Exploring the identification of multiple bacteria on stainless steel using multi-scale spectral imaging from microscopic to macroscopic.

This work investigates non-contact reflectance spectral imaging techniques, i.e. microscopic Fourier transform infrared (FTIR) imaging, macroscopic visible-near infrared (VNIR), and shortwave infrared (SWIR) spectral imaging, for the identification of bacteria on stainless steel. Spectral images of two Gram-positive (GP) bacteria (Bacillus subtilis (BS) and Lactobacillus plantarum (LP)), and three Gram-negative (GN) bacteria (Escherichia coli (EC), Cronobacter sakazakii (CS), and Pseudomonas fluorescens (PF)), were collected from dried suspensions of bacterial cells dropped onto stainless steel surfaces. Through the use of multiple independent biological replicates for model validation and testing, FTIR reflectance spectral imaging was found to provide excellent GP/GN classification accuracy (> 96%), while the fused VNIR-SWIR data yielded classification accuracy exceeding 80% when applied to the independent test sets. However, classification within gram type was far less reliable, with lower accuracies for classification within the GP (< 75%) and GN (≤ 51%) species when calibration models were applied to the independent test sets, underlining the importance of independent model validation when dealing with samples of high biological variability.

Characterisation and Classification of Foodborne Bacteria Using Reflectance FTIR Microscopic Imaging.

This work investigates the application of reflectance Fourier transform infrared (FTIR) microscopic imaging for rapid, and non-invasive detection and classification between Bacillus subtilis and Escherichia coli cell suspensions dried onto metallic substrates (stainless steel (STS) and aluminium (Al) slides) in the optical density (OD) concentration range of 0.001 to 10. Results showed that reflectance FTIR of samples with OD lower than 0.1 did not present an acceptable spectral signal to enable classification. Two modelling strategies were devised to evaluate model performance, transferability and consistency among concentration levels. Modelling strategy 1 involves training the model with half of the sample set, consisting of all concentrations, and applying it to the remaining half. Using this approach, for the STS substrate, the best model was achieved using support vector machine (SVM) classification, providing an accuracy of 96% and Matthews correlation coefficient (MCC) of 0.93 for the independent test set. For the Al substrate, the best SVM model produced an accuracy and MCC of 91% and 0.82, respectively. Furthermore, the aforementioned best model built from one substrate was transferred to predict the bacterial samples deposited on the other substrate. Results revealed an acceptable predictive ability when transferring the STS model to samples on Al (accuracy = 82%). However, the Al model could not be adapted to bacterial samples deposited on STS (accuracy = 57%). For modelling strategy 2, models were developed using one concentration level and tested on the other concentrations for each substrate. Results proved that models built from samples with moderate (1 OD) concentration can be adapted to other concentrations with good model generalization. Prediction maps revealed the heterogeneous distribution of biomolecules due to the coffee ring effect. This work demonstrated the feasibility of applying FTIR to characterise spectroscopic fingerprints of dry bacterial cells on substrates of relevance for food processing.

Spectral imaging for characterization and detection of plastic substances in branded teabags.

The addition of plastic substances in teabags is of increasing concern for conscious consumers due to the harmful effects on the environment and the potential threats to human health. This work introduces an innovative and cost-effective approach to detect and quantify plastic substances in teabags by applying near infrared hyperspectral imaging (951-2496 nm) coupled with multivariate analysis. Teabags from 6 popular brands were investigated and categorized into three classes based on spectral unmixing and target detection results: 1) the plastic teabag primarily made of nylon 6/6; 2) those made of a composite with various polypropylene and cellulose ratios; 3) biodegradable teabags free from any plastic traces. Results demonstrated the presence of numerous plastic particles in the beverage obtained after steeping nylon teabags, but the release of particles was further amplified after microwave treatment. Nevertheless, target detection results obtained from Fourier transform infrared imaging (4000-675 cm-1) dataset evidenced that a considerable proportion of particle residues detected were the contaminants obtained from tea granules that adsorbed on the teabag. This work highlights the significant importance of performing rigorous spectral analysis for chemical characterization, which is lacking in most published microplastic studies.

Using spatial information for evaluating the quality of prediction maps from hyperspectral images: A geostatistical approach.

