Our tool demonstrated strong performance, achieving over 73% precision. Importantly, it also worked well on images from other fungal species, showing potential to be further developed as a general fungal cell morphology tool. We hope TU_MyCo-Vision will contribute to making standardized, high-throughput phenotyping of fungi accessible to a broader community.

Bacterial characterization is a crucial discipline within microbiology. Given the manual and labor-intensive nature of this task, our aim is to introduce a semi-automatic segmentation method that enhances efficiency while preserving the rich details of bacterial colonies. We propose using the k-means clusterization algorithm to analyze and segment images of bacterial cultures, specifically those of Pseudomonas koreensis and Escherichia coli. Unlike existing methods that focus primarily on colony counting, our approach emphasizes morphological characterization. In some bacterial cultures, colonies are not well-defined, making manual counting or other automated counting methods unfeasible; i.e. the bacterial growth area is not easily identifiable, thus precise growth tracking is not feasible. Our method enables bacterial growth characterization even in these cases. Our computer vision system identifies and quantifies the diverse morphologies within P. koreensis and E. coli cultures, determining their relative occupancy in an image. Our approach provides valuable insights into the composition, growth patterns, and developmental stages of bacterial colonies, designed to assist both novice and expert microbiologists in bacterial analysis.

Accurate enumeration of bacteria in a culture is the first step in both fundamental as well as applied research in microbiology. Serial dilution is an age old method used widely by researchers for enumerating viable bacteria in a culture where a specific sample volume is passaged successively to a specific diluent volume. Here, we demonstrated that a higher sample volume is a better representation of bacterial population than a lower sample volume, which was in concordance with the random nature of bacterial distribution in culture. Therefore, a bigger sample to diluent ratio during serial dilution appears more favorable for an accurate bacterial enumeration than a smaller ratio. But surprisingly, enumeration using the different dilution ratios such as 1:9, 1:99 and 1:999 in 1.0 mL final volume yielded similar results with the exception of 1:999, where 1 L sample was passaged. However, in 10.0 mL final volume of dilution, the above three dilution ratios exhibited similar bacterial enumeration. The experiment was performed using two different bacterial cultures such as Escherichia coli and Ralstonia pseudosolanacearum. Our results indicated that the advantage gained due to lesser number of passages in case of a lower sample volume could overcome the disadvantage associated with it, thereby co-aligning the different dilution ratios with regards to enumeration. Hence, although in laboratory, 1:9 dilution ratio is usually performed during serial dilution, our results suggest that dilution ratios such as 1:99 in 1 mL dilution volume and ratios such as 1:99 and 1:999 in 10 mL dilution volume are equally effective, which also reduces time, cost and labor.

Total and viable microbial cell counts are increasingly important for applications including live biotherapeutic products, food safety, and probiotics. In microbiology, cells are quantified using methods such as colony forming unit (CFU), flow cytometry, and polymerase chain reaction (PCR), but different methods measure different aspects of the cells (measurands), and results may not be directly comparable across methods. In the absence of a ground-truth reference material for cell count, one cannot quantify the accuracy of any cell counting method, which limits method performance assessments and comparisons. Herein, a modified analysis of cell counting methods based on the ISO 20391-2:2019 standard was developed and demonstrated for microbial cell samples diluted over a log-scale range of concentrations. Escherichia coli samples ranging in concentration from approximately 5 x 105 cells/mL to 2 x 107 cells/mL were quantified using CFU, Coulter principle, fluorescence flow cytometry, and impedance flow cytometry. Quality metrics modified from the ISO standard were calculated for each method and shown to be repeatable across replicate experiments. The quality metrics illustrate large differences in proportionality and variability across methods, with total cell counts in good agreement and viable cell count having more variability. As the ISO standard is meant to guide fit-for-purpose method selection, interpretation of the results and quality metrics can drive method choice and optimization. The framework introduced here will help researchers select fit-for-purpose counting methods for quantification of microbial total and viable cells across a range of applications.

