This PDF file contains the front matter associated with SPIE Proceedings Volume 12120, including the Title Page, Copyright information, Table of Contents and Conference Committed lists.

Aflatoxins are among the most carcinogenic mycotoxins and are known to contaminate a wide variety of agricultural and food commodities. This study aims to explore the effectiveness of Raman hyperspectral imaging in detecting aflatoxin contamination in corn kernels in a rapid and non-destructive manner. Four hundred kernels were used with 2 treatments, namely, 200 kernels inoculated with the AF13 fungus (aflatoxigenic), and 200 kernels inoculated with sterile distilled water as control. One hundred kernels from each treatment were incubated at 30 °C for 5 and 8 days, separately. On the specified post-inoculation day, the kernels were dried and wiped free of surface mold prior to imaging. The Raman images of kernels were acquired over the endosperm side over the 103-2831 cm-1 wavenumber range. The standard aflatoxin concentration in each kernel was determined by the VICAM AflaTest method. The original mean spectra of single kernels were extracted and preprocessed by adaptive iteratively reweighted penalized least squares, Savitzky Golay smoothing and min-max normalization. On basis of the calculated “reference” mean spectra of the aflatoxin negative and -positive categories, 14 and 17 local peaks were determined, separately. After removing the identical peaks from both peak sets, a total of 24 unique peaks were extracted and used as inputs for further discriminant model development. With 20 ppb and 100 ppb as the classification thresholds, the 2-class discriminant models established with the principal component analysis-linear discriminant analysis and partial least-squares discriminant analysis methods, obtained mean overall prediction accuracies between 77.9% and 82.0%. Further investigation is ongoing to include more diverse samples and execute different types of computation algorithms, seeking solutions to improve the discriminant models in identifying aflatoxin contamination in corn kernels.

Despite developments in the dairy industry, blowing defects in cheese due to Clostridium contamination continues to be a problem. Traditional microbiological detection methods used in dairy industries to detect spores are time consuming and limited in efficiency and sensitivity. Many alternative approaches of detecting butyric acid clostridia in milk still pose a major challenge for microbiologists. According to the fermentative butyric acid production, carbon dioxide and hydrogen are also produced as byproducts during fermentation. As a consequence, during the incubation of contaminated milk in a sealed container (such as a cuvette), the measurement of CO₂ and H₂ in the headspace can be a promising possibility for early-stage detection of the spores. A multi-gas sensor based on Raman spectroscopy was used for the measurement of CO₂ and H₂ concentration. The aim of this paper is to demonstrate the applicability of Raman gas analysis to the detection of multiple gas species in sealed, transparent samples for early detection of spore contamination in milk. Two different contaminant spores were used: Clostridium tyrobutyricum and Bacillus. The first one showed an increase of CO₂ and H₂ content within 16h of incubation. The second one showed an increase of CO₂ content within 48 h of incubation. For the first time, Raman gas spectroscopy has been applied for contactless measurement of the time resolved evolution of multiple gaseous species in milk samples for spore analysis.

This paper presents development of an automated hyperspectral system to monitor plant health for produce production in NASA’s space missions. The system was designed to inspect salad crops using both reflectance and fluorescence imaging in a spectral region of 400–1000 nm. Major hardware components of the system include a compact line-scan hyperspectral camera, two LED line lights providing broadband and UV-A light for reflectance and fluorescence measurement, and a linear translation stage. In a sensor-to-sample arrangement, the translation stage moves the camera and the lights over the plants to acquire a pair of matched reflectance and fluorescence images in a line-scan imaging cycle. Control software was developed using LabVIEW to realize hardware parameterization and data transfer functions. The imaging system was installed in a growth chamber at NASA Kennedy Space Center for plant health monitoring studies. A demonstration experiment was conducted to detect drought stress for Dragoon lettuce. Two‐band reflectance ratio images at 690 and 702 nm were able to detect the drought stress on lettuce leaves without visible symptoms.

We developed a hyperspectral elastic light scatter (ELS) phenotyping instrument to explore the relationship between the wavelength of the incident beam and the elastic light-scatter patterns of a bacterial colony, and, ultimately, to enhance the classification efficiency of non-invasive ELS-based systems employed in microbiology. The new instrument consists of a supercontinuum (SC) laser and acousto-optic tunable filter (AOTF), which enables the selection of the wavelength of interest allowing multiple spectral patterns in a single measurement. An initial experiment with microflora found on green leafy vegetables derived an encouraging result, showing over 90% of average classification accuracy when classifying colonies from two different bacterial species utilizing 70 spectral bands from SC-laser. This observation suggested the notion that colonies of varied species may form distinguishable scatter features at separate different spectral bands. The increase in the number of employed bands consequently led to the rise in the number of scatter pattern features, potentially resulting in the classifier's overfitting and decrease in real-life accuracy. Therefore, the presented hyperspectral ELS system employs feature reduction and selection procedures to enhance the robustness and ultimately lessen the complexity of data collection.

