Unlocking Preclinical Alzheimer's: A Multi-Year Label-Free In Vitro Raman Spectroscopy Study Empowered by Chemometrics.

Alzheimer's disease is a progressive neurodegenerative disorder, the early detection of which is crucial for timely intervention and enrollment in clinical trials. However, the preclinical diagnosis of Alzheimer's encounters difficulties with gold-standard methods. The current definitive diagnosis of Alzheimer's still relies on expensive instrumentation and post-mortem histological examinations. Here, we explore label-free Raman spectroscopy with machine learning as an alternative to preclinical Alzheimer's diagnosis. A special feature of this study is the inclusion of patient samples from different cohorts, sampled and measured in different years. To develop reliable classification models, partial least squares discriminant analysis in combination with variable selection methods identified discriminative molecules, including nucleic acids, amino acids, proteins, and carbohydrates such as taurine/hypotaurine and guanine, when applied to Raman spectra taken from dried samples of cerebrospinal fluid. The robustness of the model is remarkable, as the discriminative molecules could be identified in different cohorts and years. A unified model notably classifies preclinical Alzheimer's, which is particularly surprising because of Raman spectroscopy's high sensitivity regarding different measurement conditions. The presented results demonstrate the capability of Raman spectroscopy to detect preclinical Alzheimer's disease for the first time and offer invaluable opportunities for future clinical applications and diagnostic methods.

Evidence of internalized microplastics in mussel tissues detected by volumetric Raman imaging.

Microplastics are a global ecological concern due to their potential risk to wildlife and human health. Animals ingest microplastics, which can enter the trophic chain and ultimately impact human well-being. The ingestion of microplastics can cause physical and chemical damage to the animals' digestive systems, affecting their health. To estimate the risk to ecosystems and human health, it is crucial to understand the accumulation and localization of ingested microplastics within the cells and tissues of living organisms. However, analyzing this issue is challenging due to the risk of sample contamination, given the ubiquity of microplastics. Here, an analytical approach is employed to confirm the internalization of microplastics in cryogenic cross-sections of mussel tissue. Using 3D Raman confocal microscopy in combination with chemometrics, microplastics measuring 1 µm in size were detected. The results were further validated using optical and fluorescence microscopy. The findings revealed evidence of microplastics being internalized in the digestive epithelial tissues of exposed mussels (Mytilus galloprovincialis), specifically within the digestive cells forming digestive alveoli. This study highlights the need to investigate the internalization of microplastics in organisms like mussels, as it helps us understand the potential risks they pose to aquatic biota and ultimately to human health. By employing advanced imaging techniques, challenges associated with sample contamination can be overcome and valuable insights into the impact of microplastics on marine ecosystems and human consumers are provided.

Successful Applications of NIR Spectroscopy and NIR Imaging in the Food Processing Chain.

Forty years ago, Near InfraRed (NIR) was considered a sleeping technique among the spectroscopic ones [...].

The importance of choosing a proper validation strategy in predictive models. A tutorial with real examples.

Machine learning is the art of combining a set of measurement data and predictive variables to forecast future events. Every day, new model approaches (with high levels of sophistication) can be found in the literature. However, less importance is given to the crucial stage of validation. Validation is the assessment that the model reliably links the measurements and the predictive variables. Nevertheless, there are many ways in which a model can be validated and cross-validated reliably, but still, it may be a model that wrongly reflects the real nature of the data and cannot be used to predict external samples. This manuscript shows in a didactical manner how important the data structure is when a model is constructed and how easy it is to obtain models that look promising with wrong-designed cross-validation and external validation strategies. A comprehensive overview of the main validation strategies is shown, exemplified by three different scenarios, all of them focused on classification.

Misinterpretation in microplastic detection in biological tissues: When 2D imaging is not enough.