The dynamics of bacterial population decline at antibiotic concentrations above the minimum inhibitory concentration (MIC) remain poorly characterized. This is because measuring colony-forming units (CFU), the standard assay to quantify inhibition, is slow, labour-intensive, costly, and can be unreliable at high drug concentrations. Luminescence assays are widely used to quantify population dynamics at subinhibitory concentrations, yet their limitations and reliability at super-MIC concentrations remain underexplored. To fill this gap, we compared luminescence- and CFU-based rates across 20 antimicrobials. In our experiments luminescence- and CFU-based rates did not differ significantly for half of them. For the other half, CFU-based estimates of rates of decline were consistently higher. The estimates differed for two main reasons: First, because light intensity tracks biomass more closely than population size, luminescence declined more slowly than the population when bacteria filamented. Second, CFU-based estimates indicated a steeper decline when antimicrobial treatment reduced the number of colonies formed per plated bacterium. This effect can result from changes in clustering behaviour, physiological changes that impair culturability, or antimicrobial carry-over. Thus, the suitability of luminescence to quantify bacterial decline depends on the physiological effects of the antimicrobial used (e.g. filamentation) and whether the quantity of interest is cell number or biomass. Within these limitations, luminescence can serve as an efficient, high-throughput alternative for quantifying bacterial dynamics at super-MIC concentrations.

Diastatic Saccharomyces cerevisiae is a common contaminant in the brewing industry. Currently available detection methods are either time-coansuming or require specialized equipment. The aim of this study was to develop a new rapid and simple assay for the detection of diastatic yeast from beer and yeast samples. More specifically, we aimed to develop a simple and rapid assay that requires minimal laboratory equipment or training, and ideally yields results as accurate as PCR-based methods. The developed assay consisted of three main steps: DNA extraction, pre-amplification of DNA, and CRISPR-Cas12a-based detection and visualisation. We compared different preamplification and visualisation techniques, and the final assay involved a one-pot reaction where LAMP and Cas12a were consecutively used to pre-amplify and detect a fragment from the STA1 gene in a single tube. These reactions only required a heat block, a pipette, and a centrifuge. The assay result was then visualised on a lateral flow strip. We used the developed assay to monitor an intentionally contaminated beer fermentation, and it was shown to yield results as accurate as PCR using previously published primers. Furthermore, the assay yielded results in approx. 75 minutes starting from a beer sample. The developed assay therefore offers reliable and rapid quality control for breweries of all sizes and can be performed without any expensive laboratory equipment. We believe the assay will be particularly useful for smaller breweries that dont already have well-equipped laboratories and are looking to implement better quality control.

Profiling the growth of bacterial cultures over time can be a tedious and error-prone process. Here, we present the development and evaluation of the use of the ODity platform to optically measure bacterial cell densities non-invasively. The digital growth data for E. coli MG1655 was calibrated against colony forming units (CFU/mL) obtained by plating on solid media. Diauxic-like shifts of liquid E. coli MG1655 cultures grown at 37{degrees}C in LB media were observed at densities as low as 2.9 x 107 {+/-} 1.2 CFU/mL. The shift occurred at a significantly higher cell density (6.0 x 107 {+/-} 1.2 CFU/mL) when the bacteria were cultured at 31{degrees}C. These shifts were only short lived, 15.2 {+/-} 1.5 and 20.8 {+/-} 1.8 min at 37{degrees}C and 31{degrees}C, respectively, with the previous growth rate restored thereafter. We measured minimum doubling times of 17.0 {+/-} 1.1 and 24.8 {+/-} 0.9 min at 37{degrees}C and 31{degrees}C, respectively. These results demonstrate that the growth and growth rate of bacterial cultures can be accurately determined non-invasively using the ODity device.

The {beta}-lactams are critically important broad-spectrum antibiotics, widely used as first-line treatments; however, their effectiveness is increasingly compromised by {beta}-lactamase enzymes. Among these, OXA-type enzymes have expanded to over 400 variants and are highly prevalent in Enterobacteriaceae. Current phenotypic and molecular detection tests have long turnaround times or require specialized equipment, respectively. In this study, we optimize a rapid molecular assay combining a PCR with modified thermal ramp rate (TRR) along with CRISPR-Cas12a fluorescence detection for the blaOXA-1gene. Using a commercial DNA Taq polymerase (TRR: 2.2 {degrees}C/s, annealing and extension hold time: 1 s), amplification time was reduced from 80 to 30 min, enabling detection within 50 min (PCR: 30 min; CRISPR: 20 min). With a locally produced enzyme (hold: 10 s), amplification time was 44 min. The assay achieved an analytical sensitivity of 8 CFU/reaction using commercial DNA Taq polymerase. The accelerated PCR:CRISPR workflow delivers results in less than one hour without compromising technical sensitivity (attomoles range), not requiring high technical expertise, and can be implemented in laboratories with basic molecular biology equipment. HighlightsAn optimized thermal gradient can reduce the turnaround time of PCR-based detection tests CRISPR-Cas in addition to the modified PCR can detect a gene target in less than an hour The proposed workflow is suitable for implementation with basic molecular biology equipment

