Growth Identification of Aspergillus flavus and Aspergillus parasiticus by Visible/Near-Infrared Hyperspectral Imaging

Abstract

Visible/near-infrared (Vis/NIR) hyperspectral imaging (400-1000 nm) was applied to identify the growth process of Aspergillus flavus and Aspergillus parasiticus. The hyperspectral images of the two fungi that were growing on rose bengal medium were recorded daily for 6 days. A band ratio using two bands at 446 nm and 460 nm separated A. flavus and A. parasiticus on day 1 from other days. Image at band of 520 nm classified A. parasiticus on day 6. Principle component analysis (PCA) was performed on the cleaned hyperspectral images. The score plot of the second to sixth principal components (PC 2 to PC 6 ) gave a rough clustering of fungi in the same incubation time. However, in the plot, A. flavus on day 3 and day 4 and A. parasiticus on day 2 and day 3 overlapped. The average spectra of each fungus in each growth day were extracted, then PCA and support vector machine (SVM) classifier were applied to the full spectral range. SVM models built by PC 2 to PC 6 could identify fungal growth days with accuracies of 92.59% and 100% for A. flavus and A. parasiticus individually. In order to simplify the prediction models, competitive adaptive reweighted sampling (CARS) was employed to choose optimal wavelengths. As a result, nine (402, 442, 487, 502, 524, 553, 646, 671, 760 nm) and seven (461, 538, 542, 742, 753, 756, 919 nm) wavelengths were selected for A. flavus and A. parasiticus, respectively. New optimal wavelengths SVM models were built, and the identification accuracies were 83.33% and 98.15% for A. flavus and A. parasiticus, respectively. Finally, the visualized prediction images for A. flavus and A. parasiticus in different growth days were made by applying the optimal wavelength's SVM models on every pixel of the hyperspectral image.

Detection of Antibiotic Constituent in Aspergillus flavus Using Quantum Convolutional Neural Network

“…In this article authors have extracted third stage of the flavus that is responsible for the delivery of required proteins causing able metabolism. In one of the research works accomplished by Chu et al (2018) focused on the prediction of parasitic and non-parasitic flavus based on spectroscopy images in connection to human beings. In their work, they have evidently classified the growth of the flavus into different stages and when it crosses the third stage of growth, the flavus turns out to be parasitic causing effects for human health.…”

Fast recognition model of Fusarium in agaric based on hyperspectral imaging

“…The SVM method is a learning system based on a hyperplane, which uses functions to map the data to high-dimensional feature spaces (Chu et al, 2018;Gromski al., 2014;Huang et al, 2020). DA is to establish discriminant function through the relationship between grouping variable and characteristic variable for classification.…”

“…MCE is a classical indicator, which can be used to control how much misclassification is considered . Since they have been widely used in classification in previous studies (Chu et al, 2018), they were used in this study to distinguish between fungal species and different culture time.…”

“…In summary, spectral data can be collected for species differentiation at 48 h during Fusarium growth time, which can save time compared to traditional culture methods for species identification. Chu et al (2018) found that by classifying hyperspectral images of Aspergillus flavus and Aparasiticus parasiticus with different levels of contamination, the PCA-SVM model was accurate up to 92.59% and 100%. Shen et al (2019) found that different Fusarium species could be distinguished from different groups of wheat samples on days 1, 3, and 5.…”

Non-Destructive Detection of Moldy Walnuts Based on Hyperspectral Imaging Technology

“…Hyperspectral imaging can make up for the shortcomings of near-infrared spectroscopy by combining images with near-infrared spectroscopy to obtain three-dimensional optical image data comprising a range of wavelengths [ 20 ]. Ma et al [ 21 ] collected hyperspectral images of “Wen 185” walnut kernels from Xinjiang in the ranges of 862.9–1704.02 nm and 382.19–1026.66 nm and established protein and fat content prediction models, respectively.…”