Seabream quality monitoring throughout the supply chain using a portable multispectral imaging device.

Monitoring food quality throughout the supply chain in a rapid and cost-effective way allows on-time decision making, reducing food waste and increasing sustainability. In that framework, a portable multispectral imaging sensor was used, while the acquired data in combination with neural networks were evaluated for the prediction of fish fillets quality. Images of fish fillets were acquired using samples from both aquaculture and retail stores of different packaging and fish parts. The obtained products (air or vacuum packaged) were further stored at different temperature conditions. In parallel to image acquisition, microbial quality was estimated as well. The data were used for the training of predictive neural models that aimed to estimate total aerobic counts (TAC). The models were developed and validated using data from aquaculture and were externally validated with samples purchased from the retail stores. The set up allowed the evaluation of models for the different parts of the fish and conditions. The performance for the validation set was similar for flesh (RMSE: 0.402-0.547) and skin side (RMSE: 0.500-0.533) of the fish fillets. The performance for the different packaging conditions was also similar, however, in the external validation, the vacuum-packaged samples showed better performance in terms of RMSE compared to the air-packaged ones. Models irrespective of packaging condition are very important for cases where the products' history is unknown although the prediction capability was not as high as in the models per packaging condition individually. The models tested with unknown samples (i.e., from retail stores) showed poorer performance (RMSE: 1.061-1.414) compared to the models validated with data partitioning (RMSE: 0.402-0.547). Multispectral imaging sensor appeared to be efficient for the rapid assessment of the microbiological quality of fish fillets for all the different cases evaluated.

Microbiological Quality Estimation of Meat Using Deep CNNs on Embedded Hardware Systems.

Spectroscopic sensor imaging of food samples meta-processed by deep machine learning models can be used to assess the quality of the sample. This article presents an architecture for estimating microbial populations in meat samples using multispectral imaging and deep convolutional neural networks. The deep learning models operate on embedded platforms and not offline on a separate computer or a cloud server. Different storage conditions of the meat samples were used, and various deep learning models and embedded platforms were evaluated. In addition, the hardware boards were evaluated in terms of latency, throughput, efficiency and value on different data pre-processing and imaging-type setups. The experimental results showed the advantage of the XavierNX platform in terms of latency and throughput and the advantage of Nano and RP4 in terms of efficiency and value, respectively.

The use of multispectral imaging for the discrimination of Arabica and Robusta coffee beans.

Arabica coffee beans are sold at twice the price, or more, compared to Robusta beans and consequently are susceptible to economically motivated adulteration by substitution. There is a need for rapid, non-destructive, and efficient analytical techniques for monitoring the authenticity of Arabica coffee beans in the supply chain. In this study, multispectral imaging (MSI) was applied to discriminate roasted Arabica and Robusta coffee beans and perform quantitative prediction of Arabica coffee bean adulteration with Robusta. The Orthogonal Partial Least Squares Discriminant Analysis (OPLS-DA) model, built using selected spectral and morphological features from individual coffee beans, achieved 100% correct classification of the two coffee species in the test dataset. The OPLS regression model was able to successfully predict the level of adulteration of Arabica with Robusta. MSI analysis has potential as a rapid screening tool for the detection of fraud issues related to the authenticity of Arabica coffee beans.

Fluorescence spectroscopy and multispectral imaging for fingerprinting of aflatoxin-B1 contaminated (Zea mays L.) seeds: a preliminary study.

Cereal seeds safety may be compromised by the presence of toxic contaminants, such as aflatoxins. Besides being carcinogenic, they have other adverse health effects on humans and animals. In this preliminary study, we used two non-invasive optical techniques, optical fiber fluorescence spectroscopy and multispectral imaging (MSI), for discrimination of maize seeds naturally contaminated with aflatoxin B1 (AFB1) from the uncontaminated seeds. The AFB1-contaminated seeds exhibited a red shift of the emission maximum position compared to the control samples. Using linear discrimination analysis to analyse fluorescence data, classification accuracy of 100% was obtained to discriminate uncontaminated and AFB1-contaminated seeds. The MSI analysis combined with a normalized canonical discriminant analysis, provided spectral and spatial patterns of the analysed seeds. The AFB1-contaminated seeds showed a 7.9 to 9.6-fold increase in the seed reflectance in the VIS region, and 10.4 and 12.2-fold increase in the NIR spectral region, compared with the uncontaminated seeds. Thus the MSI method classified successfully contaminated from uncontaminated seeds with high accuracy. The results may have an impact on development of spectroscopic non-invasive methods for detection of AFs presence in seeds, providing valuable information for the assessment of seed adulteration in the field of food forensics and food safety.