We developed convolutional neural network (CNN) models using drone pictures to estimate vineyard leaf area index (LAI) and percent canopy cover. These parameters are traditionally measured using hand-held devices (e.g., AccuPAR for LAI and Ceptometer for percent canopy cover) and calculated manually, which is labor intensive and hard to apply to large-scale areas. We collected airborne images or videos by flying a low-altitude drone with a built-in digital camera over a large-scale vineyard. The airborne images convey all necessary information for developing CNN models. To date, we have collected data from the same vineyard over a couple of years. The ground truth values were manually measured using AccuPAR and Ceptometer at the same time of airborne imaging. Specifically, we trained five CNN models to estimate percent canopy cover and leaf area index (LAI). The estimated results over a large vineyard will help guide planting intercrops or cover crops to prevent soil erosion, calculating the correct amounts of expected residue in fall, and foliar sprays of pesticides and fungicides, and characterization of vegetation-atmosphere interactions.

Plants have sensory receptors through which they sense and respond to environmental cues. This work focuses on correlating the electrical responses of living plants with variations in environmental conditions. A pair of sharp metal probes were attached to the stem of a bell pepper plant to monitor its time-domain voltage characteristics. The plant was subjected to heat and pressure stimuli, water stress, and humidity variations. For every 50-degree centigrade increment in leaf temperature, there was an up to 3-fold enhancement in the measured voltage. A burst of voltage spikes was also observed when the leaf was pressed with a finger. Spikes in the measured voltage were observed in less than 5 seconds after exposing the plant to heat/pressure stimulus. The voltage spiked at the beginning of the 7-day stress cycle for a water stressed plant but subsided after the first day. The electrical responses of the plant were also investigated under three different humidity levels. The measured signals were analyzed through Morlet, Gabor, Bump, Dubecheis, and Haar wavelets to identify the dominant frequency components for each stress condition. In addition, autocorrelation and cross correlation values were computed to analyze the similarity between the measured time-series signals. The knowledge gained from this project will help engineering plants that are hypersensitive to environmental cues. These super-sensitive plants can be used as an early warning system in the event of a wildfire, toxic gas leakage, water/nutrient shortage, or pest attack. Our findings support that probing the electrical responses of plants will enable local climate monitoring and allow the growers to intervene immediately to reduce crop losses resulting from environmental stresses.

Infection with foodborne pathogens such as Salmonella spp. is of high risk for people with a weakened immune system. Microbiological culture method has been used in general for detection of pathogens from the food matrix; however, it is time consuming and requires experience and good level of laboratory skills. In the food safety field, various techniques which allows the rapid and simple detection have been developed at the level of a user-friendly tool for detecting the foodborne pathogens. Quartz crystal microbalance (QCM) are mass-based biosensor which measures the microgram level mass changes, enabling a user to observe the presence of the pathogen simply and rapidly. When the pathogens are bound on vibrating quartz surface, the resonant frequency of a quartz crystal will be changed due to the mass of the pathogens. In this study, the QCM detected killed Salmonella Typhimurium in the range of 〖10〗^5-〖10〗^9 CFU/mL, correlating to the averaged frequency shifts. The actual concentrations of Salmonella from the culture method were compared to the difference in the resonant frequency. The QCM sensor were treated with 11-Mercaptoundecanoic acid (11-MUDA), and EDC-NHS following by antibodies and bovine serum albumin (BSA) to utilize the antibody-antigen reaction. With a usage of peristaltic pump, the solutions could be introduced to the surface while frequencies could be monitored for each step in real-time. To acquire the evidence of Salmonella, the surfaces of the quartz crystal with the fluoresce labeled antibody were captured by the fluorescence microscope. The QCM biosensor showed the possibility of detection of Salmonella in less time, compared with the conventional method.

Salicylic acid (SA) is a vital regulator of induced defense mechanisms in plants. Progressive variations in the SA levels have been reported in many droughts, salt, and cold/heat-stressed plants, as well as after pathogen/herbivore attacks. Hence, real-time and in situ monitoring of this phytohormone will facilitate the early identification of crop stresses so that immediate interventions can be implemented to mitigate productivity losses and maintain the quality of the product (e.g., quality of produce and cotton fibers). However, the lack of in situ and localized analysis capabilities in current technologies limits their use in crop fields. Toward this endeavor, this work reports an LSPR (localized surface plasmon resonance)- based fiber-optic sensor functionalized with a conjugate of gold nanoparticles and copper-based metal-organic framework (CuMOF) for selectively measuring SA levels in sap. For this purpose, the distal end of an optical fiber was coated with gold nanoparticles that exhibited LSPR upon excitation by light’s electromagnetic field. The gold nanoparticles-coated fiber was covalently functionalized with CuMOF, which selectively oxidized SA. This oxidation reaction resulted in a variation in the local refractive index, thereby leading to a change in the LSPR intensity level. The fiber-tip sensor exhibited a sensitivity of up to 0.0117 % light reflection variations per μM concentration of SA and a limit of detection of 37 μM. Plant sap was used to calibrate the sensor for precision measurements. The fiber-tip sensor was also demonstrated to measure the SA levels in the stem of a live plant. Considering excellent dynamic detection range (100-1000 μM) and sensitive thiol chemistry, the developed sensor could hold promises in future in situ probe development for real-time measurements of phytohormones in sap.

