Near-Infrared Spectroscopy for Rapid Differentiation of Fresh and Frozen-Thawed Common Carp (Cyprinus carpio).

This study aimed to investigate near-infrared spectroscopy (NIRS) in combination with classification methods for the discrimination of fresh and once- or twice-freeze-thawed fish. An experiment was carried out with common carp (Cyprinus carpio). From each fish, test pieces were cut from the dorsal and ventral regions and measured from the skin side as fresh, after single freezing at minus 18 °C for 15 ÷ 28 days and 15 ÷ 21 days for the second freezing after the freeze-thawing cycle. NIRS measurements were performed via a NIRQuest 512 spectrometer at the region of 900-1700 nm in Reflection mode. The Pirouette 4.5 software was used for data processing. SIMCA and PLS-DA models were developed for classification, and their performance was estimated using the F1 score and total accuracy. The predictive power of each model was evaluated for fish samples in the fresh, single-freezing, and second-freezing classes. Additionally, aquagrams were calculated. Differences in the spectra between fresh and frozen samples were observed. They might be assigned mainly to the O-H and N-H bands. The aquagrams confirmed changes in water organization in the fish samples due to freezing-thawing. The total accuracy of the SIMCA models for the dorsal samples was 98.23% for the calibration set and 90.55% for the validation set. For the ventral samples, respective values were 99.28 and 79.70%. Similar accuracy was found for the PLS-PA models. The NIR spectroscopy and tested classification methods have a potential for nondestructively discriminating fresh from frozen-thawed fish in as methods to protect against fish meat food fraud.

Identifying Meat from Grazing or Feedlot Yaks Using Visible and Near-infrared Spectroscopy with Chemometrics.

The quality of meat can differ between grazing and feedlot yaks. The present study examined whether spectral fingerprints by visible and near-infrared (Vis-NIR) spectroscopy and chemo-metrics could be employed to identify the meat of grazing and feedlot yaks. Thirty-six 3.5-year-old castrated male yaks (164 ± 8.38 kg) were divided into grazing and feedlot yaks. After 5 months on treatment, liveweight, carcass weight, and dressing percentage were greater in the feedlot than in grazing yaks. The grazing yaks had greater protein content but lesser fat content than feedlot yaks. Principal component analysis (PCA) was able to identify the meat of the two groups to a great extent. Using either partial least squares discriminant analysis (PLS-DA) or the soft independent modeling of class analogies (SIMCA) classification, the meat could be differentiated between the groups. Both the original and processed spectral data had a high discrimination percentage, especially the PLS-DA classification algorithm, with 100% discrimination in the 400-2500 nm band. The spectral preprocessing methods can improve the discrimination percentage, especially for the SIMCA classification. It was concluded that the method can be employed to identify meat from grazing or feedlot yaks. The unerring consistency across different wavelengths and data treatments highlights the model's robustness and the potential use of NIR spectroscopy combined with chemometric techniques for meat classification. PLS-DA's accurate classification model is crucial for the unique evaluation of yak meat in the meat industry, ensuring product traceability and meeting consumer expectations for the authenticity and quality of yak meat raised in different ways.

Characterization and classification of oleogels and edible oil using vibrational spectroscopy in tandem with one-class and multiclass chemometric methods.

