Skip to main content
ARS Home » Southeast Area » Athens, Georgia » U.S. National Poultry Research Center » Quality and Safety Assessment Research Unit » Research » Publications at this Location » Publication #282264

Title: Classification of shiga toxin-producing escherichia coli (STEC) serotypes with hyperspectral microscope imagery

Author
item Park, Bosoon
item Windham, William
item LADELY, SCOTT - Food Safety Inspection Service (FSIS)
item GURRAM, PRUDHVI - Us Army Research
item KWON, HEESUNG - Us Army Research
item Yoon, Seung-Chul
item Lawrence, Kurt
item Narang, Neelam
item Cray Jr, William

Submitted to: Proceedings of SPIE
Publication Type: Proceedings
Publication Acceptance Date: 4/23/2012
Publication Date: N/A
Citation: N/A

Interpretive Summary: Shiga toxin-producing Escherichia coli (STEC) are a type of enterohemorrhagic E. coli (EHEC) bacteria that could cause illness ranging from mild intestinal disease to severe kidney complications. Other types of EHEC include the relatively important serotype E. coli O157:H7, and more than 100 other non-O157 strains such as O26, O45, O103, O111, O121 and O145. Those serotypes have been recognized recently in serious outbreak's that cause human illness due to their toxicity. Although a conventional microbiological method for cell counting is still accurate, rapid methods for foodborne pathogen detection are needed for better performance. As one optical detection method, hyperspectral microscopic imaging can be an effective tool for identifying pathogenic bacteria because of its capability to characterize bacterial cells from micro colony samples, which could reduce incubation time needed for cell grow. The objective of this research is to develop a hyperspectral microscopic imaging method to evaluate spectral characteristics of foodborne pathogen specifically STEC. In this research, the acousto-optic tunable filters (AOTF)-based hyperspectral microscope imaging method was used to identify STEC serotypes with classification algorithms including support vector machine (SVM) and sparse kernel-based ensemble learning (SKEL) were presented.

Technical Abstract: Non-O157:H7 Shiga toxin-producing Escherichia coli (STEC) strains such as O26, O45, O103, O111, O121 and O145 are recognized as serious outbreak to cause human illness due to their toxicity. Since a conventional microbiological method for cell counting is laborious and time-consuming process, optical detection method with hyperspectral microscope imagery was developed for real-time, in-situ foodborne pathogen detection. In this research, acousto-optical tunable filters (AOTF)-based hyperspectral microscopic imaging (HMI) platform was developed for identifying pathogenic bacteria because of its capability to differentiate both spatial and spectral characteristics of each bacterial cell from microcolony samples. Using the AOTF-based HMI method, a total of 89 contiguous spectral images could be acquired within approximately 30 seconds with 250 ms exposure time. From this study, we have successfully developed the protocol for live-cell immobilization on glass slides to acquire quality spectral images from STEC bacterial cells using the modified dry method. Among the contiguous spectral imagery between 450 and 800 nm, the intensity of spectral images at 458, 498, 522, 546, 570, 586, 670 and 690 nm were distinctive for STEC bacteria. With a SVM algorithm, the highest accuracy (92%) was obtained from serotype O45, followed by O26 (89%), O145 (84%), and O111 (72%), respectively. The classification accuracies, however, were very low from O103 (57%) and O121 (16%). Similar classification accuracies were obtained with a SKEL algorithm.