Detection of fish fillet substitution and mislabeling using multimode hyperspectral imaging techniques

In 2011, poly- and perfluoroalkyl substances (PFASs) were analyzed in surface water and fish fillet samples taken from Pool 2 of the Upper Mississippi River, a 33-mile stretch inclusive of the Minneapolis/St. Paul, Minnesota (USA) metropolitan area. Approximately 100 each of bluegill, freshwater drum, smallmouth bass, and white bass were sampled within the study area. Surface water samples were also collected from each of the 10 sampling reaches established for the study. Water and fillet samples were analyzed for perfluorinated carboxylic acids (C4-C12), perfluorinated sulfonic acids (C4, C6, and C8), and perfluorooctane sulfonamide. Perfluorooctane sulfonate (PFOS) was observed with the greatest frequency in fish fillets and ranged from 3.0 to 760 ng/g wet weight. Mean (geometric) PFOS concentrations in bluegill, freshwater drum, smallmouth bass, and white bass were 20, 28, 29, and 58 ng/g wet weight, respectively. When compared with fish data collected in 2009, a significant reduction (p < 0.05) in PFOS concentrations was noted. This finding was confirmed based on data from studies conducted in 2012 and 2013. Overall, between 2009 and 2013, PFOS concentrations decreased by 65, 76, and 50% for bluegill, freshwater drum, and white bass, respectively (44% decrease for smallmouth bass from 2009 to 2012). These declines in fish PFOS concentrations are consistent with ongoing efforts to effectively control sources of PFASs to the Mississippi River. Environ Toxicol Chem 2017;36:3138-3147. © 2017 SETAC.

The aim of this work is to investigate olive oil by-product [vegetation water (VW)] inclusion in rainbow trout diet and its effect on the productive traits and the quality of the product. Two levels of VW inclusion were used and one control group was included. Fish diets were isonitrogeonous (crude protein 40%) and isoenergetic (18 MJ kg−1 DM). Two thousand and four hundred rainbow trout were used. An in vivo digestibility experiment was performed in order to determine diets’ digestibility. All the fish diets and fillets were analysed to determine the proximate and fatty acid composition. On final fish fillet, lipid oxidation was determined at 0, 24, 48, 72, 96 and 192 h of storage using the 2-thiobarbituric acid method (thiobarbituric acid reactive substances). Aroma analyses on the final cooked and raw fillet were performed using an electronic nose. The VW inclusion partially reduced protein digestibility. The fish growth varied between 1.08% and 1.1% day. The supplemental level of VW led to a better antioxidant status of fish fillet, in particular, in the fillet sample after 72 h of fillet conservation. Principal component analysis in raw and cooked fish fillet indicates that the VW inclusion in the fish diet led to aroma modification on fish fillet.

Prediction of low heavy metal concentrations in agricultural soils using visible and near-infrared reflectance spectroscopy

Dietary influence of replacing fish meal and oil with canola protein concentrate and vegetable oils on growth performance, fatty acid composition and organochlorine residues in rainbow trout ( Oncorhynchus mykiss)

The aim of this study is to assess the possibilities of the VNIR (Visible and Near InfraRed) and SWIR (Short Wavelength InfraRed) satellite data for estimating intra-plot patterns of soil electrical resistivity consistent with ground measurements. The methodology is based on optical reflectances that constitute the input variables of random forest, alone or in combination with parameters derived from a digital elevation model (DEM). Over a field located in southwestern France, the results show high level of accuracy for the 0–50 and 0–100 cm soil layers (with R² of 0.69 and 0.59, and a relative RMSE of 18% and 16%, respectively), the performances being lower for the 0–170 cm layer (R² of 0.39, relative RMSE of 20%). The combined use of optical reflectances with parameters derived from the DEM slightly improves the performances, whatever the considered layer. The influence of each reflectance on soil electrical resistivity estimates is finally analyzed, showing that the wavelengths acquired in the SWIR have a relative higher importance than VNIR reflectance.

