Submitted to: Sensing and Instrumentation for Food Quality and Safety
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: March 9, 2008
Publication Date: June 25, 2008
Citation: Yao, H., Hruska, Z., Kincaid, R., Brown, R.L., Cleveland, T.E. 2008. Differentiation of Toxigenic Fungi Using Hyperspectral Imagery. Sensing and Instrumentation for Food Quality and Safety. 2:215-224 Interpretive Summary: Aflatoxins are poisons produced by the fungus Aspergillus flavus after it infects agricultural commodities such as corn. Traditionally, corn kernels being assessed for aflatoxin contamination are examined with an ultra-violet light for evidence of bright greenish-yellow fluorescence (BGYF), indicative of the presence of A. flavus. This approach is time and labor intensive and somewhat inaccurate. The present research focused on using a new technology, hyperspectral imaging, for fungal detection. An experiment was conducted to differentiate fungal species using a tabletop hyperspectral imaging system. A total of five toxin-producing fungal species were used in the experiment: Penicillium chrysogenum, Fusarium moniliforme, Aspergillus parasiticus, Trichoderma viride and Aspergillus flavus. All were grown on agar and imaged with the hyperspectral system on the 5th day of growth. Results indicate that the five fungi are separable to a high degree of accuracy. Further development may render this technology useful for rapidly detecting the presence of aflatoxin-producing fungi in corn being sold and without destroying the grain. Corn growers, buyers and consumers would, therefore, be provided with greater food and feed protection and with less financial loss.
Technical Abstract: Pathogenic fungi, including Aspergillus flavus, are mycotoxin-producing and can contaminate grain products, such as corn. The contaminated food poses a threat to humans and animals. Traditionally, corn kernels have been examined for evidence of bright greenish-yellow fluorescence (BGYF), indicative of the presence of A. flavus, when illuminated with a high-intensity ultra-violet light. The BGYF approach is time and labor intensive and somewhat inaccurate. Several previous studies have examined spectral-based, non-destructive methods for the detection of fungi and toxins. The present report focused on using spectral image data for fungal detection. A tabletop hyperspectral imaging system, VNIR-100E, was used for high spectral and high spatial resolution data acquisition. A total of five toxin producing fungal species were used in a two-part experiment: Penicillium chrysogenum, Fusarium moniliforme, Aspergillus parasiticus, Trichoderma viride and Aspergillus flavus. All fungal isolates were cultured on agar in Petri-dishes under lab conditions and were imaged on day 5 of growth. The objective of the study was to use hyperspectral imagery for classification of different fungal species. Results indicate that the five fungi are separable to a high degree of accuracy. In addition, all five toxigenic fungi could be classified by using only three narrow bands (bandwidth = 2.43 nm) centered at 743 nm, 458 nm and 541 nm.