Oleogel represents a promising healthier alternative to act as a substitute for conventional fat in various food products. Oil selection is a crucial factor in determining the technological properties and applications of oleogels due to their distinct fatty acid composition, molecular weight, and thermal properties, as well as the presence of antioxidants and oxidative stability. Hence, the relevance of monitoring oleogel properties by non-destructive, eco-friendly, portable, fast, and effective techniques is a relevant task and constitutes an advance in the evaluation of oleogels quality. Thus, the present study aims to classify oleogels rapidly and reliably, without the use of chemicals, comparing two handheld near infrared (NIR) spectrometers and one portable Raman device. Furthermore, two different multivariate methods are compared for oleogel classification according to oil type. Three types of oleogels were prepared, containing 95 % oil (sunflower, soy, olive) and 5 % beeswax as a structuring agent, melted at 90 °C. Polarized light microscopy (PLM) images were acquired, and fatty acid composition, peroxide index and free fatty acid content were determined using official methods. A total of 240 oleogel and 92 oil spectra were obtained for each instrument. After spectra pretreatment, Principal Component Analysis (PCA) was performed, and two classification methods were investigated. The Data Driven - Soft Independent Modelling of Class Analogy (DD-SIMCA) and Partial Least Squares Discriminant Analysis (PLS-DA) models demonstrated 95 % to 100 % of accuracy for the external test set. In conclusion, the use of vibrational spectroscopy using handheld and portable instruments in tandem with chemometrics showed to be an efficient alternative for classifying oils and oleogels and could be extended to other food samples. Although the classification of vegetable oils by NIR is widely used and known, this work proposes the classification of different types of oil in oleogel matrices, which has not yet been explored in the literature.

Identification of lentils (Lens culinaris Medik) from Eglouvi (Lefkada, Greece) based on rare earth elements profile combined with chemometrics.

An analytical approach has been developed to verify the authenticity of premium lentils originating from Eglouvi, Lefkada, Greece. The method relies on the digestion of samples followed by the analysis of their rare earth elements (REEs) content. Lentils originating from Eglouvi exhibit higher content in most REEs compared to lentils from other regions as well as distinct Sc/Y and Sc/Yb concentration ratios. Principal component analysis effectively segregates "Eglouvi" lentils into a distinct cluster. Soft Independent Modelling of Class Analogy (SIMCA) successfully models "Eglouvi" lentils. Significant enhancement in model specificity was achieved upon inclusion of Sc/Y and Sc/Yb concentration ratios as additional variables. The model is capable of detecting adulteration in blends of Eglouvi lentils, with a minimum rejection threshold of 4.6% w/w for Greek lentil adulterants and 6.0% w/w for imported lentil adulterants.

A compact Fourier-transform near-infrared spectrophotometer and chemometrics for characterizing a comprehensive set of seized ecstasy samples.

A comprehensive data set of ecstasy samples containing MDMA (N-methyl-3,4-methylenedioxyamphetamine) and MDA (3,4-methylenedioxyamphetamine) seized by the Brazilian Federal Police was characterized using spectral data obtained by a compact, low-cost, near-infrared Fourier-transform based spectrophotometer. Qualitative and quantitative characterization was accomplished using soft independent modeling of class analogy (SIMCA), linear discriminant analysis (LDA) classification, discriminating partial least square (PLS-DA), and regression models based on partial least square (PLS). By applying chemometric analysis, a protocol can be proposed for the in-field screening of seized ecstasy samples. The validation led to an efficiency superior to 96 % for ecstasy classification and estimating total actives, MDMA, and MDA content in the samples with a root mean square error of validation of 4.4, 4.2, and 2.7 % (m/m), respectively. The feasibility and drawbacks of the NIR technology applied to ecstasy characterization and the compromise between false positives and false negatives rate achieved by the classification models are discussed and a new approach to improve the classification robustness was proposed considering the forensic context.

Detection of impurities in m-cresol purple with Soft Independent Modeling of Class Analogy for the quality control of spectrophotometric pH measurements in seawater.

