Paul Heinemann

Modeling the inactivation of Salmonella Typhimurium, Listeria monocytogenes, and Salmonella Enteritidis on poultry products exposed to pulsed UV-light

Pulsed UV light inactivation of Salmonella Typhimurium on unpackaged and vacuum-packaged chicken breast, Listeria monocytogenes on unpackaged and vacuum-packaged chicken frankfurters, and Salmonella Enteritidis on shell eggs was explained by log-linear and Weibull models using inactivation data from previous studies. This study demonstrated that the survival curves of Salmonella Typhimurium and L. monocytogenes were nonlinear exhibiting concavity. The Weibull model was more successful than the log-linear model in estimating the inactivations for all poultry products evaluated, except for Salmonella Enteritidis on shell eggs, for which the survival curve was sigmoidal rather than concave, and the use of the Weibull model resulted in slightly better fit than the log-linear model. The analyses for the goodness of fit and performance of the Weibull model produced root mean square errors of 0.059 to 0.824, percent root mean square errors of 3.105 to 21.182, determination coefficients of 0.747 to 0.989, slopes of 0.842 to 1.042, bias factor values of 0.505 to 1.309, and accuracy factor values of 1.263 to 6.874. Overall, this study suggests that the survival curves of pathogens on poultry products exposed to pulsed UV light are nonlinear and that the Weibull model may generally be a useful tool to describe the inactivation patterns for pathogenic microorganisms affiliated with poultry products.

Detection of Apple Deterioration Using an Electronic Nose and zNosetm

Damage in apples can cause fruit spoilage, reduce commodity economic value, and give rise to food quality and safety concerns. This research investigated use of electronic nose (Enose, Cyranose 320) and zNoseTM -based nondestructive protocols for rapid detection of deterioration in apples. Key compounds associated with apple aroma were identified using gas chromatography and mass spectrometry, and the differences were observed after 6 days exposure to artificially induced damage in the form of a cut. High-dimensional data were compressed by principal component analysis (PCA) and partial least squares (PLS). Linear discriminant analysis (LDA) and canonical variate analysis (CVA) models were developed based on the compressed data. Experiments showed that both the Enose and zNose were able to effectively detect the volatile differences between undamaged and damaged apples four or more days after the cut. Differences in number of cuts had some effect on volatile compound emissions. Apples subjected to two cuts and three cuts generated volatile profiles that were significantly different from uncut apples. Varying the orientation of cut apples did not give significant differences in the volatile profile. The PLS-LDA model produced the best correct classification rates: 96% using the zNose, and 85% using the Enose.

NON-COMPOSTED GRAIN-BASED SUBSTRATES FOR MUSHROOM PRODUCTION (AGARICUS BISPORUS)

Non-composted grain-based substrates were evaluated for the cultivation of mushrooms (Agaricus bisporus) with the goal of eliminating the need for the lengthy and often malodorous composting process. Millet grain, millet grain mixed with soybean, and commercial rye grain spawn were used as substrates. Treatments included different proportions (100%, 75%, 50%, and 25%) of millet grain, or grain spawn with perlite as an inert bulking material. For the millet and soybean mixtures, the biological matter (millet and soybean) was set at 75% by volume, while the ratios of millet to soybean were varied. To induce fructification, all substrates were overlain with a sterilized mixture of peat and calcium carbonate (casing) containing 25% activated charcoal (v/v), which was shown to be as effective as a commercial non-sterile casing. Among the various treatments, the highest mushroom yield among all three grain treatments was recorded for the 100% millet/0% perlite treatment (8.7 kg/m2), which was comparable to that of compost (7.7 kg/m2). In contrast, the millet/soybean mixtures failed to produce any mushrooms when soybean was added to the substrate. The highest recorded mushroom yield for commercial grain spawn was 5.3 kg/m2 for the 100% spawn/0% perlite treatment. This yield was lower when compared to the millet grain and compost substrates. However, biological efficiency (fresh weight of mushroom divided by dry weight of substrates × 100) was 117% for the 25% spawn/75% perlite treatment, while that of compost and the 75% millet/25% perlite treatment were 98% and 55%, respectively. The results suggest that mushrooms can be grown in a non-composted substrate, but further economic analysis will need to be performed to determine economic viability of alternative substrates.

