Submitted to: Meeting Abstract
Publication Type: Abstract Only
Publication Acceptance Date: April 4, 2014
Publication Date: May 7, 2014
Citation: Park, B., Eady, M.B., Choi, S., Hinton Jr, A., Yoon, S.C., Lawrence, K.C. 2014. Rapid identification of salmonella serotypes with stereo and hyperspectral microscope imaging Methods. IAFP's 2014 European Symposium on Food Safety. Meeting Abstract [abstract]. Technical Abstract: The hyperspectral microscope imaging (HMI) method can reduce detection time within 8 hours including incubation process. The early and rapid detection with this method in conjunction with the high throughput capabilities makes HMI method a prime candidate for implementation for the food industry. This research was conducted to determine if the spectral signature at 24 hours incubation time can be compared to the spectra of earlier time frames from 8 to 12 hours of incubation, and to determine if the five Salmonella serotypes could be differentiated by their spectral signatures. HMI spectral data from five Salmonella serotypes (Enteritidis, Heidelberg, Infants, Kentucky, and Typhimurium) at various incubation times from 8 to 24 hours were analyzed. Bacterial colonies were picked from agar plates with a stereo microscope. A total of 89 contiguous images were acquired between 450-800 nm. Preprocessing of the spectral data was performed by a global data transformation algorithm, and followed by a principle component analysis (PCA). The Mahalanobis distance (MD) was calculated from PCA score plots for analyzing cluster of serotypes. Partial least squares regression (PLSR) was used for calibration and prediction of the model, while soft independent modeling of class analogy (SIMCA) was used for classification of serotype clusters. Pearson correlation values indicated spectral patterns for varying incubation times ranging from 0.993 to 0.999. PCA score plots showed cluster separation with average MD for incubation times ranging from 1.116 to 52.937. PLSR had a maximum RMSEC value of 0.084 and RMSEV value of 0.089. SIMCA classification values were above 98.1%, and validation values above 97.7%. The early and rapid detection abilities could identify contaminated products before being released to the public marketplace and causing widespread disease. The HMI with analytical methods can be used for quality assurance for in-house product safety assessments.