Accurate spectrophotometric pH measurements in seawater are critical to documenting long-term changes in ocean acidity and carbon chemistry, and for calibration of autonomous pH sensors. The recent development of purified indicator dyes greatly improved the accuracy of spectrophotometric pH measurements by removing interfering impurities that cause biases in pH that can grow over the seawater pH range to >0.01 above pH 8. However, some batches of purified indicators still contain significant residual impurities that lead to unacceptably large biases in pH for oceanic and estuarine climate quality measurements. While high-performance liquid chromatography (HPLC) is the standard method for verifying dye purity, alternative approaches that are simple to implement and require less specialized equipment are desirable. We developed a model to detect impurities in the pH indicator m-cresol purple (mCP) using a variant of the classification technique Soft Independent Modeling of Class Analogy (SIMCA). The classification model was trained with pure mCP spectra (350 nm to 750 nm at 1 nm resolution) at pH 12 and tested on independent samples of unpurified and purified mCP with varying levels of impurities (determined by HPLC) and measured on two different spectrophotometers. All the dyes identified as pure by the SIMCA model were sufficiently low in residual impurities that their apparent biases in pH were < 0.002 in buffered artificial seawater solutions at a salinity of 35 and over a pH range of 7.2 to 8.2. Other methods that can also detect residual impurities relevant to climate quality measurements include estimating the impurity absorption at 434 nm and assessing the apparent pH biases relative to a reference purified dye in buffered solutions or natural seawater. Laboratories that produce and distribute purified mCP should apply the SIMCA method or other suitable methods to verify that residual impurities do not significantly bias pH measurements. To apply the SIMCA method, users should download the data and model developed in this work and measure a small number of instrument standardization and model validation samples. This method represents a key step in the development of a measurement quality framework necessary to attain the uncertainty goals articulated by the Global Ocean Acidification Observing Network (GOA-ON) for climate quality measurements (i.e., ±0.003 in pH).

A one-class classification approach for authentication of specialty coffees by inductively coupled plasma mass spectroscopy (ICP-MS).

Due to the lucrative nature of specialty coffees, there have been instances of adulteration where low-cost materials are mixed in to increase the overall volume, resulting in illegal profit. A widely used and recommended approach to detect possible adulteration is the application of one-class classifiers (OCC), which only require information about the target class to build the models. Thus, this work aimed to identify adulterations in specialty coffees with low-quality coffee using multielement analysis determined by ICP-MS and to evaluate the performance of one-class classifiers (dd-SIMCA, OCRF, and OCPLS). Therefore, authentic specialty coffee samples were adulterated with low-quality coffee in 25 % to 75 % (w/w) proportions. Samples were subjected to acid decomposition for analysis by ICP-MS. OCPLS method presented the best performance to detect adulterations with low-quality coffee in specialty coffees, showing higher specificity (SPE = 100 %) and reliability rate (RLR = 94.3 %).

Multivariate curve resolution-soft independent modelling of class analogy (MCR-SIMCA).

BACKGROUND: Various classification, class modeling, and clustering techniques operate within abstract spaces, utilizing Principal Components (e.g., Linear Discriminant Analysis (LDA), Principal Component Analysis (PCA)) or latent variable spaces (e.g., Partial Least Squares Discriminant Analysis (PLS-DA)). It's important to note that PCA, despite being a mathematical tool, defines its Principal Components under certain mathematical constraints, it has a wide range of applications in the analysis of real-world systems. In this research, we assess the viability of employing the Multivariate Curve Resolution (MCR) subspace within class modeling techniques, as an alternative to the PC subspace. (92). RESULTS: This study evaluates the use of the MCR subspace in class modeling methods, specifically in tandem with soft independent modeling of class analogy (SIMCA), to investigate the advantages of employing the meaningful physico-chemical subspace of MCR over the mathematical subspace of PCA. In the MCR-SIMCA strategy, the model is constructed by applying MCR to training samples from a target class. The MCR model effectively partitions the data into two smaller sub-matrices: the contribution matrix and the corresponding response matrix. In the next step, the contribution matrix resulting from the decomposition of the training set develops a distance plot (DP). First, the theory of the MCR-SIMCA model is discussed in detail. Next, two real experimental datasets were analyzed, and their performance was compared with the DD-SIMCA model. In most cases, the results were as good as or even more satisfactory than those obtained with the DD-SIMCA model. (146). SIGNIFICANCE: The suggested class modeling method presents a promising avenue for the analysis of real-world natural systems. The study's results emphasize the practical utility of the MCR approach, underscoring the significance of the MCR subspace advantages over the PCA subspace. (39).

