A simple, sensitive and rapid chemiluminescent fiber optic biosensor utilizing monoclonal antibodies to S. aureus was developed to detect the pathogen in food. The S. aureus cells were selectively labeled with a monoclonal-horseradish peroxidase (POD) conjugate, collected by membrane filtration, and detected with a luminometer and an enhanced chemiluminescent luminol reagent. Two different diameter membranes, 25 mm and 13 mm, were first tested in a luminometer tube format assay. A hand operated syringe filtration unit was used to capture cells and the membrane was then transferred to a luminometer tube for the chemiluminescent reaction. An improved system utilized a simple but efficient microwell plate vacuum filtration unit with an 8 mm membrane sealed at the bottom of the sample well. The sample was concentrated on the membrane and positioned directly in front of a fiber optic light guide to effectively collect and transmit the signal to the luminometer. Labeling S. aureus in solution proved to be much more effective than on the membrane surface. Using the microwell plate filtration system resulted in less sample handling, better reproducibility, and dramatically reduced assay time. The variability for 25 mm and 13 mm assays were 24.7% and 13.3%, while the microwell plate assay reduced this to 4.0%. The ability of the fiber optic probe to effectively collect the signal meant the sensitivity of the assay was not compromised with smaller membrane and sample size. The sensitivity of the biosensor was 3.8 X 10⁴ CFU/ml, adequate to detect the organism at concentrations lower than the level that could result in food poisoning. The performance of the biosensor was not effected by the food materials and by the presence of other bacteria.

A competitive exclusion (CE) culture of chicken cecal anaerobes has been developed and used in this laboratory for control of Salmonella typhimurium in chickens. The CE culture consists of 29 different species of micro-organisms, and is known as CF3. Detection of one of the CF3 bacteria, Eubacteria, and S. typhimurium were demonstrated using a commercial immunomagnetic (IM) electrochemiluminescence (ECL) sensor, the ORIGEN^R Analyzer. Analysis was achieved using a sandwich immunoassay. Bacteria were captured on antibody- conjugated 280 micron sized magnetic beads followed by binding of reporter antibodies labelled with ruthenium (II) tris(dipyridyl) chelate [Ru(bpy)₃²⁺]. The magnetic beads were then trapped on an electrode in the reaction cell of the ORIGEN^R Analyzer by a magnet, and the ECL was evoked from Ru(bpy)₃²⁺ on the tagged reporter antibodies by an electrical potential at the electrode. Preliminary IM-ECL assays with Eubacteria yielded a detection limit of 10⁵ cfu/mL. Preliminary IM-ECL assays with S. typhimurium yielded a similar detection limit of 10⁵ cfu/mL.

A panel of species specific monoclonal antibodies were raised to Campylobacter coli, Campylobacter jejuni and Campylobacter lari. The isotypes, and cross-reactivity profiles of each monoclonal antibody against an extensive panel of micro- organisms, were determined.

Live cells of E. coliO157:H7 were captured by goat anti-E. coliO157 serum coated on the surface of polystyrene based immunomagnetic beads (IMB). The captured bacteria were labeled by 4',6-diamidino-2-phenylindole (DAPI), a nucleic acid stain, for observation by epifluorescent microscopy. The beads with captured bacteria were then concentrated by magnetic separators. The efficiency of this magnetic concentration step was less than that of using high speed centrifugation. The antibody-captured and IMB-immobilized bacteria were then applied on HF-treated, bovine serum albumin (BSA)-coated microscope slides mounted on an automated stage, and magnetically aligned before fluorescence distribution was measured by a cooled CCD attached to an inverted microscope. Since the beads were concentrated and linearly aligned along the edge of the magnetic field, image capture along the edge for a few field widths was sufficient to account for most of captured bacteria. We applied this approach to determine the bacterial counts in spiked beef hamburger patties. The results showed that after a 6-hour enrichment, sufficient number of the bacteria could be detected from the samples spiked with 1 CFU of E. coliO157:H7 per gram of the hamburger.

Microsphere-based immunoassays were devised for compounds of agricultural and biomedical interest (e.g., digoxin, theophylline, and zearalenone). Commercially available microspheres with surface functional groups for chemical derivatization were used as solid carriers. After immobilizing the target substances, the surface of the haptenized microspheres was blocked by a protein to reduce aspecific binding. Competitive immunoassays were performed using the functionalized microspheres and antibodies labeled with horseradish peroxidase. Immunofluorescence signal amplification was achieved by enzyme-catalyzed reporter deposition (CARD). An epifluorescence microscope, a CCD camera interfaced with a computer, and microscopy image analysis software were employed for quantitative detection of fluorescent light emitted from individual microspheres. Integration of several such immunoassays and application of an optical encoding method enabled multianalyte determination. These immunoassays can also be utilized in an immunosensor array format. This immunoarray format could facilitate miniaturization and automation of multianalyte immunoassays.