Bacteria increase in biomass and divide, but determining precisely when cell division completes is technically challenging. To aid time-lapse imaging and cell-cycle tracking, we set out to identify a protein in Bacillus subtilis, which when fused with a fluorophore would cause the membrane to fluoresce in a manner that was constitutive, uniform, and bright. A forward genetic transposon-based approach combined with fluorescence-activated cell sorting was used to identify a fluorescent fusion to the glucose PTS transport transmembrane protein PtsG with all desired properties. Moreover, PtsG-GFP was constitutive and neutral to growth under all conditions tested and also labeled membranes during sporulation. We used PtsG-GFP to track cell growth in microfluidic channels and determine when cytokinesis occurred, defined as when fluorescence reached a local maximum at the division plane. Simultaneous imaging with a compatible fluorescent fusion to the cell division protein FtsZ indicated that FtsZ peak intensity occurred midway through septum constriction and that Z-ring recycling coincided with cytokinesis. We conclude that PtsG-GFP is a powerful tool for membrane imaging and cell cycle tracking. As such, we provide constructs with fluorophores that emit across the visible spectrum and antibiotic resistance cassettes to facilitate deployment in B. subtilis. IMPORTANCEBacterial cells are fully divided when new membrane separates the cytoplasm of each daughter. Reproducibly staining of bacterial membranes with exogenous labels for fluorescence microscopy can be challenging, particularly during chemostatic growth in microfluidic devices. Here, we report that fusion of a fluorescent protein to the glucose transport protein PtsG causes the membrane of Bacillus subtilis to give off bright and even fluorescence under a variety of conditions. We use PtsG-GFP to operationally define when cytokinesis occurs during growth, and we note that a fluorescent PtsG fusion would likely make fluorescent staining of the membrane more facile theoretically in any organism.

Objectivesthis study aimed to develop a fully automated, non-destructive and label-free identification method of bacterial colonies, directly on agar plates, using a combination of digital holography and artificial intelligence and to evaluate its performances. Methodhigh-resolution holographic images of individual colonies on translucent brain-heart agar plates were taken every 30 minutes throughout an 18-hour incubation period (530 MPx for the full plate) using a large field 1x magnification system, a partially coherent LED light source and a high-resolution CMOS sensor. A database containing 49 490 digital holograms of individual colonies from 276 clinical strains belonging to ten of the most prevalent pathogenic bacterial species was used to train the convolutional neural network (CNN). Improvement in the accuracy of the prediction from the CNN algorithms was achieved using the information at different phylogenetic levels. Resultsthe performance of the BAIO-DX solution was assessed on 232 strains belonging to the 10 species used to train the algorithms but also on 64 strains from 8 species not included in the training database. For the species included in the training dataset, this new method identified 86.6% of the strains at the species level with a positive-percent agreement of 96.5%. An additional 48% of the strains not identified at the species level could be identified at the genus level thanks to the phylogenetic interpretation of the results. Conclusionsthese first results validate this approach as a candidate to obtain a fully automated non-destructive and label-free solution for bacterial identification in clinical microbiology laboratories. IMPORTANCE STATEMENTIdentification of pathogenic bacteria by culture-based methods are typically performed using MALDI-TOF mass spectrometry or biochemical systems. While automation and interpretive algorithms based on agar plate imaging and artificial intelligence (AI) has reduced manual steps, bacterial identification is still labor-intensive. Here we developed a fully automated, non-destructive and label-free identification method of bacterial colonies at the species level, directly on agar plates, using a combination of digital holography and convolutional neural network algorithms. After training the system with 276 strains belonging to ten of the most frequent pathogenic bacterial species, the BAIO-DX solution was able to identify 86.6% of new strains from these 10 species with a positive-percent agreement of 96.5%. These thorough proof of concept shows that imaging methods coupled to AI algorithms are promising to reach a fully automated identification of a significant proportion of pathogenic bacteria and has potential to enhance diagnostic workflows in clinical microbiology.

