LAUSR

BACKGROUND Fungal contamination in food products leads to mustiness, biochemical changes, and undesirable odor, which results in lower food quality and market value. In order to develop a rapid method for fungi detection, hyperspectral imaging (HSI) was applied to identify five fungi inoculated on plates (Aspergillus niger, Aspergillus flavus, Penicillium chrysogenum, Aspergillus fumigatus, Aspergillus ochraceus). Meanwhile, near infrared (NIR), mid infrared (MIR) and electronic nose (E-nose) were applied to detect and identify freeze-dried A. bisporus infected with five fungi. RESULTS For HSI spectrum of the five fungi on plates, PLSR models were used to distinguish these groups. A. ochraceus group had the highest calibration performance (Rc 2 =0.786, RMSEC=0.125 log CFU/g), while A. flavus group had the highest prediction performance (Rp 2 = 0.821, RMSEP = 0.083 log CFU/g). RPD values of all models were higher than 2.0 for NIR, MIR and E-nose results of freeze-dried A. bisporus infected with different fungi. Fungal species and infection degree can be distinguished by principle component analysis (PCA) and partial least squares discriminant analysis (PLS-DA) using NIR, MIR and E-nose, since the discrimination accuracy was more than 90%. In addition, NIR methods had higher recognition rate than MIR and E-nose methods. CONCLUSION PCA and PLSR models based on full spectra of HSI can achieve good discrimination results for the five fungi growth on plates. Moreover, NIR, MIR and E-nose were proven to be effective in the monitoring of fungal contamination on freeze-dried A. bisporus. However, NIR could be a more accurate method. This article is protected by copyright. All rights reserved.