Detection of aflatoxin B1 content and revelation of its dynamic accumulation process using visible/near-infrared hyperspectral and microscopic imaging

Authors: GUO, XIAOHUAN - China Agricultural University; WANG, WEI - China Agricultural University; JIA, BEIBEI - Chinese Academy Of Inspection And Quarantine; Ni, Xinzhi; Zhuang, Hong; Yoon, Seung; Gold, Scott; Pokoo-Aikins, Anthony; Mitchell, Trevor; Bowker, Brian; YE, JIAWEI - China Agricultural University

Submitted to: Food Microbiology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 1/12/2025
Publication Date: 3/2/2025
Citation: Guo, X., Wang, W., Jia, B., Ni, X., Zhuang, H., Yoon, S.C., Gold, S.E., Pokoo-Aikins, A., Mitchell, T.R., Bowker, B.C., Ye, J. 2025. Detection of aflatoxin B1 content and revelation of its dynamic accumulation process using visible/near-infrared hyperspectral and microscopic imaging. Food Microbiology. https://doi.org/10.1016/j.ijfoodmicro.2025.111065.
DOI: https://doi.org/10.1016/j.ijfoodmicro.2025.111065

Interpretive Summary: In this study, the Aspergillus flavus (AFB1) development and the levels of AFB1 grown on Potato Dextrose Agar (PDA), PDA+L-glutamine (Gln), and PDA+rapamycin (RAPA) culture media were monitored using visible/near-infrared (Vis/NIR) hyperspectral imaging (HSI) and microscopic imaging and analyzed with chemometric methods. The study found that AFB1 accumulation was heterogeneous in temporal and spatial scope (colony diameter direction) and sensitive to substrates. The cycle areas and the aflatoxin levels were altered by adding enhancers (Gln) and inhibitors (RAPA) to the substrate. Gln promoted colony growth and spore formation and increased the accumulation of AFB1. Conversely, RAPA inhibited the colony development and AFB1 production. At an incubation time of 120 h, the distribution of heterogeneous regions, the morphology characteristics of hyphae and spores, and aflatoxin concentration were also affected in the parameter space influenced by Gln and RAPA. Each colony was segmented to determine AFB1 levels and observe the microscopic morphology. The study found that the AFB1 content in the area of hyphae aging and dissolution decreased, and the difference seemed to be related to the morphology of the hyphae. Through principal component analysis, spectral bands related to the color changes of hyphae and spores during colony growth (such as 480, 623, 674, 726 nm), and key wavelengths related to changes in nutritional composition (such as 840, 867, 882, 967 nm) were identified. By combining Vis/NIR-HSI with chemometric algorithms, a prediction model for AFB1 levels on culture media has been successfully constructed. The study outcome suggested that HSI would be useful for monitoring fungal growth and detecting aflatoxin contamination, ensuring food safety, and detecting grain seeds contaminated by fungi and aflatoxins.

Technical Abstract: Understanding and controlling the dynamic process of aflatoxin B1 (AFB1) accumulation by Aspergillus flavus (A.flavus) remains challenging. In the temporal and spatial domains, fungal colonies exhibited concentric circular response patterns. In this study, the A.flavus development and the levels of AFB1 by inoculation on Potato Dextrose Agar (PDA), PDA+L-glutamine (Gln), and PDA+rapamycin (RAPA) culture media were monitored using visible/near-infrared (Vis/NIR) hyperspectral imaging (HSI). A continuous increase in AFB1 content was observed in the PDA and PDA+Gln groups, and the growth of A.flavus and the accumulation of aflatoxin were promoted in the PDA+Gln group. In the PDA+RAPA group, the development of fungi and the production of AFB1 were inhibited. Meanwhile, the levels of AFB1 in various heterogeneous areas of the colony were measured and their microscopic morphological characteristics were observed. The study revealed that the distribution of AFB1 across the fungal colony was uneven, and this heterogeneity appeared to be associated with the aging and autolysis of the hyphae. Spectral bands (480, 623, 674, 726 nm) related to the color changes of hyphae and spores during colony growth, and key wavelengths (840, 882, 867, 972 nm) related to changes in nutritional composition were identified by principal component analysis (PCA) analysis. Multiple preprocessing techniques and modeling methods were employed to construct regression models for predicting AFB1 contents, with first-derivative (1stD) and partial least squares regression (PLSR) demonstrating the best performance. Finally, a spatial map of AFB1 levels was established using the optimal model, showing a spatial pattern similar to the measurement results. This study highlights the potential of Vis/NIR HSI for monitoring the growth of A.flavus and AFB1 detection.