The presence of microplastics in the food chain is a public concern worldwide, and its analysis is an analytical challenge. In our research, we apply Raman imaging to study the presence of 1 µm polystyrene microplastics in cryosections of Mytilus galloprovincialis due to its wide geographic distribution, widespread occurrence in the food web, and general high presence in the environment. Ingested microplastics are accumulated in the digestive tract, but a large number can also be rapidly eliminated. Some authors state that the translocation of microplastics to the epithelial cells is possible, increasing the risk of microplastics transmission along the food chain. However, as seen in our study, a surface imaging approach (2D) is probably not enough to confirm the internalization of particles and avoid misinterpretation. In fact, while some microplastic particles were detected in the epithelium by 2D Raman imaging, further 3D Raman imaging analysis demonstrated that those particles were dragged from the lumens to the epithelium during sample preparation due to the blade drag effect of the cryotome, and subsequently located on the surface of the analyzed cryosection, discarding the translocation to the epithelial cells. This effect can also happen when the samples are fortuitously contaminated during sample preparation. Several research articles that use similar analytical techniques have shown the presence of microplastics in different types of tissue. It is not our intention to put such results in doubt, but the present work points out the necessity of appropriate three-dimensional analytical methods including data interpretation and the need to go a step further than just surface imaging analysis.

Batch-wise versus continuous dough mixing of Danish butter cookies. A near infrared hyperspectral imaging study.

The Danish buttered cookie is a famous confectionery product. Its success makes manufacturing of the large volumes required challenging, introducing the need for different strategies to increase production while maintaining a high-quality standard. Two manufacturing lines used are batch-wise and continuous dough mixing. Despite the recipe being the same, the outcome of the two production types differs in texture and external appearance. While this does not infringe on the quality, changes in texture are observable. This manuscript analyses the physicochemical differences of the cookies after baking using Near Infrared hyperspectral imaging and Chemometrics. The study demonstrates that the changes in texture between batch and continuous production are mostly due to the difference in crystalline sucrose emerging in invisible spots on or near the surface of the cookies and a higher tendency of migrated butter-fat spots on the surface of the cookies for the continuous manufacturing procedure.

Data reduction by randomization subsampling for the study of large hyperspectral datasets.

Large amount of information in hyperspectral images (HSI) generally makes their analysis (e.g., principal component analysis, PCA) time consuming and often requires a lot of random access memory (RAM) and high computing power. This is particularly problematic for analysis of large images, containing millions of pixels, which can be created by augmenting series of single images (e.g., in time series analysis). This tutorial explores how data reduction can be used to analyze time series hyperspectral images much faster without losing crucial analytical information. Two of the most common data reduction methods have been chosen from the recent research. The first one uses a simple randomization method called randomized sub-sampling PCA (RSPCA). The second implies a more robust randomization method based on local-rank approximations (rPCA). This manuscript exposes the major benefits and drawbacks of both methods with the spirit of being as didactical as possible for a reader. A comprehensive comparison is made considering the amount of information retained by the PCA models at different compression degrees and the performance time. Extrapolation is also made to the case where the effect of time and any other factor are to be studied simultaneously.

Combination of Two Synchrotron Radiation-Based Techniques and Chemometrics to Study an Enhanced Natural Remineralization of Enamel.

The limitations to assess dental enamel remineralization have been overcome by a methodology resulting from the appropriate combination of synchrotron radiation-based techniques on both, infrared microspectroscopy and micro X-ray diffraction, with the help of specific data mining. Since amelogenin plays a key role in modulating the mineralization of tooth enamel, we propose a controlled ion release for fluorapatite structural ions (Ca2+, PO43-, and F-, also including Zn2+) by using weak acid and weak base ion-exchange resins in the presence of amelogenin to remineralize the surface of etched teeth. This combination provides the necessary ions for enamel remineralization and a guide for crystal growth due to the protein. Remineralized tooth samples were analyzed by applying the indicated methodology. The synchrotron data were treated using principal component analysis and multivariate curve resolution to analyze the mineral layer formed in the presence and absence of amelogenin. The remineralizing treatment created a fluorapatite layer free of carbonate impurities and with a similar orientation to that of the natural enamel thanks to amelogenin contribution.

Tooth whitening effects on dental enamel, oxidation or reduction? Comparison of physicochemical alterations in bovine enamel using Synchrotron-based Micro-FTIR.