In microbiology, the estimation of the growth rate of microorganisms is a critical parameter to describe a new strain or characterize optimal growth conditions. Traditionally, this parameter is estimated by selecting subjectively the exponential phase of the growth, and then determining the slope of this curve section, by linear regression. However, for some experiments, the number of points to describe the growth can be very limited, and consequently such linear model will not fit, or the parameters estimation can much lower and strongly variable. In this paper, we propose a tools to estimate growth parameters using a logistic Verhulst model that take into account the entire growth curve for the estimation of the growth rate. The efficiency of such model is compared to the linear model. Finally, the novelty of our work is to propose a "Shiny-web application", online, without any programming or modelling skills, to allow estimating growth parameters including growth rate, maximum population, and beginning of the exponential phase, as well as an estimation of their variability. The final results can be displayed in the form of a scatter plot representing the model, its efficiency and the estimated parameters are downloadable.

While culturing fungal pathogens on artificial media, the phenotype, i.e. colony aspect, growth kinetics and in some case the pigment production can be used as discrimination criteria. Our aim was to enable the comparison between the growth kinetics of 8 fungal species, cultured simultaneously on 3 different media, and imaged under standardised conditions. Using a camera stand, taking pictures was quick and easy, producing homogeneous pictures, facilitating their comparison. The colony growth kinetics varied widely across the 8 species. The mycelium aspect and pigment production depended on the media and changed over time. Having standardised the imaging setup and grown species simultaneously allows proposing reference sets of pictures.

Microbial flow cytometry allows to rapidly characterize microbial communities. Recent research has demonstrated a moderate to strong connection between the cytometric diversity and taxonomic diversity based on 16S rRNA gene amplicon sequencing data. This creates the opportunity to integrate both types of data to study and predict the microbial community diversity in an automated and efficient way. However, microbial flow cytometry data results in a number of unique challenges that need to be addressed. The results of our work are threefold: i) We expand current microbial cytometry fingerprinting approaches by proposing and validating a model-based fingerprinting approach based upon Gaussian Mixture Models, which we called PhenoGMM. ii) We show that microbial diversity can be rapidly estimated by PhenoGMM. In combination with a supervised machine learning model, diversity estimations based on 16S rRNA gene amplicon sequencing data can be predicted. iii) We evaluate our method extensively by using multiple datasets from different ecosystems and compare its predictive power with a generic binning fingerprinting approach that is commonly used in microbial flow cytometry. These results demonstrate the strong connection between the genetic make-up of a microbial community and its phenotypic properties as measured by flow cytometry. Our workflow facilitates the study of microbial diversity and community dynamics using flow cytometry in a fast and quantitative way. ImportanceMicroorganisms are vital components in various ecoystems on Earth. In order to investigate the microbial diversity, researchers have largely relied on the analysis of 16S rRNA gene sequences from DNA. Flow cytometry has been proposed as an alternative technique to characterize microbial community diversity and dynamics. It is an optical technique, able to rapidly characterize a number of phenotypic properties of individual cells. So-called fingerprinting techniques are needed in order to describe microbial community diversity and dynamics based on flow cytometry data. In this work, we propose a more advanced fingerprinting strategy based on Gaussian Mixture Models. When samples have been analyzed by both flow cytometry and 16S rRNA gene amplicon sequencing, we show that supervised machine learning models can be used to find the relationship between the two types of data. We evaluate our workflow on datasets from different ecosystems, illustrating its general applicability for the analysisof microbial flow cytometry data. PhenoGMM facilitates the rapid characterization and predictive modelling of microbial diversity using flow cytometry.