Authenticity identification of high - Temperature Daqu Baijiu through multi-channel visual array sensor of organic dyes combined with smart phone App.

High - temperature Daqu Baijiu faces a challenge from illegal adulteration of high-grade Baijiu bottles with low-grade Baijiu, affecting its quality and value. This study developed a rapid identification method for high temperature Daqu Baijiu with the same aroma type using a four-channel visual array sensor and detection of color changes caused by competition coordination with Zn2+ and color-changing organic dyes. The array sensor demonstrated high stability and repeatability in targeting flavor components and achieved 97.78 % or more accuracy combined with DD-SIMCA model in detecting adulteration across the Baijiu with same aroma type. The results of GC-MS and Quantum Chemical Calculation showed that esters, acids, and pyrazines played a crucial role. The smart phone App could quickly identify the authenticity of Baijiu with accuracy achieved 93 %. This research provides a foundation for rapid and reliable assessment of Baijiu quality and authenticity, enabling the industry to combat fraudulent practices effectively.

Simple analytical method using ultraviolet spectral dataset and chemometrics for the authentication of Indonesian specialty ground roasted coffee with different botanical and geographical indications.

The possible application of a simple analytical method based on a UV (ultraviolet) spectral dataset coupled with SIMCA (soft independent modeling of class analogy) for authentication of Indonesian specialty ground roasted coffee with different botanical and geographical indications (GIs) was demonstrated. Three types of Indonesian specialty ground roasted coffee were used: GIs arabica coffee from Gayo Aceh (96 samples), GIs liberica coffee from Meranti-Riau (119 samples), and GIs robusta coffee from Lampung (150 samples) with 1 g weight of each sample. All samples were extracted using hot distilled water and 3 mL aqueous filtered samples were pipetted into a 10 mm quartz cell. Original UV spectral datasets were recorded in the range of 190-399 nm. The pre-processed spectral dataset was generated using three simultaneous different preprocessing techniques: moving average smoothing with 11 segments, standard normal variate (SNV), and Savitzky-Golay (SG) first derivative with window size and polynomial order value of 11 and 2. The supervised classification based on the SIMCA method was applied for preprocessed selected spectral data (250-399 nm). The PCA data showed that GIs coffee with different botanical and geographical indications can be well separated. The SIMCA classification was accepted with 100 % of correct classification (100 % CC). This dataset demonstrated the potential use of UV spectroscopy with chemometrics to perform simple and affordable authentication of Indonesian specialty ground roasted coffee with different botanical and geographical indications (GIs).

Identification of Variety and Prediction of Chemical Composition in Cocoa Beans (Theobroma cacao L.) by FT-MIR Spectroscopy and Chemometrics.

Cocoa is rich in polyphenols and alkaloids that act as antioxidants, anticarcinogens, and anti-inflammatories. Analytical methods commonly used to determine the proximal chemical composition of cocoa, total phenols, and antioxidant capacity are laborious, costly, and destructive. It is important to develop fast, simple, and inexpensive methods to facilitate their evaluation. Chemometric models were developed to identify the variety and predict the chemical composition (moisture, protein, fat, ash, pH, acidity, and phenolic compounds) and antioxidant capacity (ABTS and DPPH) of three cocoa varieties. SIMCA model showed 99% reliability. Quantitative models were developed using the PLS algorithm and favorable statistical results were obtained for all models: 0.93 < R2c < 0.98 (R2c: calibration determination coefficient); 0.03 < SEC < 4.34 (SEC: standard error of calibration). Independent validation of the quantitative models confirmed their good predictive ability: 0.93 < R2v < 0.97 (R2v: validation determination coefficient); 0.04 < SEP < 3.59 (SEP: standard error of prediction); 0.08 < % error < 10.35). SIMCA model and quantitative models were applied to five external cocoa samples, obtaining their chemical composition using only 100 mg of sample in less than 15 min. FT-MIR spectroscopy coupled with chemometrics is a viable alternative to conventional methods for quality control of cocoa beans without using reagents, and with the minimum sample preparation and quantity.