Multispectral and X-ray images for characterization of Jatropha curcas L. seed quality.

BACKGROUND: The use of non-destructive methods with less human interference is of great interest in agricultural industry and crop breeding. Modern imaging technologies enable the automatic visualization of multi-parameter for characterization of biological samples, reducing subjectivity and optimizing the analysis process. Furthermore, the combination of two or more imaging techniques has contributed to discovering new physicochemical tools and interpreting datasets in real time. RESULTS: We present a new method for automatic characterization of seed quality based on the combination of multispectral and X-ray imaging technologies. We proposed an approach using X-ray images to investigate internal tissues because seed surface profile can be negatively affected, but without reaching important internal regions of seeds. An oilseed plant (Jatropha curcas) was used as a model species, which also serves as a multi-purposed crop of economic importance worldwide. Our studies included the application of a normalized canonical discriminant analyses (nCDA) algorithm as a supervised transformation building method to obtain spatial and spectral patterns on different seedlots. We developed classification models using reflectance data and X-ray classes based on linear discriminant analysis (LDA). The classification models, individually or combined, showed high accuracy (> 0.96) using reflectance at 940 nm and X-ray data to predict quality traits such as normal seedlings, abnormal seedlings and dead seeds. CONCLUSIONS: Multispectral and X-ray imaging have a strong relationship with seed physiological performance. Reflectance at 940 nm and X-ray data can efficiently predict seed quality attributes. These techniques can be alternative methods for rapid, efficient, sustainable and non-destructive characterization of seed quality in the future, overcoming the intrinsic subjectivity of the conventional seed quality analysis.

Multispectral imaging of hand eczema.

BACKGROUND: Hand eczema is a disease with large variation in clinical presentation and severity. Scoring systems for quantitative severity assessment exist. However, they are observer-dependent. An objective quantitative tool for scoring of hand eczema would improve categorization of hand eczema. OBJECTIVE: To investigate the usefulness of multispectral imaging in assessing severity of hand eczema with respect to extent and the different morphological features. METHODS: Patients with hand eczema (n = 60) and healthy controls (n = 28) were included. The severity of hand eczema was assessed by a dermatologist using the Hand Eczema Severity Index (HECSI) and a global assessment (Physician Global Assessment [PGA]). Multispectral imaging of the hand was performed on all patients and controls using the VideometerLab Instrument. RESULTS: Areas of the morphological elements identified by multispectral imaging were statistically significantly correlated with the PGA scores. Analyzed by Cohen's kappa, a moderate agreement between imaging-based severity assessment and PGA was found. The imaging-based severity assessment was also correlated with HECSI (Spearman rho 0.683, P < .001). Still, the imaging-based algorithm was not capable of differentiating hand eczema patients from controls. CONCLUSIONS: Multispectral imaging allows quantitative measurements of different skin parameters to be performed. In its present form, multispectral imaging cannot replace the clinical assessment of a dermatologist. However, after refinement, this or similar technologies could prove useful.

Using Polarized Spectroscopy to Investigate Order in Thin-Films of Ionic Self-Assembled Materials Based on Azo-Dyes.

Three series of ionic self-assembled materials based on anionic azo-dyes and cationic benzalkonium surfactants were synthesized and thin films were prepared by spin-casting. These thin films appear isotropic when investigated with polarized optical microscopy, although they are highly anisotropic. Here, three series of homologous materials were studied to rationalize this observation. Investigating thin films of ordered molecular materials relies to a large extent on advanced experimental methods and large research infrastructure. A statement that in particular is true for thin films with nanoscopic order, where X-ray reflectometry, X-ray and neutron scattering, electron microscopy and atom force microscopy (AFM) has to be used to elucidate film morphology and the underlying molecular structure. Here, the thin films were investigated using AFM, optical microscopy and polarized absorption spectroscopy. It was shown that by using numerical method for treating the polarized absorption spectroscopy data, the molecular structure can be elucidated. Further, it was shown that polarized optical spectroscopy is a general tool that allows determination of the molecular order in thin films. Finally, it was found that full control of thermal history and rigorous control of the ionic self-assembly conditions are required to reproducibly make these materials of high nanoscopic order. Similarly, the conditions for spin-casting are shown to be determining for the overall thin film morphology, while molecular order is maintained.