Assessing the size selectivity of capture gears for reef fishes using paired stereo-baited remote underwater video

Relative to standard red/green/blue (RGB) imaging systems, hyperspectral imaging systems offer superior capabilities but tend to be expensive and complex, requiring either a mechanically complex push-broom line scanning method, a tunable filter, or a large set of light emitting diodes (LEDs) to collect images in multiple wavelengths. This paper proposes a new methodology to support the design of a hypothesized system that uses three imaging modes—fluorescence, visible/near-infrared (VNIR) reflectance, and shortwave infrared (SWIR) reflectance—to capture narrow-band spectral data at only three to seven narrow wavelengths. Simulated annealing is applied to identify the optimal wavelengths for sparse spectral measurement with a cost function based on the accuracy provided by a weighted k-nearest neighbors (WKNN) classifier, a common and relatively robust machine learning classifier. Two separate classification approaches are presented, the first using a multi-layer perceptron (MLP) artificial neural network trained on sparse data from the three individual spectra and the second using a fusion of the data from all three spectra. The results are compared with those from four alternative classifiers based on common machine learning algorithms. To validate the proposed methodology, reflectance and fluorescence spectra in these three spectroscopic modes were collected from fish fillets and used to classify the fillets by species. Accuracies determined from the two classification approaches are compared with benchmark values derived by training the classifiers with the full resolution spectral data. The results of the single-layer classification study show accuracies ranging from ~68% for SWIR reflectance to ~90% for fluorescence with just seven wavelengths. The results of the fusion classification study show accuracies of about 95% with seven wavelengths and more than 90% even with just three wavelengths. Reducing the number of required wavelengths facilitates the creation of rapid and cost-effective spectral imaging systems that can be used for widespread analysis in food monitoring/food fraud, agricultural, and biomedical applications.

Per- and polyfluoroalkyl substances (PFAS) are a class of thousands of man-made chemicals that are persistent and highly stable in the environment. Fish consumption has been identified as a key route of PFAS exposure for humans. However, routine fish monitoring targets only a handful of PFAS, and non-targeted analyses have largely only evaluated fish from heavily PFAS-impacted waters. Here, we evaluated PFAS in fish fillets from recreational and drinking water sources in central North Carolina to assess whether PFAS are present in these fillets that would not be detected by conventional targeted methods. We used liquid chromatography, ion mobility spectrometry, and mass spectrometry (LC-IMS-MS) to collect full scan feature data, performed suspect screening using an in-house library of 100 PFAS for high confidence feature identification, searched for additional PFAS features using non-targeted data analyses, and quantified perfluorooctanesulfonic acid (PFOS) in the fillet samples. A total of 36 PFAS were detected in the fish fillets, including 19 that would not be detected using common targeted methods, with a minimum of 6 and a maximum of 22 in individual fish. Median fillet PFOS levels were concerningly high at 11.6 to 42.3 ppb, and no significant correlation between PFOS levels and number of PFAS per fish was observed. Future PFAS monitoring in this region should target more of these 36 PFAS, and other regions not considered heavily PFAS contaminated should consider incorporating non-targeted analyses into ongoing fish monitoring studies.

Abstract. In this paper, visible and near infrared (Vis-NIR) spectral technique and chemometrics were investigated for the fast and non-destructive determination of soil available nitrogen content. The Vis-NIR diffuse reflectance spectra of 2 mm dried soil surface was collected by USB2000+ spectrometer, and the available nitrogen of soil was detected by alkaline hydrolysis diffusion method. Three different pretreatment methods were used to process the raw spectral data, including Savitzky-Golay (SG) smoothing, multiplicative scatter correction (MSC), standardized normal variable (SNV), and then the preprocessed spectral data was combined with partial least squares regression (PLSR) method to establish the prediction model. In addition, the precision of prediction model was evaluated by the correlation coefficient and the standard deviation of the model. The results showed that the optimal pretreatment method was Savitzky-Golay (SG), the highest precision model was established by Savitzky-Golay (SG) smoothing pretreatment method combined with PLSR, and the correlation coefficient of the validation set (Rv) was 0.8389, the validation standard deviation (SEP) was 9.6564. Therefore, it is feasible to apply Vis-NIR spectral technique in fast and non-destructive determination of soil available nitrogen content, which provides a theoretical basis for further development in real-time and in-situ detection equipment for soil available nitrogen content.

