An official website of the United States government.
<OL> <LI> Develop a hyperspectral imaging system to predict meat quality. Initial focus will be on beef tenderness prediction. a. Develop and calibrate a hyperspectral imaging system b. Develop and validate multivariate models to predict shear force tenderness scores c. Test and validate the models to predict sensory tenderness scores <LI>Integrate heat transfer and dynamic microbial models to describe the growth of pathogens in food systems during processing. Initial focus will be on Salmonella Enteritidis in shell eggs during cooling. a. Develop and validate a predictive model to describe growth of SE in shell eggs b. Develop and validate a heat transfer model for cooling of shell eggs c. Integrate microbial and heat transfer models and validate the integrated model to predict temperatures of shell eggs and consequent growth of SE in shell eggs. <LI>Develop near-infrared spectroscopy to detect yolk contamination in egg white. a. Develop methods for quantitatively measuring yolk contamination in liquid egg white by near-infrared spectroscopy. b. Develop and validate chemometric models using various spectral pretreatments and calibration techniques. <LI> Develop spectral imaging techniques to detect intentional contamination in food a. Develop a spectral and imaging database of target food materials, potential chemicals, and contaminated food materials. b. Develop and validate multivariate models to detect the presence of chemical contaminants.
Non-Technical Summary: Improving food quality and safety is essential to enhance competitiveness of our agricultural and food industry in the global market. After the September 11 incident, food defense has been a major concern. This project develops (i)an instrument for cattle producers and processors to assess beef product quality to meet consumer expectations that would result in enhanced economic opportunities, (ii)a predictive modeling tool for the egg processing industry to implement Hazard Analysis and Critical Control Points (HACCP) and to evaluate the adequacy of risk reduction strategies, (iii)an instrument for egg processors to detect yolk contamination in egg white, and (iv)a spectral imaging system for food industries and inspectors from regulatory agencies to screen food products for contamination due to bioterrorism. <P> Approach: 1. Hyperspectral Imaging to Predict Meat Quality: A hyperspectral imaging system consisting of a CCD digital video camera and a spectrograph was developed and calibrated. A lighting chamber provided diffuse lighting. A large image database (n=1000) of hyperspectral images of beef steaks with varying tenderness scores will be created. After image acquisition, slice shear force values will be collected as the tenderness reference for model development. Image features will be extracted and models will be developed to predict tenderness scores. Finally, the developed models will be validated to predict sensory tenderness scores. 2. Predictive Models to Describe Growth of Pathogens in Food Systems: Growth of five strains of Salmonella Enteritidis (SE) in egg yolk will collected at ten different growth temperatures (7 to 45 degrees C). A minimum of two replications will be performed for each temperature. The Baranyi model will be used to describe the growth of SE as a function of time at constant environmental conditions. The growth rate from the Baranyi model then will be modeled as a function of temperature. Both models will be integrated and solved numerically to describe microbial growth under constantly varying temperature conditions. A heat transfer model will be developed using finite element method to accurately estimate the temperature distribution of eggs at various locations on the pallet. The center point temperature will be predicted during the cooling period for various air temperature and speed. This center point temperature will be used in dynamic microbial modeling. Methods described by Amezquita (2004) will be used to integrate the heat transfer model and the microbiological predictive model. Scenarios with egg cooling parameters will be developed and the output with egg cooling temperature histories will be validated. 3. Near-infrared Spectroscopy to Detect Yolk Contamination in Egg White: Samples for calibration development and validation will be prepared by mixing appropriate amounts of the yolk (0.00 - 0.30 percent) into the liquid egg white. Near-infrared transmission spectra will be collected from the calibration and validation samples using a NIR spectrometer. Chemometric models will be developed and validated. The NIR models will also be applied to commercial egg white samples, and the NIR results will be compared with yolk contamination levels measured by the method of Bergquist and Wells (1956). 4. Spectral Imaging to Detect Intentional Contamination in Food: Of the numerous foods that the U.S. consumer utilizes annually, products that serve niche roles within our economy would likely be the major targets for terrorists. We will use easily accessible chemicals and potential chemical agents listed by the CDC (http://www.bt.cdc.gov/) as contaminants. Spectral profiles will be collected used in visible, near-infrared, and mid-infrared regions. Multivariate data analysis will be developed and evaluated by validating with new samples. The detection limit of each contaminant in various food matrices will be determined.