UV imaging of multiple unit pellet system (MUPS) tablets: A case study of acetylsalicylic acid stability.

The applicability of multispectral ultraviolet (UV) imaging in combination with multivariate image analysis was investigated to monitor API degradation within multiple unit pellet system (MUPS) tablets during storage. For this purpose, acetylsalicylic acid (ASA) layered pellets were coated with Eudragit® RL PO and compressed to MUPS tablets. These tablets were stored under four different conditions with different levels of relative humidity (0 and 75%) and temperature (21 and 40°C) and analysed at seven storage time points (0, 15, 40, 140, 165, 265, and 330d). The UV imaging results for estimation of the salicylic acid (SA) concentration as degradation product of ASA in the tablets were compared to the SA concentration measured by high performance liquid chromatography with a partial least squares regression resulting in an RMSEP of 4.86% and an R2 of 0.9812. The estimation of the SA concentration based on mean UV reflectance spectra was possible even through the coating of the API pellets and at low concentration levels. In addition, the distribution of the SA concentration on the tablet surfaces for different storage time periods was visualized. UV imaging as fast and non-destructive method appears to offer significant potential for monitoring of API degradation during stability studies.

Multispectral UV Imaging for Determination of the Tablet Coating Thickness.

The applicability of off-line multispectral UV imaging in combination with multivariate data analysis was investigated to determine the coating thickness and its distribution on the tablet surface during lab-scale coating. The UV imaging results were compared with the weight gain measured for each individual tablet and the corresponding coating thickness and its distribution measured by terahertz pulsed imaging (TPI). Three different tablet formulations were investigated, 2 of which contained UV-active tablet cores. Three coating formulations were applied: Aquacoat® ECD (a mainly translucent coating) and Eudragit® NE (a turbid coating containing solid particles). It was shown that UV imaging is a fast and nondestructive method to predict individual tablet weight gain as well as coating thickness. The coating thickness distribution profiles determined by UV imaging correlated to the results of the TPI measurements. UV imaging appears to hold a significant potential as a process analytical technology tool for determination of the tablet coating thickness and its distribution resulting from its high measurement speed, high molar absorptivity, and a high scattering coefficient, in addition to relatively low costs.

Multispectral UV imaging for surface analysis of MUPS tablets with special focus on the pellet distribution.

In the present study the applicability of multispectral UV imaging in combination with multivariate image analysis for surface evaluation of MUPS tablets was investigated with respect to the differentiation of the API pellets from the excipients matrix, estimation of the drug content as well as pellet distribution, and influence of the coating material and tablet thickness on the predictive model. Different formulations consisting of coated drug pellets with two coating polymers (Aquacoat® ECD and Eudragit® NE 30 D) at three coating levels each were compressed to MUPS tablets with various amounts of coated pellets and different tablet thicknesses. The coated drug pellets were clearly distinguishable from the excipients matrix using a partial least squares approach regardless of the coating layer thickness and coating material used. Furthermore, the number of the detected drug pellets on the tablet surface allowed an estimation of the true drug content in the respective MUPS tablet. In addition, the pellet distribution in the MUPS formulations could be estimated by UV image analysis of the tablet surface. In conclusion, this study revealed that UV imaging in combination with multivariate image analysis is a promising approach for the automatic quality control of MUPS tablets during the manufacturing process.

Rapid Assessment of Tablet Film Coating Quality by Multispectral UV Imaging.