The purpose of this work is to develop a rapid, automated system for nucleic acid purification and concentration from environmental and food processing samples. Our current approach involves off-line filtration and cell lysis (ballistic disintegration) functions in appropriate buffers followed by automated nucleic acid capture and purification on renewable affinity matrix microcolumns. Physical cell lysis and renewable affinity microcolumns eliminate the need for toxic organic solvents, enzyme digestions or other time- consuming sample manipulations. Within the renewable affinity microcolumn, we have examined nucleic acid capture and purification efficiency with various microbead matrices (glass, polymer, paramagnetic), surface derivitization (sequence-specific capture oligonucleotides or peptide nucleic acids), and DNA target size and concentration under variable solution conditions and temperatures. Results will be presented comparing automated system performance relative to benchtop procedures for both clean (pure DNA from a laboratory culture) and environmental (soil extract) samples, including results which demonstrate 8 minute purification and elution of low-copy nucleic acid targets from a crude soil extract in a form suitable for PCR or microarray-based detectors. Future research will involve the development of improved affinity reagents and complete system integration, including upstream cell concentration and cell lysis functions and downstream, gene-based detectors. Results of this research will ultimately lead to improved processes and instrumentation for on-line, automated monitors for pathogenic micro-organisms in food, water, air, and soil samples.

An automated system for filtration capture and immunoelectrochemical detection of bactria in liquid samples is described. The detector incorporates a porous electrode in contact with the filter, rather than the solid electrode used previously, to allow sample and reagent solutions to be delivered in a flowing stream. This eliminated the need for manual assembly of the electrode and filter for each assay and allowed repetitive assays on a single filter/electrode. The electrochemical response of the novel gold grid electrode under static and flow conditions was found to be consistent with theory for a planar electrode operating in laminar flow conditions. A computer-controlled fluid handling system was coupled to the detector for delivery of samples and reagents at controlled flow rates and times. The combination of flow detector and fluid handling system allows for automation of the previous assay protocol as well as providing new operating modes with enhanced background rejection and improved sensitivity. The use of these operating modes is demonstrated by a simple assay for Escherichia coli O157:H7 with virtually no background current.

Recent foodborne outbreaks have emphasized that microbes are evolving strategies that overcome our traditional processing and preservation techniques. Conventional treatments to decontaminate products containing human pathogens are largely ineffective, and new interventions are needed. Such innovative technologies are needed to assure the production and processing of high quality, fresh food items that are less likely to support pathogen growth. Novel chemical and physical treatments are needed that will reduce the risk of microbial contamination while not adversely affecting the quality of the products. This paper presents Natick Lab's efforts in exploring non-traditional methods such as electrolyzed oxidizing water, high intensity light, modified atmospheres and microwave and irradiation technologies, non-traditional chemical treatments including novel sanitizing solutions and natural antimicrobial agents, and non-traditional biological treatments such as the use of bacteriocins. Microbial tests showed significant antifungal, antibacterial effects of these methods, individually and synergistically, with minimum deterioration of food quality as measured by the sensory evaluations. The methods are useful for both military and civilian applications.

Escherichia coli O157:H7 outbreaks were mostly due to consumption of undercooked contaminated beef which resulted in severe illness and several fatalities. Recalls of contaminated meat are costly for the meat industry. Our research attempts to understand the mechanisms of bacterial adhesion on animal carcass in order to eliminate or reduce pathogens in foods. We have reported the interactions of immobilized E. coli O157:H7 cells with extracellular matrix (ECM) components using a surface plasmon resonance biosensor (BIAcore). These studies showed that immobilized bacterial cells allowed the study of real-time binding interactions of bacterial surface with the ECM compounds, collagen I, laminin and fibronectin. Collagen I and laminin bound to the E. coli sensor surface with dissociation and association rates ranging from 10⁶ to 10⁹. Binding of collagen I and laminin mixture resulted in synergistic binding signals. An inhibition model was derived using collagen-laminin as the ligand which binds with E. coli sensor. A select group of naturally occurring food additives was evaluated by determining their effectivity in inhibiting the collagen-laminin binding to the bacterial sensor. Bound collagen-laminin was detached from the E. coli sensor surface with the aid of an organic acid. The biosensor results were verified with cell aggregation assays which were observed with optical and electron microscopes. These biosensor studies provided understanding of bacterial adhesion to connective tissue macromolecules. It also provided a model system for the rapid assessment of potential inhibitors that can be used in carcass treatment to inhibit or reduce bacterial contamination.

