Proceedings of The World Conference on Engineering and Technology
Visible/near-infrared Spectroscopy as a Novel Technology for Nondestructive Detection of Escherichia coli ATCC 8739 in Lettuce Samples
Sahar Rahi, Hossein Mobli, Bahareh Jamshidi, Aslan Azizi and Mohammad Sharifi
Microbial contamination of vegetables as a major safety challenge is known in recent years. Detection of Escherichia coli (E. coli ATCC 8739) in vegetables especially lettuce is very important issue for post-harvest management in Iran. Non-destructive detection and evaluation of E. coli contamination in fresh lettuce leaf as a main aim of this work was explored using Visible Near-infrared spectroscopy (Vis/NIR). Reflectance spectra in intractance mode were measured using a spectrometer at the range of 350–1100 nm, and 200 fresh lettuce leaves given three different treatments of E. coli solution were used for spectra measurements and total E. coli determination. Classification of lettuce samples based on different concentration of E. coli solution was performed by Partial least Squares Discriminant Analysis (PLS-DA) with different pre-processing methods into the two different groups of “safe” and “unsafe” samples. The best model was recommended using standard normal variation (SNV) + second derivate (D2) pre-processing methods with the minimum standard deviation of cross-validation (SECV = 0.256). Besides, a good correlation (R2 of 0.93) between Vis/NIR spectral data and the presence of E. coli contamination proved the possibility of Vis/NIR spectroscopy for microbial detection and evaluation in lettuce.
Keywords: Visible/Near-infrared (Vis/NIR) spectroscopy; pre-processing; Optical; Food safety.