Chemical imaging techniques are beneficial for control of tablet coating layer quality as they provide spectral and spatial information and allow characterization of various types of coating defects. The purpose of this study was to assess the applicability of multispectral UV imaging for assessment of the coating layer quality of tablets. UV images were used to detect, characterize, and localize coating layer defects such as chipped parts, inhomogeneities, and cracks, as well as to evaluate the coating surface texture. Acetylsalicylic acid tablets were prepared on a rotary tablet press and coated with a polyvinyl alcohol-polyethylene glycol graft copolymer using a pan coater. It was demonstrated that the coating intactness can be assessed accurately and fast by UV imaging. The different types of coating defects could be differentiated and localized based on multivariate image analysis and Soft Independent Modeling by Class Analogy applied to the UV images. Tablets with inhomogeneous texture of the coating could be identified and distinguished from those with a homogeneous surface texture. Consequently, UV imaging was shown to be well-suited for monitoring of the tablet coating layer quality. UV imaging is a promising technique for fast quality control of the tablet coating because of the high data acquisition speed and its nondestructive analytical nature.

Quantitative surface topography assessment of directly compressed and roller compacted tablet cores using photometric stereo image analysis.

Surface topography, in the context of surface smoothness/roughness, was investigated by the use of an image analysis technique, MultiRay™, related to photometric stereo, on different tablet batches manufactured either by direct compression or roller compaction. In the present study, oblique illumination of the tablet (darkfield) was considered and the area of cracks and pores in the surface was used as a measure of tablet surface topography; the higher a value, the rougher the surface. The investigations demonstrated a high precision of the proposed technique, which was able to rapidly (within milliseconds) and quantitatively measure the obtained surface topography of the produced tablets. Compaction history, in the form of applied roll force and tablet punch pressure, was also reflected in the measured smoothness of the tablet surfaces. Generally it was found that a higher degree of plastic deformation of the microcrystalline cellulose resulted in a smoother tablet surface. This altogether demonstrated that the technique provides the pharmaceutical developer with a reliable, quantitative response parameter for visual appearance of solid dosage forms, which may be used for process and ultimately product optimization.

Multispectral UV imaging for fast and non-destructive quality control of chemical and physical tablet attributes.

Monitoring of tablet quality attributes in direct vicinity of the production process requires analytical techniques that allow fast, non-destructive, and accurate tablet characterization. The overall objective of this study was to investigate the applicability of multispectral UV imaging as a reliable, rapid technique for estimation of the tablet API content and tablet hardness, as well as determination of tablet intactness and the tablet surface density profile. One of the aims was to establish an image analysis approach based on multivariate image analysis and pattern recognition to evaluate the potential of UV imaging for automatized quality control of tablets with respect to their intactness and surface density profile. Various tablets of different composition and different quality regarding their API content, radial tensile strength, intactness, and surface density profile were prepared using an eccentric as well as a rotary tablet press at compression pressures from 20MPa up to 410MPa. It was found, that UV imaging can provide both, relevant information on chemical and physical tablet attributes. The tablet API content and radial tensile strength could be estimated by UV imaging combined with partial least squares analysis. Furthermore, an image analysis routine was developed and successfully applied to the UV images that provided qualitative information on physical tablet surface properties such as intactness and surface density profiles, as well as quantitative information on variations in the surface density. In conclusion, this study demonstrates that UV imaging combined with image analysis is an effective and non-destructive method to determine chemical and physical quality attributes of tablets and is a promising approach for (near) real-time monitoring of the tablet compaction process and formulation optimization purposes.

Comparison of a multispectral vision system and a colorimeter for the assessment of meat color.

The color assessment ability of a multispectral vision system is investigated by a comparison study with color measurements from a traditional colorimeter. The experiment involves fresh and processed meat samples. Meat is a complex material; heterogeneous with varying scattering and reflectance properties, so several factors can influence the instrumental assessment of meat color. In order to assess whether two methods are equivalent, the variation due to these factors must be taken into account. A statistical analysis was conducted and showed that on a calibration sheet the two instruments are equally capable of measuring color. Moreover the vision system provides a more color rich assessment of fresh meat samples with a glossier surface, than the colorimeter. Careful studies of the different sources of variation enable an assessment of the order of magnitude of the variability between methods accounting for other sources of variation leading to the conclusion that color assessment using a multispectral vision system is superior to traditional colorimeter assessments.

Chemical imaging and solid state analysis at compact surfaces using UV imaging.