MODELING OF COMPOST TEMPERATURE AND INACTIVATION OF SALMONELLA AND E. COLI O157:H7 DURING WINDROW FOOD WASTE COMPOSTING

A simulation model was developed to predict temperature and inactivation of E. coli O157:H7 and Salmonella during windrow composting. In particular, the model included an energy balance to estimate the change in temperature based on heat generated by biological decomposition and heat losses by convection, conduction, evaporation, and radiation. The model was validated with the measured data for the effects of seasonal variation on compost temperature and pathogen reduction. Sensitivity analysis was performed on the model to evaluate the variations in both seasons (winter and summer) and moisture contents (40% to 80%). The model showed the highest variation between experimental and predicted data only in winter composts. The results suggested that moisture content of 40% to 60% was appropriate for summer and 40% to <60% for winter composting. Higher moisture levels did not demonstrate pathogen inactivation during winter conditions, whereas it took a month to eliminate the pathogens in summer according to the model predictions. Overall, the model was promising for evaluation of the composting process for different conditions. Further research is needed to improve the model predictions using measured process parameters under different environmental conditions.

Detection of Apple Defects Using an Electronic Nose and zNose

Apple defects and spoilage not only reduce commodity economic value, but cause food safety concerns as well. It is essential for fruit quality assurance and safety to rapidly detect fruit physical damage and spoilage. This article presents the application of an electronic nose (Cyranose 320) and zNose to the development of a nondestructive, rapid and cost effective system for the detection of defects of apples. The key compounds associated with apple aroma were identified and the “smellprints” of these key compounds were established by the electronic nose and zNose. Healthy and damaged apples were kept in 2L glass jars for 6 hours for preconcentration before measuring. Principal Component Analysis (PCA) models were developed based on the Enose and zNose data. Maholanobis distance was applied for discriminant analysis. Experiments showed that the Enose and zNose are both capable of detecting the volatile differences between healthy apples and damaged apples. After five days deterioration, the correct classification rate for the Enose was 83.3%, and for the zNose was 100%. After seven days, the correct classification rate was 100% for both instruments. For the next stage, a non-linear model and sensor fusion technique will be developed.

Non-destructive apple bruise on-line test and classification with Raman spectroscopy

A prototype automated inspection system was developed to classify apples based on bruising in real time. Raman spectroscopy was used for apple bruise testing. Fifty ‘Delicious’ apples were randomly divided into two groups, 25 apples for light bruise group (falling height 50 mm), and 25 for heavy bruise (falling height 300 mm); each group was divided into a modeling set and a testing set. A Nicolet FT-Raman Spectroscope was employed to obtain apple spectra. The Spectrascope utilized OMNIC E. S. P. 5.1 software. The unbruised and bruised spectra were analyzed and classified by WinDAS using canonical variate analysis (CVA) and principle component analysis (PCA) models, on both the training and testing sets. The PCA and CVA model analysis satisfactorily classified the apples by bruise. Therefore, the results show that the Raman spectroscope permits non-destructive bruise determination with good results.

AUTOMATED MICROPROPAGATED SUGARCANE SHOOT SEPARATION BY MACHINE VISION

A prototype vision-guided separation mechanism for Stage 2 micropropagated sugarcane shoots was developed and tested. Micropropagated sugarcane plantlets grown between parallel plates showed two-dimensional structure which greatly simplified shoot identification using machine vision. Two identification methods were developed to locate the shoot positions in the shoot clump image. The shoot locations identified by the vision system were used to guide a pair of stepper motor-driven x-y tables to separate and transplant shoots. Machine vision algorithms and computer control algorithms were developed, integrated, and tested in the prototype sugarcane shoot separation system. The best combination of identification algorithm and separation method resulted in 85% of total shoots successfully separated. A set of shoots separated by the vision-guided system and a second set manually separated were compared in continued culture. A total of 76% of shoots survived that were separated by the automated system versus 83% for manual separation.