A portable bacterial colony classification tool based on colonies’ reflective elastic light-scatter (ELS) patterns has been developed using a smartphone, a green laser, and a projection screen material. As the collimated beam from the laser illuminates the bacterial colony, backscattered photons interfere and generate a unique pattern on the screen material determined by the unique morphology of the colony. The phone camera, which is located behind the screen, captures the pattern. The collected patterns are utilized to extract the distinctive scatter-related features across different organisms for the classification process. Unlike other tools that use transmitted ELS patterns, the novel device measures the reflective signal, and therefore this ELS technique can be applied to organisms that are grown on opaque media such as blood agar, chocolate agar, which normally prohibits the transmission of the light and generation of forward ELS patterns. The adaptation of the smartphone camera as an imaging device dramatically reduced the system to a palm-size instrument. This made it wholly portable and easy to carry. For validation of the instrument, two different bacteria species, E. coli and L. innocua were grown on opaque agar media and tested. The results showed over 90% of overall accuracy in differentiating the organisms.

Fungal species such as Aspergillus, Fusarium and Alternaria can contaminate agricultural commodities in the field or during storage and produce mycotoxins. They usually pose threats to human and animal health and can result in significant economic loss. Specifically, Fusarium graminearum, the major causative agent of Fusarium head blight (FHB) of small cereals produces mycotoxins including deoxynivalenol, nivalenol, and zearalenone. Conventional detection methods are time-consuming, expensive and require large-scale instruments and skilled technicians. Furthermore, detection of the toxins in post-harvested grain is a process that can only be accomplished after the grain is harvested. Therefore, our goal was to develop a molecular point-of-detection (POD) platform which was sensitive and specific to detect low levels of toxin-producing fungi within agricultural products in the field and could also be used directly in food products. Herein, we investigated a rapid molecular POD assay called loop mediated isothermal amplification (LAMP) to detect low levels of genomic DNA extracted from Fusarium graminearum, which is often associated with toxicological potential and food safety issues. Both fluorescent and colorimetric LAMP assays were characterized and optimized to detect low-level of pathogens within 70 and 50 minutes respectively. In summary, LAMP offers an efficient assay format for rapid and specific nucleic acid-based detection of mycotoxins in-field use. Coupled with our custom-designed microchip, our platform provides a proof-of-principle to achieve low-cost and widespread foodborne pathogens testing at the POD which is highly desirable to keep analysis time and costs low, but more importantly be a field use application.

Due to its ease of sample preparation and rapid processing speed, laser-induced breakdown spectroscopy (LIBS) has emerged as a promising new technique for food analysis. Food adulteration detection is critical for fair trade and protecting customers from food fraud. As a result, there is a high demand for a rapid and portable detection method for authenticating and evaluating the safety of marketed food and beverage products. An increased prevalence of food fraud which frequently entails the substitution of inferior ingredients for high-quality products has necessitated the development of innovative measures for detecting and preventing fraud. In this report, we describe an authentication approach utilizing a custom designed benchtop LIBS system. We focused on high-value regional food products such as European alpine-style cheeses and Italian balsamic vinegars. Liquid samples were measured on paper without any pretreatment, and solid samples were ablated directly on the sample surface by LIBS. The pre-processed LIBS spectra were utilized for training and validating various classifiers for sample categorization and validation. The development of an elastic net (ENET) classification model is also reported in the study. In summary, our research highlighted the potential of the LIBS technique combined with chemometric methods for solid and liquid high-value food authenticity certification. The results show that LIBS enables rapid analysis and accurate food sample classification without the requirement for sample pretreatment.

Rapid detection and disinfection of microbial contamination is an ongoing concern across various food processing industries. Numerous methods exist for detection, including nucleic acid-based, fluorescence microscopy, and immunological-based tests. There is an emerging interest in using optical techniques to perform detection and disinfection simultaneously. This study reports on experimental results of energy density effects on disinfection of gram-negative organisms using a commercially available portable device. Three different gram-negative organisms were cultured and diluted over a four-log range. Samples of different concentrations were plated and exposed to UVC with increasing energy densities. A summary of the disinfection rate is presented. We identified an appropriate energy density condition that was required depending upon the concentration and type of microorganisms. The results showed that the tested portable device could serve be a valuable alternative for in-field screening and disinfection.

Emission of greenhouse gases such as CO₂ is highly dependent on the energy systems of the countries. In this regard, for accurate analysis of CO₂ emission it is necessary to take into account the factors related to energy consumption. In the present paper, three countries including Turkey, Bulgaria and Greece are considered as case studies to model their CO₂ by using an economic indicator (GDP) and consumptions of different energy sources. In this regard, Group Method of Data Handling (GMDH) is applied as the method and the required data for modeling between 2000 and 2019 are gathered. The results indicated that R² values of the proposed model for training, test and overall datasets are 0.9997, 0.9991 and 0.9995, respectively. In addition, AARD of the mentioned datasets were around 0.71%, 1.18% and 0.85%, respectively. These values reveal significant exactness of the proposed which can be attributed to proper selection of both inputs and modeling method.