Fast non-destructive multi-wavelength UV imaging together with multivariate image analysis was utilized to visualize distribution of chemical components and their solid state form at compact surfaces. Amorphous and crystalline solid forms of the antidiabetic compound glibenclamide, and microcrystalline cellulose together with magnesium stearate as excipients were used as model materials in the compacts. The UV imaging based drug and excipient distribution was in good agreement with hyperspectral NIR imaging. The UV wavelength region can be utilized in distinguishing between glibenclamide and excipients in a non-invasive way, as well as mapping the glibenclamide solid state form. An exploratory data analysis supported the critical evaluation of the mapping results and the selection of model parameters for the chemical mapping. The present study demonstrated that the multi-wavelength UV imaging is a fast process analytical technique with the potential for real-time monitoring of critical quality attributes.

OvaSpec - A vision-based instrument for assessing concentration and developmental stage of Trichuris suis parasite egg suspensions.

BACKGROUND: OvaSpec is a new, fully automated, vision-based instrument for assessing the quantity (concentration) and quality (embryonation percentage) of Trichuris suis parasite eggs in liquid suspension. The eggs constitute the active pharmaceutical ingredient in a medicinal drug for the treatment of immune-mediated diseases such as Crohn׳s disease, ulcerative colitis, and multiple sclerosis. METHODS: This paper describes the development of an automated microscopy technology, including methodological challenges and design decisions of relevance for the future development of comparable vision-based instruments. Morphological properties are used to distinguish eggs from impurities and two features of the egg contents under brightfield and darkfield illumination are used in a statistical classification to distinguish eggs with undifferentiated contents (non-embryonated eggs) from eggs with fully developed larvae inside (embryonated eggs). RESULTS: For assessment of the instrument׳s performance, six egg suspensions of varying quality were used to generate a dataset of unseen images. Subsequently, annotation of the detected eggs and impurities revealed a high agreement with the manual, image-based assessments for both concentration and embryonation percentage (both error rates <1.0%). Similarly, a strong correlation was demonstrated in a final, blinded comparison with traditional microscopic assessments performed by an experienced laboratory technician. CONCLUSIONS: The present study demonstrates the applicability of computer vision in the production, analysis, and quality control of T. suis eggs used as an active pharmaceutical ingredient for the treatment of autoimmune diseases.

Potential of multispectral imaging technology for rapid and non-destructive determination of the microbiological quality of beef filets during aerobic storage.

The performance of a multispectral imaging system has been evaluated in monitoring aerobically packaged beef filet spoilage at different storage temperatures (0, 4, 8, 12, and 16°C). Spectral data in the visible and short wave near infrared area (405-970nm) were collected from the surface of meat samples and correlated with microbiological data (log counts), for total viable counts (TVCs), Pseudomonas spp., and Brochothrix thermosphacta. Qualitative analysis (PLS-DA) was employed for the discrimination of meat samples in three microbiological quality classes based on the values of total viable counts, namely Class 1 (TVC<5.5log10CFU/g), Class 2 (5.5log10CFU/g7.0log10CFU/g). Furthermore, PLS regression models were developed to provide quantitative estimations of microbial counts during meat storage. In both cases model validation was implemented with independent experiments at intermediate storage temperatures (2 and 10°C) using different batches of meat. Results demonstrated good performance in classifying meat samples with overall correct classification rate for the three quality classes ranging from 91.8% to 80.0% for model calibration and validation, respectively. For quantitative estimation, the calculated regression coefficients between observed and estimated counts ranged within 0.90-0.93 and 0.78-0.86 for model development and validation, respectively, depending on the microorganism. Moreover, the calculated average deviation between observations and estimations was 11.6%, 13.6%, and 16.7% for Pseudomonas spp., B. thermosphacta, and TVC, respectively. The results indicated that multispectral vision technology has significant potential as a rapid and non-destructive technique in assessing the microbiological quality of beef fillets.

Mucosal host responses to bacteriuria in colonic and ileal neobladders.