OBJECTIVES: To compare the side effects of typical whitening treatments (by means of oxidation) compared to the new treatment developed by the authors through reduction. The aim is to provide information about the chemical interactions of the encapsulated reductant agent (metabisulfite, MBS) with the enamel structure compared with carbamide peroxide (CP) and to study their penetration in the hydroxyapatite (HAP) and the changes produced in the mineral and its hardness. METHODS: Chemical imaging is performed by synchrotron-based micro Fourier transformed infrared spectroscopy (SR-µFTIR). Continuous Stiffness Measurements (CSM) were used to determine the depth reached by the treatments in order to delimitate the area of study. RESULTS: The SR-µFTIR studies showed that MBS treatments softened the first 10 µm of enamel, as happens in the initial stages of tooth decay. Principal component analysis (PCA) showed that the main differences between treatments were found in the intensity of the ν3 PO43- peak related to tooth demineralization. CP and MBS promoted different changes in the HAP mineral, observed as opposite shifts of the peak: CP shortened the P-O bond while MBS seemed to elongate it. Moreover, MBS promoted the loss of carbonates while CP did not, which is probably related to the solution's pH. When comparing MBS and MBS Liposomes, it was observed how liposomes favoured the diffusion of MBS to inner layers, since the effects of MBS were observed in deeper enamel. Thus, the encapsulated MBS whitening effect is highly improved in terms of time when compared to MBS alone or CP. SIGNIFICANCE: The obtained results indicated that using oxidizing (CP) or reducing (MBS) treatments, promote different HAP mineral changes, and that liposomes favour the diffusion of MBS into the enamel. It is the first time that synchrotron light is used to map the bovine incisor's enamel chemically, and to determine the effect of a whitening treatment in terms of chemical HAP modifications, and the extent in deep of these effects.

Data Mining of Polymer Phase Transitions upon Temperature Changes by Small and Wide-Angle X-ray Scattering Combined with Raman Spectroscopy.

The complex physical transformations of polymers upon external thermodynamic changes are related to the molecular length of the polymer and its associated multifaceted energetic balance. The understanding of subtle transitions or multistep phase transformation requires real-time phenomenological studies using a multi-technique approach that covers several length-scales and chemical states. A combination of X-ray scattering techniques with Raman spectroscopy and Differential Scanning Calorimetry was conducted to correlate the structural changes from the conformational chain to the polymer crystal and mesoscale organization. Current research applications and the experimental combination of Raman spectroscopy with simultaneous SAXS/WAXS measurements coupled to a DSC is discussed. In particular, we show that in order to obtain the maximum benefit from simultaneously obtained high-quality data sets from different techniques, one should look beyond traditional analysis techniques and instead apply multivariate analysis. Data mining strategies can be applied to develop methods to control polymer processing in an industrial context. Crystallization studies of a PVDF blend with a fluoroelastomer, known to feature complex phase transitions, were used to validate the combined approach and further analyzed by MVA.

Staling of white wheat bread crumb and effect of maltogenic α-amylases. Part 3: Spatial evolution of bread staling with time by near infrared hyperspectral imaging.

This paper explores how the staling of white bread affects the behavior of the whole crumb surface and how that mechanism is interrupted/changed by the addition of maltogenic α-amylases. This is done using near infrared hyperspectral imaging, machine learning methodologies and the knowledge acquired in the previous two manuscripts. Methods like principal component analysis and multivariate curve resolution demonstrate how the constituents of the bread being stored (for 21 days) evolve differently depending on the presence/absence of maltogenic α-amylases and also which parts of the crumb are primarily exposed to changes. The spatial distribution of the hardness is calculated in the entire surface of the slice area during staling by using partial least square regression. This manuscript comprehends one of the largest studies made on white bread staling and proposes a complete methodology using near infrared hyperspectral imaging and machine learning.

Near infrared hyperspectral imaging and spectral unmixing methods for evaluation of fiber distribution in enriched pasta.

Pasta is mostly composed by wheat flour and water. Nevertheless, flour can be partially replaced by fibers to provide extra nutrients in the diet. However, fiber can affect the technological quality of pasta if not properly distributed. Usually, determinations of parameters in pasta are destructive and time-consuming. The use of Near Infrared-Hyperspectral Imaging (NIR-HSI), together with machine learning methods, is valuable to improve the efficiency in the assessment of pasta quality. This work aimed to investigate the ability of NIR-HSI and augmented Multivariate Curve Resolution-Alternating Least Squares (MCR-ALS) for the evaluation, resolution and quantification of fiber distribution in enriched pasta. Results showed R2V between 0.28 and 0.89, %LOF < 6%, variance explained over 99%, and similarity between pure and recovered spectra over 96% and 98% in models using pure flour and control as initial estimates, respectively, demonstrating the applicability of NIR-HSI and MCR-ALS in the identification of fiber in pasta.

