Location: Food and Feed Safety Research
Project Number: 6054-42000-027-006-S
Project Type: Non-Assistance Cooperative Agreement
Start Date: Sep 3, 2023
End Date: Sep 2, 2025
Identify and quantify corn kernels contaminated with aflatoxin-producing fungi, using non-destructive hyperspectral imaging systems for elimination from food supply. Produce spectral libraries for fungus alone and corn kernels infected with the fungus. Determine spectral differences between resistant and susceptible corn varieties to aflatoxin contamination and infected and uninfected corn kernels with aflatoxin producing fungi. Develop rapid, non-destructive hyperspectral and multispectral imaging methodology to measure fungal growth and aflatoxins and spectral signatures associated with traits for resistance to fungal infection and aflatoxin contamination in corn kernels. Test systems’ effectiveness in identification of contaminated kernels under laboratory and field situations.
Different corn varieties lab and/or field-infected with aflatoxin-producing and non-producing fungi will be collected and imaged using tabletop hyperspectral scanning imaging systems including fluorescence, visible/near infrared (VIS/NIR), shortwave near infrared (SWIR), and Raman. Infected kernels will be spectrally analyzed to determine differences in fungal infection captured by each system. Cultures of different aflatoxin-producing and non-producing fungi will also be imaged, and the spectral fingerprints will be collected and added to our existing "spectral library" of the different strains of fungi. These data will be used to determine if hyperspectral imaging can then be used to differentiate and possibly quantitate the varying fungal strains and/or their aflatoxin production both in pure fungal culture and in fungally infected kernels of different corn varieties. Techniques also will be investigated during ongoing experiments to determine the best imaging environment to accomplish hyperspectral and multispectral analyses, such as type and direction of lighting, automated multispectral-based detection, for different wavelength ranges. Once appropriate algorithms are developed, the imaging systems will be tested under various laboratory and field experimental conditions to determine the efficacy of the imaging systems.