Large screens of bacterial strain collections to identify potential biocontrol agents are often time consuming, costly, and fail to provide quantitative results. In this study, we present two quantitative and high-throughput methods to assess the inhibitory capacity of bacterial biocontrol candidates against fungal phytopathogens. One method measures the inhibitory effect of bacterial culture supernatant components on the fungal growth, while the other accounts for direct interaction between growing bacteria and the fungus by co-cultivating the two organisms. The antagonistic supernatant method quantifies the culture components antifungal activity by calculating the cumulative impact of supernatant addition relative to a non-treated fungal control, while the antagonistic co-cultivation method identifies the minimal bacterial cell concentration required to inhibit fungal growth by co-inoculating fungal spores with bacterial culture dilution series. Thereby, both methods provide quantitative measures of biocontrol efficiency and allow prominent fungal inhibitors to be distinguished from less effective strains. The combination of the two methods shed light on the type of inhibition mechanisms and provide the basis for further mode of action studies. We demonstrate the efficacy of the methods using Bacillus spp. with different levels of antifungal activities as model antagonists and quantify their inhibitory potency against classic plant pathogens. ImportanceFungal phytopathogens are responsible for tremendous agricultural losses on annual basis. While microbial biocontrol agents represent a promising solution to the problem, there is a growing need for high-throughput methods to evaluate and quantify inhibitory properties of new potential biocontrol agents for agricultural application. In this study, we present two high-throughput and quantitative fungal inhibition methods that are suitable for commercial biocontrol screening.

BackgroundDetecting bacteria at the strain level is crucial in microbiology. Although qPCR is widely used, designing strain-specific primers remains a challenge due to nucleotide sequence similarities among related strains. Methods and ResultsThis paper introduces a simplified, web-based workflow for designing strain-specific primers using publicly available microbial genomes. The method does not require advanced bioinformatics skills and can be applied using a basic computer. Primers designed using this workflow are applied to assess the survival of two close Bacillus strains in soil microcosms. ConclusionThe workflow offers an accessible solution for accurate bacterial strain detection, and fills a gap for researchers without specialized training in bioinformatics.

Many modern molecular analysis methods utilize DNA content values as part of the measurement process, and thus, the distribution of genome copies per cell within a population of cells is important. Genome copy distributions can be measured via flow cytometry by thresholding (or "gating") a subset of cells from which estimates of the targeted properties (e.g., genome copy number) can be calculated. This manuscript introduces a new approach that gives separate estimates of signal and noise, the former of which is used for gating and analysis, and the latter is used to quantify uncertainty. In this approach stochastic regression was used to quantify subpopulations of cells that have distinctly different genome copies per cell within a heterogenous population of Escherichia coli (E. coli) cells. By separating the signal and noise components, they can be used independently to evaluate measurement quality across different experimental conditions.

IntroductionThe high number of uncontrollable variables in microbiological systems increases experimental complexity and reduces accuracy, potentially leading to data misinterpretation or uncorrectable errors. During an interlaboratory calibration analysis it was observed that the microbial logarithmic reduction (LR) caused by disinfectants depends not only on the type of disinfectant but also on the initial microbial load in the fabric carriers, which can produce a misinterpretation of the results. Fabric carriers are commonly used in standard tests such as EN16616 and ASTM2274. ObjectiveA method based on statistical calibration is proposed using a regression line between N0 (initial microbial load in the carrier) and LR to eliminate the influence of one on the other. ResultsAn example with Candida albicans is presented. Once the method was applied, the influence of N0 on LR was eliminated and the new LR values can be used for factorial experiments, for example, to check the efficacy of disinfectants or detergents without depending on the microbial load placed in the carrier.

Loop-mediated isothermal amplification (LAMP) is a widely used rapid and affordable molecular DNA amplification method with minimal resource requirements. However, visual interpretation of results is subjective and prone to errors, leading to potential false-positive and negative results. To address this limitation, a machine-learning approach is proposed for automated LAMP classification based on digital images. The approach utilizes You Only Look Once (YOLOv8), a fast and robust object detection algorithm to locate and classify tubes within LAMP images, enabling automated categorization as positive or negative. The trained model achieved a high overall accuracy of 95.5% in classifying LAMP images into positive or negative. Additionally, the approach had a 98.0% precision and 92.7% recall for positive cases and 93.4% precision and 98.2% recall for negative cases, demonstrating its potential for real-time LAMP diagnosis and enhanced assay performance. This project demonstrated the platforms suitability for real-time testing, offering an easy operation and rapid results.