Abstract. The objective of this work was to evaluate the potential of Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER), subsystems VNIR (Visible and Near Infrared) and SWIR (Short Wave Infrared) images, for discrimination and mapping of amethyst mining residues (basalt) in the Ametista do Sul Region, Rio Grande do Sul State, Brazil. This region provides the most part of amethyst mining of the World. The basalt is extracted during the mining process and deposited outside the mine. As a result, mounts of residues (basalt) rise up. These mounts are many times smaller than ASTER pixel size (VNIR – 15 meters and SWIR – 30 meters). Thus, the pixel composition becomes a mixing of various materials, hampering its identification and mapping. Trying to solve this problem, multispectral algorithm Maximum Likelihood (MaxVer) and the hyperspectral technique SAM (Spectral Angle Mapper) were used in this work. Images from ASTER subsystems VNIR and SWIR were used to perform the classifications. SAM technique produced better results than MaxVer algorithm. The main error found by the techniques was the mixing between "shadow" and "mining residues/basalt" classes. With the SAM technique the confusion decreased because it employed the basalt spectral curve as a reference, while the multispectral techniques employed pixels groups that could have spectral mixture with other targets. The results showed that in tropical terrains as the study area, ASTER data can be efficacious for the characterization of mining residues.

The influence of vegetation cover on the spectroscopic estimation of soil properties

PurposeThis work aimed to improve breast screening program accuracy using automated classification. The goal was to determine if whole image features represented in the discrete cosine transform would provide a basis for classification. Priority was placed on avoiding false negative findings.MethodsOnline datasets were used for this work. No informed consent was required. Programs were developed in Mathematica and, where necessary to improve computational performance ported to C++. The use of a discrete cosine transform to separate normal from cancerous breast tissue was tested. Features (moments of the mean) were calculated in square sections of the transform centered on the origin. K-nearest neighbor and naive Bayesian classifiers were tested.ResultsForty-one features were generated and tested singly, and in combination of two or three. Using a k-nearest neighbor classifier, sensitivities as high as 98% with a specificity of 66% were achieved. With a naive Bayesian classifier, sensitivities as high as 100% were achieved with a specificity of 64%.ConclusionWhole image classification based on discrete cosine transform (DCT) features was effectively implemented with a high level of sensitivity and specificity achieved. The high sensitivity attained using the DCT generated feature set implied that these classifiers could be used in series with other methods to increase specificity. Using a classifier with near 100% sensitivity, such as the one developed in this project, before applying a second classifier could only boost the accuracy of that classifier.

Non-destructive prediction of the moisture content of individual wheat kernels combining hyperspectral imaging and WGAN data augmentation algorithm.

The timely estimation of growth and photosynthetic-related traits in an easy and nondestructive manner using hyperspectral data will become imperative for addressing the challenges of environmental stresses inherent to the agricultural sector in arid conditions. However, the handling and analysis of these data by exploiting the full spectrum remains the determining factor for refining the estimation of crop variables. The main objective of this study was to estimate growth and traits underpinning photosynthetic efficiency of two wheat cultivars grown under simulated saline field conditions and exposed to three salinity levels using hyperspectral reflectance information from 350–2500 nm obtained at two years. Partial least squares regression (PLSR) based on the full spectrum was applied to develop predictive models for estimating the measured parameters in different conditions (salinity levels, cultivars, and years). Variable importance in projection (VIP) of PLSR in combination with multiple linear regression (MLR) was implemented to identify important waveband regions and influential wavelengths related to the measured parameters. The results showed that the PLSR models exhibited moderate to high coefficients of determination (R2) in both the calibration and validation datasets (0.30–0.95), but that this range of R2 values depended on parameters and conditions. The PLSR models based on the full spectrum accurately and robustly predicted three of four parameters across all conditions. Based on the combination of PLSR-VIP and MLR analysis, the wavelengths selected within the visible (VIS), red-edge, and middle near-infrared (NIR) wavebands were the most sensitive to all parameters in all conditions, whereas those selected within the shortwave infrared (SWIR) waveband were effective for some parameters in particular conditions. Overall, these results indicated that the PLSR analysis and band selection techniques can offer a rapid and nondestructive alternative approach to accurately estimate growth- and photosynthetic-related trait responses to salinity stress.

Rapid Assessment of Quality Parameters for Frozen Cod Using Near Infrared Spectroscopy