Detection and identification of Cannabis sativa L. using near infrared hyperspectral imaging and machine learning methods. A feasibility study.

Remote identification of illegal plantations of Cannabis sativa Linnaeus is an important task for the Brazilian Federal Police. The current analytical methodology is expensive and strongly dependent on the expertise of the forensic investigator. A faster and cheaper methodology based on automatic methods can be useful for the detection and identification of Cannabis sativa L. in a reliable and objective manner. In this work, the high potential of Near Infrared Hyperspectral Imaging (HSI-NIR) combined with machine learning is demonstrated for supervised detection and classification of Cannabis sativa L. This plant, together with other plants commonly found in the surroundings of illegal plantations and soil, were directly collected from an illegal plantation. Due to the high correlation of the NIR spectra, sparse Principal Component Analysis (sPCA) was implemented to select the most important wavelengths for identifying Cannabis sativa L. One class Soft Independent Class Analogy model (SIMCA) was built, considering just the spectral variables selected by sPCA. Sensitivity and specificity values of 89.45% and 97.60% were, respectively, obtained for an external validation set subjected to the s-SIMCA. The results proved the reliability of a methodology based on NIR hyperspectral cameras to detect and identify Cannabis sativa L., with only four spectral bands, showing the potential of this methodology to be implemented in low-cost airborne devices.

Staling of white wheat bread crumb and effect of maltogenic α-amylases. Part 2: Monitoring the staling process by using near infrared spectroscopy and chemometrics.

This second paper provides chemical insight of the different phenomena occurring in bread during storage with and without anti-staling enzymes by using near infrared spectroscopy and chemometrics. The target, thus, is three-fold: (1) To monitor the staling process in the top, middle and bottom parts of the white bread loaf by near infrared spectroscopy and to extract chemical information of the different chemical mechanisms occurring in the staling process; (2) to assess the correlation between the near infrared spectroscopy and spatial texture profile analysis in terms of hardness, and (3) to demonstrate the anti-staling effect of the enzymes by showing a collapse of the correlation between near infrared (NIR) spectra and hardness as measured by texture analysis. It is found that NIR spectroscopy in combination with chemometrics (Partial Least Squares Regression) can predict the hardness development of the control bread.

Chemometric approaches for document dating: Handling paper variability.

A non-destructive methodology based on Fourier Transformed Infrared Spectroscopy (FTIR) is proposed in this research to estimate the age of documents of different ages. Due the variability in the samples caused by their different chemical compositions, chemometric approaches were proposed to build one unique regression model able to determine the age of the paper regardless of its composition. PLS models were built employing Generalized Least Squares Weighting (GLSW) and Orthogonal Least Squares (OLS) filters to reduce the variability of samples from the same year. Afterwards, sparse PLS, which is an extension of the PLS model including a variable selection step, was applied to compare its performance with the preprocessing filters. All techniques proposed were compared to the initial PLS models, showing the potential of the chemometric approaches applied to FTIR data to estimate the age of unknown documents.

Shear force analysis by core location in Longissimus steaks from Nellore cattle using hyperspectral images - A feasibility study.

Industry requires non-destructive real-time methods for quality control of meat in order to improve production efficiency and meet consumer expectations. Near Infrared Hyperspectral Images were used for tenderness evaluation of Nellore beef and the construction of tenderness distribution maps. To investigate whether the selection of the region of interest (ROI) in the image at the exact location where the shear force core was collected improves tenderness prediction and classification models, 50 samples from Longissimus muscle were imaged (1000-2500 nm) and shear force were measured (Warner-Bratzler). The data were analyzed by chemometric techniques (Partial Least Squares together with discriminant analysis - PLS-DA). Classification models using local ROI presented better performance than the ROI models of the whole sample (external validation sensitivity for the tough class = 33% and 70%, respectively), but none could be considered as successful model. However, the more general model had better performance in the tenderness distribution maps, with 72% of predicted images correctly classified.

Multi-spectral imaging for the estimation of shooting distances.

