An official website of the United States government.
Ensuring a safe, secure, and high quality food supply is of critical importance to us. Intelligent quality sensors refer to devices that are robust, adaptable, accurate and can be used to provide critical quality information about food and agricultural products. To be practical, these sensors need to be cost-effective, non-destructive, and portable in order to be used with food products. Thus, a multi-disciplinary team of scientists from North Dakota State University is involved in ongoing research with the long-term goal of developing miniaturized portable sensors to provide quality information to users about specific food and agricultural products. We hypothesize that the volatile chemicals/gases generated by fungal and bacterial metabolism of food products could be used as indicators of food quality and safety. Over the last several years, we have been developing or adapting electronic nose systems using detectors based on metal oxide, composite polymers, mixed metal oxide or thin films of metal oxide as food safety/quality sensors.
NON-TECHNICAL SUMMARY: Ensuring a safe, secure, and high quality food supply is of critical importance to us. We propose development of intelligent quality sensors to provide critical quality information about food and agricultural products. <P>
APPROACH: This current research project focuses on the application and evaluation of infrared sensing (by means of a portable FTIR spectrometer) for quality/safety evaluation of selected food products. Parallel work will involve further development of pattern recognition techniques for different electronic nose modules and their implementation in embedded systems. Exploratory investigations will be conducted for additional characterization of thin films of tin oxide and nano-particles of metal oxide (tin-oxide) for their utility in sensing specific gaseous compounds associated with selected food products.
<P>
PROGRESS: 2003/06 TO 2006/05<BR>
The long- term goal is to develop miniaturized portable sensors that can provide quality information to users about specific food and agricultural products. Specific emphasis was given to assess the capabilities of IR-based olfactory sensing for evaluating the condition of the packaged beef for spoilage and contamination with Salmonella. A Fourier-Transform infrared spectrometer (FT-IR) spectrometer was purchased from a commercial source. An IR-based olfactory sensing system module was developed using the FTIR spectrometer and other accessories, such as a pump, computer, and valves etc. This developed system extracts the gas from the headspace of a meat package and generates the FTIR spectrum of the compounds present. Experiments were conducted to evaluate the performance of the FT-IR-based olfactory sensing system for discriminating a givenspoilage in packaged beef for its spoilage. A meat sample was considered spoiled if the bacterial count was > = 6 log10(cfu/g). Algorithms and techniques were developed to process the acquired FT-IR spectraum of headspace gas of from meat samples. Statistical models were developed to classify a given meat sample asinto spoiled or not. A maximum total average accuracy of 89 % was obtained for classifying meat samples into two groups (spoiled or not spoiled) for samples stored at 50 degrees F. Experiments were also conducted for classifying packaged meat samples for Salmonella contamination. Vacuum packaged beef stored at 20 degree C were used. A meat sample was considered to be Salmonella contaminated if the Salmonella count was equal to or more than 0.7 log10 (cfu/g). The overall maximum accuracy for classifying a given sample into either contaminated or not, was 88.7% and the statistical model used all the peak information in five different selected ranges between 500- 400 wave numbers. Vacuum packaged beef stored at 20 degree C were used. The same modeling technique and storage temperature provided a maximum overall accuracy of 86.5% when fresh beef samples were used. These findings shows potential of using FTIR-based olfactory sensing technique for classification of packaged meat samples for spoilage and Salmonella contamination using headspace gases. Additional validation is recommended. Parallel studies were conducted for developing a theoretical framework for a potential NDIR (non-dispersive infrared) sensor for meat contamination and spoilage. Simulation of gas flow was done using a commercial MEMS (micro-electromechanical-systems) software program. Parallel study was conducted on evaluating FT-IR-based olfactory sensing to discriminate Fusarium infected barley from the uncontaminated ones. Statistical models with bootstrapping technique provided a maximum overall accuracy of 89%. Additional studies were also conducted in evaluating different pattern recognition techniques such as wavelet transformation and genetic programming. Studies related to Trypsin inhibitor activities of soymilk were conducted. The Langumuir-Blodgett thin films of n-tetraphenyl porphine manganese (III) chloride mixed with Arachidic acid were evaluated for their sensing behavior towards gaseous compounds associated with meat spoilage and contamination. For ethanol and acetic acid, red and blue shifts were observed, respectively. Additional studies were also conducted in evaluating different pattern recognition techniques such as wavelet transformation and Genetic programming.
<P>
IMPACT: 2003/06 TO 2006/05<BR>
Development of robust and reliable sensors for food safety applications is a critical need. Our research has proven the concept of our hypothesis that olfactory sensing of headspace of packaged meat could distinguishclassify Salmonella contaminated product from non-contaminated ones. Once validated further with larger datasets,. T this technique has potential to be used as a first- line alert system. Our technique has advantagess into reducinge or eliminatinge the problem of sampling associated with other sensing techniques. Thus, it has potential to be used atin different critical points in the food supply chain (for selected food products) as first- line alert system.