EVALUATION OF THE EFFECTS OF FORCED AERATION DURING PHASE I MUSHROOM SUBSTRATE PREPARATION: PART 1. MODEL DEVELOPMENT

The purpose of this research was to evaluate the effects of forced aeration during Phase I mushroom substrate preparation on substrate temperature, oxygen concentration, and heat generation. The work is described in two parts: (1) model development and thermal property determination, and (2) air temperature, oxygen, and airflow measurements and heat generation calculations using the model. This article represents part 1. To determine the rate of heat generation during substrate preparation, an energy balance model was developed for an environmentally controlled Phase I system. Thermal properties of mushroom substrate in the Phase I process were measured. Thermal conductivity ranged from 0.19 to 0.49 W/m-K. Thermal diffusivity ranged from 0.20 to 0.31 × 10-6 m2/s. Using an average bulk density of 275 kg/m3, the specific heat of Phase I mushroom substrate was calculated to be 5.23 kJ/kg-K.

GENETIC ALGORITHMS (GAS) AND EVOLUTIONARY STRATEGY TO OPTIMIZE ELECTRONIC NOSE SENSOR SELECTION

The high dimensionality of electronic nose data increases the difficulty of their use in classification models. Reducing this high dimensionality helps reduce variable numbers, potentially improve classification accuracy by removing irrelevant sensors, and reduce computation time and sensor cost. In this research, the Cyranose 320 electronic nose was optimized for apple defect detection by selecting the most relevant of its 32 internal sensors using various selection methods. The contribution of each sensor was first evaluated statistically by calculating the F-value. By keeping only the top 90% cumulative F-values, 25 sensors were selected and the classification error rate was 25.4%. Sequential forward/backward search methods reduced the minimum classification error rate to 6.1%. Two more heuristic optimization algorithms, genetic algorithm (GA) and the covariance matrix adaptation evolutionary strategy (CMAES), were applied and compared. Although both algorithms gave a best classification error rate of 4.4%, the average classification error rate of CMAES over 30 random seed runs was 5.0% (SD = 0.006), which was better than the 5.2% (SD = 0.004) from the GA. The final optimal solution sets obtained by using an integer GA showed that including more sensors did not guarantee better classification performance. The best reduction in classification error rate was 10%, while the number of sensors was reduced by 75%. This study provided a robust and efficient optimization approach to reduce the high dimensionality of electronic nose data, which substantially improved electronic nose performance in apple defect detection while potentially reducing the overall electronic nose cost for future specific applications.

Agaricus bisporus Grain Spawn Substrate with S41 and S44 Nutrient Supplements

A substrate predominantly composed of mushroom grain spawn and nutrient supplements was tested for Agaricus mushroom productivity with the goal of eliminating composting from mushroom production. Different rates of S41 and S44 mushroom nutrient supplements mixed with grain spawn were used (20%, 15%, 10%, 5%, 1%, and 0%). Two casing treatment were tested for mushroom yield; a non-sterile and a sterile casing with 25% activated carbon. Thiophanate methyl was used in one set of treatments to control the incidence of fungal contamination, whereas another set of treatments had no thiophanate methyl. The results show that mushroom yield from a nonsterile casing overlain on a 10% S41 grain spawn substrate was statistically comparable(p<0.05) to the same substrate with a sterile casing with 25% activated carbon, 6.4 kg/m2 and 8.6 kg/m2, respectively. Adding thiophanate methyl lowered the yield for both S41 and S44 (p<0.05). The S41 supplement on average produced greater mushroom yield compared to S44 (p<0.04). When perlite was added to the substrate as a water-holding material, the yield was increased greatly, producing 13kg/m2 compared to 7.5 kg/m2 for the treatment without perlite (p<0.05).