Click on “Download PDF” for the PDF version or on the title for the HTML version.

If you are not an ASABE member or if your employer has not arranged for access to the full-text, Click here for options.

Feasibility of using visible/near-infrared (Vis/NIR) spectroscopy to detect aflatoxigenic fungus and aflatoxin contamination on corn kernels

Published by the American Society of Agricultural and Biological Engineers, St. Joseph, Michigan

Citation:  2018 ASABE Annual International Meeting  1801006.(doi:10.13031/aim.201801006)
Authors:   Feifei Tao, Haibo Yao, Fengle Zhu, Zuzana Hruska, Yongliang Liu, Kanniah Rajasekaran, Deepak Bhatnagar
Keywords:   aflatoxin, aflatoxigenic fungus, corn kernel, visible/near-infrared spectroscopy, PLS-DA, PCA-QDA.

Abstract. The demand for developing a rapid and non-destructive method for sensing aflatoxin contamination and/or aflatoxigenic fungal infection that is suitable to real-time and on-line detection has received significant attentions. In this study, we utilized the visible/near-infrared (Vis/NIR) spectroscopy over the spectral range of 400-2500 nm to detect AF13-inoculated corn kernels and the corresponding aflatoxin contamination. A total of 180 corn kernels were used with 3 treatments included, namely, 60 kernels inoculated with AF13 (aflatoxigenic) strain, 60 kernels inoculated with AF38 (non-aflatoxigenic) strain, and 60 kernels inoculated with sterile distilled water as control. Both chemometric methods of partial least squares discriminant analysis (PLS-DA) and principal component analysis combined with quadratic discriminant analysis (PCA-QDA) were employed to develop the classification models. The obtained results indicated the potential of Vis/NIR spectroscopy combined with appropriate chemometric methods in differentiating the AF13-inoculated corn kernels from the AF38-inoculated and control kernels, and identifying the corresponding aflatoxin contamination. The best overall accuracy in classifying the AF13-inoculated and “control+AF38-inoculated” corn kernels achieved 91.1% using the PLS-DA method. Specifically, the accuracy attained in identifying the AF13-inoculated corn kernels was 100.0% using this model. The best overall accuracy obtained in classifying the healthy and aflatoxin-contaminated corn kernels was over 82.0%, using either the threshold of 20 ppb or 100 ppb. The best accuracy in identifying the aflatoxin-contaminated corn kernels achieved 90.9% and 88.9%, with the classification threshold of 20 ppb and 100 ppb, respectively.

(Download PDF)    (Export to EndNotes)