OBJECTIVE: To study the local host response in patients with colonic and ileal neobladders, with or without bacteriuria. METHODS: Twenty-three patients with colonic neobladders and 19 with ileal neobladders were included. Eleven radical prostatectomy patients and seven healthy male volunteers were used as controls. Six urine samples were obtained from all patients and controls over a six-month period. The samples were cultured semiquantitatively, and the number of neutrophils and concentrations of the inflammatory mediators interleukin 6 and 8 (IL-6, IL-8) in the urine were determined. RESULTS: The prostatectomy patients and healthy volunteers had sterile urine, and concentrations of IL-6 and IL-8 were below the detection limit. Most (>70%) of the urine samples from patients with colonic and ileal neobladders showed anaerobic or aerobic bacterial growth, and uropathogens were identified in about 45% of the samples. The local host response was minimal or undetectable in the sterile urine samples. However, the host response was markedly induced by uropathogenic strains in the urine, but not by urinary carriage of nonpathogenic or anaerobic strains. IL-8, but not IL-6, was increased in colonic neobladders, which corresponds to the mucosal host responses in patients with intact lower urinary tracts and asymptomatic bacteriuria. In ileal neobladders, the IL-8 responses were higher, and levels of IL-6 were significantly increased. CONCLUSION: Neobladders exhibit a significant local host response to colonization with bacterial uropathogens. This reaction is more pronounced and includes IL-6 activation in ileal neobladders than in colonic neobladders.

Microbial flora in ileal and colonic neobladders.

OBJECTIVE: To describe bacterial colonization in patients with ileal and colonic neobladders. METHODS: Twenty-three patients with right colon neobladders, 30 with ileal neobladders, 11 who had undergone radical prostatectomy, and 6 healthy controls were included. Culture of clean-catch, midstream urine specimens was done weekly for 3 weeks, and this was repeated after 6 months. Residual urine was measured, and the patients were interviewed about leakage. All patients and controls were antibiotic free during the study except for 13 of the ileal neobladder patients, who were treated with trimethoprim 100mg daily. RESULTS: Urine cultures from controls and prostatectomy patients were negative for bacteria, whereas 67% of the specimens from patients with neobladders, not on antibiotic therapy, were culture positive, and half of these contained uropathogenic species, such as Escherichia coli, Klebsiella pneumoniae, Pseudomonas aeruginosa, and Enterococcus faecalis. Bacterial colonization (including uropathogenic strains) was strongly correlated with residual urine (p<0.005), but not with leakage. Anaerobic strains were found more frequently (p=0.04) in urine from ileal neobladders than in urine from colonic neobladders. The 13 patients with ileal neobladders and on prophylactic antibiotic therapy carried bacteriuria in 80% of the samples, the majority being anaerobic strains. Uropathogenic strains, mainly Enterecoccus faecalis was revealed in 30% of the samples. CONCLUSIONS: The lower urinary tract of patients with ileal or colonic neobladders is heavily colonized with potentially uropathogenic and anaerobic bacteria. Complete bladder emptying reduces the bacterial burden. Anaerobic colonization is increased in neobladders reconstructed from ileum. Prophylactic antibiotic therapy does not seem to reduce the bacterial burden, but interferes with the bacterial composition.

Visual clone identification of Penicillium commune isolates.

A method for visual clone identification of Penicillium commune isolates was developed. The method is based on images of fungal colonies acquired after growth on a standard medium and involves a high degree of objectivity, which in future studies will make it possible for non-experts to perform a qualified identification of different species as well as clones within a species. A total of 77 P. commune isolates from a cheese dairy were 3-point inoculated on Yeast Extract Sucrose (YES) agar and incubated for 7 days at 25 degrees C. After incubation, the isolates were classified into groups containing the same genotype determined by DNA fingerprinting (AFLP). Each genotype also has a specific phenotype such as different colony colours. By careful image acquisition, colours were measured in a reproducible way. Prior to image analysis, each image was corrected with respect to colour, geometry and self-illumination, thereby gaining a set of directly comparable images. A method for automatic extraction of a given number of concentric regions was used. Using the positions of the regions, a number of relevant features--capturing colour and colour-texture from the surface of the fungal colonies--was extracted for further analysis. We introduced the Jeffreys-Matusitas (JM) distance between the feature distributions to express the similarity between regions in two colonies, and to evaluate the overall (weighted) similarity. The nearest neighbour (NN) classification rule was used. On a dataset from 137 isolates, we obtained a "leave-one-out" cross-validation identification rate of approximately 93-98% compared with the result of DNA fingerprinting.