Authentication and discrimination of new Brazilian Canephora coffees with geographical indication using a miniaturized near-infrared spectrometer.

New Brazilian Canephora coffees (Conilon and Robusta) of high added value from specific origins have been protected by geographical indication to guarantee their origin and quality. Recently, benchtop near-infrared (NIR) spectroscopy combined with chemometrics has demonstrated its usefulness to discriminate them. It was the first study, however, and therefore the possibility exists to develop a new portable NIR method for this purpose. This work assessed a miniaturized NIR as a cheaper spectrometer to discriminate and authenticate new Brazilian Canephora coffees with certified geographical origins and to differentiate them from specialty Arabica. Discriminant chemometric and class modeling techniques have been applied and have obtained good predictive ability on external test sets. In addition, models with similar classification purpose were compared with those obtained in previous research carried out with benchtop NIR for the same samples, obtaining comparable results. In this context, the portable method was used as a laboratory technique and has the advantage of being cheaper than benchtop NIR spectrometer. Furthermore, it brings a high possibility to be implemented in small coffee cooperatives, industries or control agencies in the future that do not have high economic resources.

Gasoline discrimination using infrared spectroscopy and virtual samples based on measurement uncertainty.

In a previous work, we proposed a methodology for pair-wise discrimination of gasoline samples by creating virtual samples based on physicochemical assays or distillation curves. Satisfactory results were achieved, although specialist and specific apparatus (not commonly available at police laboratories) were required. The present study goes a step further and for the first time investigates the possibility of infrared (IR) spectroscopy to enable a virtual samples-based methodology for comparison of gasoline samples in pairs. IR spectroscopy feasibility for in situ applications is attractive for forensic investigations. The performances of one handheld NIR device and one dual-range (FT-NIR and FT-IR) benchtop spectrometer were evaluated. The estimation of uncertainty in infrared spectral measurement (needed to generate virtual samples) is barely discussed in literature. So far, there are no literature reports describing quantification and comparison of measurement uncertainties for the spectral acquisitions evaluated here, especially regarding their use for generating virtual samples. A stepwise procedure to quantify uncertainties associated with IR spectral acquisition, at each wavenumber, is described. This method can be useful for understanding both the sources of variability in IR measurements and the system under investigation. Uncertainty estimation was based on experimental data and considered intermediate precision, repeatability and variations in sample temperature as sources of variability. Virtual samples were employed in a discrimination approach using SIMCA models. Results for portable NIR, FT-NIR and FT-IR data sets showed complete discrimination for 96.3%, 93.4% and 93.7% of the 1431 pairs of gasoline samples evaluated, respectively. These results were comparable and similar to those obtained for the physicochemical properties data set (95.7%), although slightly inferior to the result obtained for distillation curves (99.2%). Using IR non-destructive methods in this case could enable faster investigations and simpler analysis, especially for the low-cost handheld spectrometer. In a screening approach, atmospheric distillation assays can be employed only if infrared techniques are not capable of distinguishing the samples subject to comparison. In this work, a pair of samples was considered to be completely discriminated only when a null false positive error (FPR) was achieved, although a more flexible criterium may be acceptable in practice. Finally, the methodology could be extended to other applications where sample comparison is important.

Class modelling by Soft Independent Modelling of Class Analogy: why, when, how? A tutorial.

This article contains a comprehensive tutorial on classification by means of Soft Independent Modelling of Class Analogy (SIMCA). Such a tutorial was conceived in an attempt to offer pragmatic guidelines for a sensible and correct utilisation of this tool as well as answers to three basic questions: "why employing SIMCA?", "when employing SIMCA?" and "how employing/not employing SIMCA?". With this purpose in mind, the following points are here addressed: i) the mathematical and statistical fundamentals of the SIMCA approach are presented; ii) distinct variants of the original SIMCA algorithm are thoroughly described and compared in two different case-studies; iii) a flowchart outlining how to fine-tune the parameters of a SIMCA model for achieving an optimal performance is provided; iv) figures of merit and graphical tools for SIMCA model assessment are illustrated and v) computational details and rational suggestions about SIMCA model validation are given. Moreover, a novel Matlab toolbox, which encompasses routines and functions for running and contrasting all the aforementioned SIMCA versions is also made available.