Multispectral images of clothing targets shot at seven different distances (from 10 to 220cm) were recorded at 18 specific wavelengths in the 400-1000nm range to visualize the gunshot residue (GSR) pattern. Principal component analysis (PCA) showed that the use of violet-blue wavelengths (430, 450 and 470nm) provided the largest contrast between the GSR particles and the white cotton fabric. Then, the correlation between the amount of GSR particles on clothing targets and the shooting distance was studied. By selecting the blue frame of multispectral images (i.e. the blue frame in the red-green-blue (RGB) system which falls at 470nm), the amount of pixels containing GSR particles was accounted based on the intensity of pixels in that frame. Results demonstrated that the number of pixels containing GSR exponentially decreases with the shooting distance from 30 to 220cm following a particular exponential equation. However, the targets shot at the shortest distance (10cm) did not satisfy the above equation, probably due to the noticeable differences of the GSR-pattern of these targets (e.g. high presence of soot). Then, the equation was applied to validation samples to estimate the shooting distances, obtaining results with an error below 10%.

Development and validation of a method for the determination of regulated fragrance allergens by High-Performance Liquid Chromatography and Parallel Factor Analysis 2.

This work presents the development and validation of a multivariate method for quantitation of 6 potentially allergenic substances (PAS) related to fragrances by ultrasound-assisted emulsification microextraction coupled with HPLC-DAD and PARAFAC2 in the presence of other 18 PAS. The objective is the extension of a previously proposed univariate method to be able to determine the 24 PAS currently considered as allergens. The suitability of the multivariate approach for the qualitative and quantitative analysis of the analytes is discussed through datasets of increasing complexity, comprising the assessment and validation of the method performance. PARAFAC2 showed to adequately model the data facing up different instrumental and chemical issues, such as co-elution profiles, overlapping spectra, unknown interfering compounds, retention time shifts and baseline drifts. Satisfactory quality parameters of the model performance were obtained (R2≥0.94), as well as meaningful chromatographic and spectral profiles (r≥0.97). Moreover, low errors of prediction in external validation standards (below 15% in most cases) as well as acceptable quantification errors in real spiked samples (recoveries from 82 to 119%) confirmed the suitability of PARAFAC2 for resolution and quantification of the PAS. The combination of the previously proposed univariate approach, for the well-resolved peaks, with the developed multivariate method allows the determination of the 24 regulated PAS.

Prediction of pork quality parameters by applying fractals and data mining on MRI.

This work firstly investigates the use of MRI, fractal algorithms and data mining techniques to determine pork quality parameters non-destructively. The main objective was to evaluate the capability of fractal algorithms (Classical Fractal algorithm, CFA; Fractal Texture Algorithm, FTA and One Point Fractal Texture Algorithm, OPFTA) to analyse MRI in order to predict quality parameters of loin. In addition, the effect of the sequence acquisition of MRI (Gradient echo, GE; Spin echo, SE and Turbo 3D, T3D) and the predictive technique of data mining (Isotonic regression, IR and Multiple linear regression, MLR) were analysed. Both fractal algorithm, FTA and OPFTA are appropriate to analyse MRI of loins. The sequence acquisition, the fractal algorithm and the data mining technique seems to influence on the prediction results. For most physico-chemical parameters, prediction equations with moderate to excellent correlation coefficients were achieved by using the following combinations of acquisition sequences of MRI, fractal algorithms and data mining techniques: SE-FTA-MLR, SE-OPFTA-IR, GE-OPFTA-MLR, SE-OPFTA-MLR, with the last one offering the best prediction results. Thus, SE-OPFTA-MLR could be proposed as an alternative technique to determine physico-chemical traits of fresh and dry-cured loins in a non-destructive way with high accuracy.

Transferring results from NIR-hyperspectral to NIR-multispectral imaging systems: A filter-based simulation applied to the classification of Arabica and Robusta green coffee.

Due to the differences in terms of both price and quality, the availability of effective instrumentation to discriminate between Arabica and Robusta coffee is extremely important. To this aim, the use of multispectral imaging systems could provide reliable and accurate real-time monitoring at relatively low costs. However, in practice the implementation of multispectral imaging systems is not straightforward: the present work investigates this issue, starting from the outcome of variable selection performed using a hyperspectral system. Multispectral data were simulated considering four commercially available filters matching the selected spectral regions, and used to calculate multivariate classification models with Partial Least Squares-Discriminant Analysis (PLS-DA) and sparse PLS-DA. Proper strategies for the definition of the training set and the selection of the most effective combinations of spectral channels led to satisfactory classification performances (100% classification efficiency in prediction of the test set).