A neuro-fuzzy based image classification system that utilizes color-imaging features of poultry viscera in the spectral and spatial domains was developed in this study. Poultry viscera of liver and heart were separated into four classes: normal, airsacculitis, cadaver, and septicemia. Color images for the classified poultry viscera were collected in the poultry process plant. These images in RGB color space were segmented and statistical analysis was performed for feature selection. The neuro-fuzzy system utilizes hybrid paradigms of fuzzy interference system and neural networks to enhance the robustness of the classification processes. The results showed that the accuracy for separation of normal from abnormal livers were 87.5 to 92.5% when two classes of validation data were used. For two-class classification of chicken hearts, the accuracies were 92.5 to 97.5%. When neuro-fuzzy models were employed to separate chicken livers into three classes (normal, airsacculitis, and cadaver), the accuracy was 88.3% for the training data and 83.3% for the validation data. Combining features of chicken liver and heart, a generalized neuro-fuzzy model was designed to classify poultry viscera into four classes (normal, airsacculitis, cadaver, and septicemia). The classification accuracy of 86.3% was achieved for the training data and 82.5% accuracy for the validation.

Development of effective food inspection systems is critical in successful implementation of the hazard analysis and critical control points (HACCP) program. Hyperspectral imaging or imaging spectroscopy, which combines techniques of imaging and spectroscopy to acquire spatial and spectral information simultaneously, has great potential in food quality and safety inspection. This paper reviewed the basic principle and features of hyperspectral imaging and its hardware and software implementation. The potential areas of application for hyperspectral imaging in food quality and safety inspection were identified and its limitations were discussed. A hyperspectral imaging system developed for research in food quality and safety inspection was described. Experiments were performed to acquire hyperspectral images from four classes of poultry carcasses: normal, cadaver, septicemia, and tumor. Noticeable differences in the spectra of the relative reflectance and its second difference in the wavelengths between 430 nm and 900 nm were observed between wholesome and unwholesome carcasses. Differences among the three classes of unwholesome carcasses were also observed from their respective spectra. These results showed that hyperspectral imaging can be an effective tool for safety inspection of poultry carcasses.

Optical detection of fecal contamination on poultry carcasses is addressed in this paper. Specifically, a hyperspectral imaging system with a spectral range in the visible and near- infrared is used to capture imagery of a poultry carcass. A number of image processing routines are investigated for their effectiveness and efficiency at detecting feces. Based on the need for speed, the more simplistic mathematical operations seem to hold the most promise for detecting contaminated carcasses in real-time.

The size of a poultry spleen is an indication of whether the bird is wholesomeness or has a virus-related disease. This study explored the possibility of detecting poultry spleen enlargement with a computer imaging system to assist human inspectors in food safety inspections. Images of 45-day-old hybrid turkey internal viscera were taken using fluorescent and UV lighting systems. Image processing algorithms including linear transformation, morphological operations, and statistical analyses were developed to distinguish the spleen from its surroundings and then to detect abnormal spleens. Experimental results demonstrated that the imaging method could effectively distinguish spleens from other organ and intestine. Based on a total sample of 57 birds, the classification rates were 92% from a self-test set, and 95% from an independent test set for the correct detection of normal and abnormal birds. The methodology indicated the feasibility of using automated machine vision systems in the future to inspect internal organs and check the wholesomeness of poultry carcasses.

The Instrumentation and Sensing Laboratory (ISL) has developed an industrial prototype diode-array visible/near-infrared (Vis/NIR) spectrophotometer system for inspecting poultry for diseased and defective carcasses on-line. The ISL design is based on the principle that wholesome and diseased and defective birds have different chemical compositions of tissues and may have different skin color. This visible/near- infrared spectrophotometer system has been tested off-line at 60 and 90 birds per minute. On-line trials of the visible/near-infrared chicken carcass inspection system were conducted during an 8-day period in a slaughter plant in New Holland, Pennsylvania, where spectra (470 - 960 nm) of 1174 normal and 576 abnormal (diseased and/or defective) chicken carcasses were measured. The instrument measured the spectra of veterinarian-selected carcasses as they passed on a processing line at a speed of 70 birds per minute. Classification models using principal component analysis as a data pretreatment for input into neural networks were able to classify the carcasses from the spectral data with a success rate of 95%. Data from 3 days can predict the subsequent two days' chickens with high accuracy. This accuracy was consistent with the results obtained previously on off-line studies. Thus, the method shows promise for separation of diseased and defective carcasses from wholesome carcasses in a partially automated inspection system. Details of the models using various training regimens are discussed.

