Application Study: Mold quantification on yogurt

Objective

Detecting and quantifying fungal growth on yogurt samples.

Introduction

Detection and quantification of mold growth on dairy products allows for evaluation of shelf-life or efficacy testing of bioprotective organisms such as Lactococcus lactis and natural antimicrobial preservatives like natamycin and nisin. In this application study, the growth of four different fungal spoilage organisms on yogurt samples was analyzed using the Reshape Imaging System in combination with AI-powered image analysis. We highlight the system's ability to detect the time of mold appearance, quantify fungal growth, and increase throughput.

Results

Imaging of mold growth. The Reshape Imaging System enabled simultaneous incubation and recording of 10 multiwell plates with yogurt samples inoculated with fungal spore suspensions of four different species (Fig. 1). To better visualize white pigmented fungal species, fruit coloring was added to half of the multiwell plates (Fig. 1 bottom).

_{Figure 1. Mold growth on yogurt. 3 ml yogurt sample was added to each well of a 6-well plate and inoculated with fungal spore suspensions. The assay was incubated at 27°C for 5 days and continuously recorded using the Reshape Imaging System. Left to right: Control and Fungus 1-4. Top: Plain yogurt. Bottom: Yogurt with fruit coloring.}

Automatic detection of mold growth. Reshape's AI-powered image analysis allowed automatic detection of fungal growth on yogurt samples. The annotation masks indicate the detected region of interest on every image, enabling analysis of growth development over time (Fig. 2).

_{Figure 2. Detection and annotation of mold. The fungal growth was automatically detected by AI-powered image analysis models. Annotation masks indicate the detected growth on each single image of the timelapse.}

Continuous growth analysis. The growth of the four fungal species was imaged at hourly timepoints, allowing continuous fungal growth analysis by AI-powered models (Fig. 3). Mold 4 displayed the fastest growth rate and reached maximum coverage of the well within 60 h, whereas Mold 1 showed a slower growth rate compared to the other fungi.

_{Figure 3. Growth analysis. The growth of the fungal species was imaged at hourly timepoints, allowing continuous growth analysis and quantification.}

Automatic detection of time of appearance. The time of appearance was determined for each fungal species on all samples and averaged between replicates. Mold 2 appeared first, despite Mold 4 displaying the fastest growth rate.

Improving throughput with assay downscaling. To enhance throughput and accommodate large screens, the assay was implemented on microtiter plates, facilitating continuous tracking and simultaneous assessment of 960 samples.

_{Figure 5. Downscaling mold assessment on yogurt. Four fungal species were inoculated in 96-well plates on plain yogurt samples (top) or on yogurt with fruit coloring (bottom).}

Automated HT-analysis. In high throughput assays, Reshape's color change assessment or image classification enables automated analysis of microbial growth on microtiter plates. In smalle wells, where CFU detection is not feasible, color development can serve as a measure for fungal growth (Fig. 6).

_{Figure 6. Color assessment of mold. Continuous color quantification was employed as a measure of fungal growth, exemplified with Mold 2. Standard errors are indicated with dashed lines.}

Conclusion

Reshape's AI-powered image analysis facilitates the automated detection of fungal growth on food samples and food matrices. In this application study, the growth of four different fungal species was imaged and analyzed, providing valuable insights including time of appearance, growth quantification, and growth rate. Additionally, the assay was downscaled to microtiter plates, demonstrating that Reshape's technology can efficiently aid high throughput and assay automation.