Location: Quality & Safety Assessment Research2013 Annual Report
1a. Objectives (from AD-416):
1. Develop methods and instruments to identify food safety hazards throughout various stages of poultry and egg production and processing. 2. Detect and characterize foodborne pathogens, toxins, and bacterial threat agents with rapid optical methods. 3. Develop and evaluate detection methods for foodborne pathogens and toxins with nanotechnology.
1b. Approach (from AD-416):
Various optical (imaging) methods will be used for detecting intentional and unintentional contaminants and bacterial pathogens of food products. A real-time in-line hyperspectral imaging system will be used to rapidly detect diseased and contaminated broiler carcasses in processing plants. A monochromatic imaging system will be used for detecting cracks in table and hatching eggs. Hyperspectral microscopic imaging system, Raman imaging instrument and Fourier transform infrared spectrometer will be used to detect biofilms and foreign materials on the surfaces of food processing equipment. Visible/near-infrared hyperspectral imaging systems will be used to rapidly detect and characterize foodborne pathogens associated with poultry products and bacterial threat agents. Nanotechnology will be used for detection of foodborne pathogens and toxins. Collaboration with ARS Environmental Microbial and Food Safety Laboratory, BARC, FSIS, AMS, and the University of Georgia Nano Science and Engineering Center will be used to enhance the research.
3. Progress Report:
Differentiating Non-0157 STEC Colonies from Background Microflora in Ground Beef by Hyperspectral Imaging (HSI). Detection and recovery of non-O157 Shiga-toxin producing Escherichia coli (STEC) in high background ground beef is difficult due to the competing microflora and the lack of a suitable selective and differential agar media. A HSI technique was developed to detect and differentiate each of the “big-six” non-O157 serogroups from background microflora. The results provide the location of the suspect target STEC colony on the agar plate such that an analyst can pick the colony for further conformational tests. Relationship of Eggshell Microstructure to Chemical Composition. Low eggshell strength leads to development of cracks, allowing contamination of eggs by pathogenic bacteria and loss to the industry. A Laser Induced Breakdown Spectroscopy (LIBS) method was developed to correlate the chemical composition of the eggshell with microstructural features that enhance the propensity of eggshells to become damaged. Eggshell hardness was related to the ratio of calcium components detected by the LIBS system. Classification of Salmonella serotypes with Laser Induced Breakdown Spectroscopy (LIBS). A LIBS method was developed to identify five Salmonella serotypes based on the atomic spectral signatures emitted from the laser induced plasma. Colony films of the different serotypes were created on stainless steel and the LIBS spectra of the resulting films were subsequently treated via a hierarchical clustering method which successfully identified each serotype. Classification of Foodborne Pathogens with Hyperspectral Microscope Imaging (HMI). A HMI method was developed to classify foodborne pathogens from the optical properties of bacterial cells in micro colony samples using their spectral signatures. The HMI method was able to differentiate five serotypes of Salmonella and Staphylococcus with a support vector machine classification algorithm. The results can be used for a rapid presumptive screening method for foodborne pathogens. Salmonella Typhimurium (ST) Identification with DNA aptamers using Surface Enhanced Raman Scattering (SERS). A special substrate was fabricated and tested with SERS to produce a label-free, specific detection of ST from poultry carcass rinses. The substrate uses a silver nanorod array that have special binding reagents, known as aptamers, to capture the ST cells which are then detected by the SERS system. Further optimization for binding conditions and fabrication of device is needed for improving the lower limit of detection. Nanobiosensor for Ricin Detection with Atomic Force Microscopy (AFM) and Surface Plasmon Resonance (SPR). Nucleic acid aptamers was used as binding reagents for the label-free detection of biomolecules. The specific interactions between ricin and its corresponding aptamer were measured with AFM and SPR spectrometry. The specific binding site on the aptamer molecule was clearly recognized by the AFM recognition images. Aptamers can be integrated into various detection platforms and provide versatile applications for fast detection and analysis related to biosafety and biosecurity.
1. Hyperspectral imaging for detection of foreign materials. Hyperspectral machine vision sensor to detect foreign materials in specialty crops and to remove them by a robot was jointly developed under a cooperative agreement. Foreign material detection methods and computer algorithms, including their source codes, were developed and transferred to industry. Also, the application software and robot control software were developed and transferred. Preliminary tests of product moving on a conveyor were successful. The technology was further developed with a more universal platform which can be used in real-time poultry inspection applications and other material detection applications in the future.
2. Methods and device for early and rapid detection of foodborne pathogens on agar media. A hyperspectral microscope imaging (HMI) technique was developed for early and rapid detection of Salmonella serotypes ( Enteritidis, Heidelberg, Infantis, Kentucky, and Typhimurium) at incubation times of 8, 10, 12, and 24 hrs on brilliant green sulfa (BGS) agar plates. Pearson correlation values indicate virtually identical spectral patterns for varying incubation times and advanced multivariate statistical analysis techniques show good separation at all incubation times. Additionally, a soft independent modeling of class analogy classification method also produced very high classification values. Thus, Salmonella serotypes can be identified by applying these techniques to HMI data, regardless of incubation time, which can speed up detection time by reducing the time needed for incubation.
3. Automated colony counting with hyperspectral imaging. Computer algorithms were developed to automatically detect and count bacterial colonies growing on agar plates. The software uses image segmentation algorithms to detect non-O157 Shiga toxin-producing Escherichia coli (STEC) serogroups (commonly known as the “big six”), Salmonella Enteritidis and Typhimurium, and background microflora on agar plates. An algorithm to rapidly count cells from microscope images that fluoresce was also developed.
Yoon, S.C., Windham, W.R., Ladely, S.R., Heitschmidt, G.W., Lawrence, K.C., Park, B., Narang, N., Cray Jr, W.C. 2013. Hyperspectral imaging for differentiating colonies of non-O157 shiga-toxin producing echerichia coli (STEC) serogroups on spread plates of pure cultures. Journal of Near Infrared Spectroscopy. 21(2):81-95.