Applying a calibration model onto hyperspectral (HS) images is of great interest because it produces images of chemical or physical properties. HS imaging is widely used in this way in food processing industries for monitoring product quality and process control. In this context, one of the main difficulties in the application of regression models to HS images is to evaluate the error of the obtained predictions, since in a proximal imaging set up, the size of the pixels is usually much smaller than the area required to obtain a wet chemical reference. Moreover, the selection of regression model parameters, such as the number of latent variables (LV) in a partial least squares (PLS) model, can modify the appearance of the prediction maps. The objective of this work is to propose an approach based on geostatistical indices to use spatial information of prediction maps for supporting the evaluation of regression models applied to HS images. This work stablishes a theoretical connection between linear regression model performance estimates and the spatial decomposition of variance in prediction maps, when the ground truth to estimate is spatially structured. This approach was tested in a simulated dataset and two real case studies. Geostatistical indices of the prediction maps were compared to model performance metrics for PLS models with increasing number of LV. The theoretical framework was proven by the results on the simulated dataset. In particular, the evolution of the nugget effect, C0, corresponded with the evolution of the random error of the model. Conversely, the error term of the model related with the slope of the model corresponded with the evolution of the structured variance observed in the prediction maps. On the real case studies, geostatistical indexes, extracted from the prediction maps, allowed to quantitatively evaluate the spatial structure of the estimations and complement the Root Mean Standard Error of Cross Validation (RMSECV) for the choice of optimal number of LV to consider in the model. The main advantage of this approach is that it does not require ground truth values. It could be used as a source of information for supporting the choice of optimum calibration parameters, such as the number of latent variables, or the choice of pre-treatments, complementing the traditional visual inspection of prediction maps with quantitative and objective metrics.

Effect of the Architecture of Fiber-Optic Probes Designed for Soluble Solid Content Prediction in Intact Sugar Beet Slices.

Sugar beet is the second biggest world contributor to sugar production and the only one grown in Europe. One of the main limitations for its competitiveness is the lack of effective tools for assessing sugar content in unprocessed sugar beet roots, especially in breeding programs. In this context, a dedicated near infrared (NIR) fiber-optic probe based approach is proposed. NIR technology is widely used for the estimation of sugar content in vegetable products, while optic fibers allow a wide choice of technical properties and configurations. The objective of this research was to study the best architecture through different technical choices for the estimation of sugar content in intact sugar beet roots. NIR spectral measurements were taken on unprocessed sugar beet samples using two types of geometries, single and multiple fiber-probes. Sugar content estimates were more accurate when using multiple fiber-probes (up to R2 = 0.93) due to a lesser disruption of light specular reflection. In turn, on this configuration, the best estimations were observed for the smallest distances between emitting and collecting fibers, reducing the proportion of multiply scattered light in the spectra. Error of prediction (RPD) values of 3.95, 3.27 and 3.09 were obtained for distances between emitting and collecting fibers of 0.6, 1.2 and 1.8 µm respectively. These high RPD values highlight the good predictions capacities of the multi-fiber probes. Finally, this study contributes to a better understanding of the effects of the technical properties of optical fiber-probes on the quality of spectral models. In addition, and beyond this specificity related to sugar beet, these findings could be extended to other turbid media for quantitative optical spectroscopy and eventually to validate considered fiber-optic probe design obtained in this experimental study.

Multispectral vision for monitoring peach ripeness.

UNLABELLED: The main objective of this research was to develop an automatic procedure able to classify Rich Lady commercial peaches according to their ripeness stage through multispectral imaging techniques. A classification procedure was applied to the ratio images calculated as red (R, 680 nm) divided by infrared (IR, 800 nm), that is, R/IR images. Four image-based ripeness reference classes (A: unripe to D: overripe) were generated from 380 fruit images (season 1: 2006) by a nonsupervised classification method and evaluated according to reference measurements of the ripeness of the same samples: Magness-Taylor penetrometry firmness, low-mass impact firmness, reflectance at 680 nm (R680, and soluble solids content. The assignment of unknown sample images from those season 1 images (internal validation, n = 380) and of 240 images from the 2nd season (season 2: 2007) to the ripeness reference classes (external validation) was carried out by computing the minimum Euclidean distance (classification distance, C(d)) between each unknown image histogram and the average histogram of each ripeness reference class. For both validation phases, firmness values decreased and R680 increased for increasing alphabetical order of image-based class letter, reflecting the ripening process. Moreover, 70% (season 1) and 80% (season 2) of the samples below bruise susceptibility firmness were classified into class D. PRACTICAL APPLICATION: This work proposes and validates a procedure for assessing peach ripeness through spectral imaging. The control of ripeness in this fruit is crucial for ensuring its quality and the measurement of optimum peach ripeness at harvest and postharvest is a controversial issue, which needs to be balanced between a minimum ripeness, acceptable for the consumer, and a maximum ripeness, to minimize fruit losses during the postharvest process. The proposed method is nondestructive and quick, showing thus, a good perspective for its application in fresh fruit packing lines, either for peach ripeness assessment or for other fruits (providing adequate calibration).