A multispectral imaging system with selected optical filters of 542 and 700-nm wavelength was shown feasible for detecting contaminated poultry carcasses with high accuracy. The analysis of textural features based on co-occurrence matrix (COM) was conducted to determine the performance of multispectral image analyses in discriminating unwholesome poultry carcasses from wholesome carcasses. The variance, sum average, sum variance, and sum entropy of COM were the most significant texture features (P less than 0.0005) to identify unwholesome poultry carcasses. The feature values of angular second moment, variance, sum average, sum variance, and sum entropy did not vary with the distances and directions of COM for the spectral images. When a direction was equal to 0 degrees, the contrast was lower and the inverse difference moment and difference variance were higher (P less than 0.01) than any other direction in the visible spectral images. The characteristics of variance and sum variance texture feature of spectral images varied with the wavelength of spectral images and unwholesomeness of poultry carcasses as well. The sum variance of wholesome was higher (P less than 0.005) than unwholesome carcasses at the spectral image of 542-nm. The linear discriminant model was able to identify wholesome carcasses with classification accuracy of 83.9 percent and the unwholesome carcasses could be identified by quadratic model with 97.1 percent accuracy when textural features of spectral image at 700-nm wavelength were used as input data for models.

From 1988 through 1992 Salmonellosis accounted for 27% of the total reported foodborne disease outbreaks and 57% of the outbreaks in which the pathogen was identified. The prevalence of Salmonellosis and the new requirements to monitor the organism as a marker in pathogen reduction programs will drive the need for rapid, on-site testing. A compact fiber optic fluorometer using a red diode laser as an excitation source and fiber probes for analyte detection has been constructed and used to measure Salmonella. The organisms were isolated with anti-Salmonella magnetic beads and were labeled with a secondary antibody conjugated to a red fluorescent dye. The response of the system was proportional to the concentration of Salmonella typhimurium from 3.2 X 10⁵ colony forming units (CFU)/ml to 1.6 X 10⁷ CFU/ml. The system was developed to utilize a fiber-optic magnetic focusing problem that attracted the magnetic microspheres to the surface of a sample chamber directly in front of the excitation and emission fibers. The signal obtained from a homogenous suspension of fluorescent magnetic microspheres was 9 to 10 picowatts. After focusing, the signal from the fluorescent labeled magnetic microspheres increased to 200 picowatts, approximately 20 times greater than the homogeneous suspension. The magnetic focusing assay detected 1.59 X 10⁵ colony forming units/ml of Salmonella typhimurium cultured in growth media. The process of magnetic focusing in front of the fibers has the potential to reduce the background fluorescence from unbound secondary antibodies, eliminating several rinsing steps, resulting in a simple rapid assay.

Using starch ampicillin agar, qualitative and quantitative determinations of Aeromonas spp. were made at several sites during swine carcass dressing and cutting. Aeromonas spp. were observed at all sites surveyed. Levels increased during shackling and passage through the first and middle polisher/washers, and significantly decreased during the singeing steps. Passage through the final polisher/washer caused a small increase in levels in Aeromonas spp. and these levels then remained constant during the rest of the carcass dressing operation. Aeromonas spp. were also isolated from the room where the carcasses were cut into wholesale cuts and cuts for further processing. Presumptive Aeromonas spp. cultures isolated from the different sites were confirmed as belonging to the genus Aeromonas and then speciated using the biochemical scheme of Joseph and Carnahan; 81% of the cultures were identified at A. hydrophila. Since most isolates were A. hydrophila, determination of the origin of isolates from different sites in the processing plant must await utilizing molecular biotyping techniques on the cultures. These results indicate the Aeromonas spp. occurs extensively in the swine dressing environment and thus represents a possible public health hazard and potential spoilage concern. Changes in cleaning and sanitizing of equipment may be necessary during swine carcass dressing and cutting to guard against this pathogen.