Multi-block DD-SIMCA as a high-level data fusion tool.

Multi-block classification method based on the Data Driven Soft Independent Modeling of Class Analogy (DD-SIMCA) is presented. A high-level data fusion approach is used for the joint analysis of data collected with the help of different analytical instruments. The proposed fusion technique is very simple and straightforward. It uses a Cumulative Analytical Signal which is a combination of outcomes of the individual classification models. Any number of blocks can be combined. Although the high-level fusion eventually leads to a rather complex model, the analysis of partial distances makes it possible to establish a meaningful relationship between the classification results and the influence of individual samples and specific tools. Two real world examples are used to demonstrate the applicability of the multi-block algorithm and the consistency of the multi-block method with its predecessor, a conventional DD-SIMCA.

Green Chemometric-Assisted Characterization of Common and Black Varieties of Celery.

Celery (Apium graveolens L., var. Dulce), is a biennial herbaceous plant belonging to the Apiaceae family, cultivated in humid soils in the Mediterranean basin, in Central-Southern Europe, and in Asia. Despite its wide diffusion and although it is well-known that cultivar/origin strongly influences plant composition, only a few studies have been carried out on the different types of celery. The present work aims to investigate four different Italian types of celery (two common, Elne and Magnum celery, and two black, Torricella Peligna Black and Trevi Black celery), and to test, whether the combination of FT-IR spectroscopy and chemometrics allows their ecotype discrimination. The peculiarity of this study lies in the fact that all the analyzed celeries were grown in the same experimental field under the same soil and climate conditions. Consequently, the differences captured by the FT-IR-based tool are mainly imputable to the different ecotypes. In order to achieve this goal, FT-IR profiles were handled by two diverse classifiers: sequential preprocessing through ORThogonalization (SPORT) and soft independent modeling by class analogy (SIMCA). Eventually, the highest classification rate (90%, on an external set of 100 samples) has been achieved by SPORT.

Adaptive Spectral Model for abnormality detection based on physiological status monitoring of dairy cows.

This research study developed milk spectral data-driven approach, called Adaptive Spectral Model for Abnormality Detection - ASMAD, for detection of physiological abnormalities of individual dairy cows. The algorithm is based on the logic that milk spectra of each individual cow is highly animal-specific, which means it could be used as a respective individual marker for identification. When the algorithm fails to recognize the milk spectra as coming from a certain animal, instead of being treated as a mistake, this outcome is accepted as a deviation of the respective individual marker, and a potential indicator of abnormal physiological state. For the purpose of ASMAD development, near infrared spectra of milk of seven dairy cows have been collected daily during 1-year period. The abnormality detection model is built using supervised recognition method Soft Independent Modeling of Class Analogies - SIMCA, and optimized with respect to spectral pre-processing, choice of the wavelength region and size of the time-window when developing the adaptive model. The sensitivity and specificity of ASMAD were dependent on the animal, and in the ranges 40.00-64.29% and 87.23-98.86%, respectively. Considering significant level of day-to-day spectral variation and multitude of physiological and environmental factors influence on milk constituents and spectra, these results represent a significant potential for creating a health-status monitoring and detection of abnormal physiological states in dairy animals. The adaptive modeling based on the time series of spectral data collected from the individual organism utilized in this work for monitoring physiological status and abnormality detection in dairy cows, has a good potential to be used for similar purposes in other animals and humans.

Characterization and Authentication of "Ricotta" Whey Cheeses through GC-FID Analysis of Fatty Acid Profile and Chemometrics.

The fatty acid (FA) profiles of 240 samples of ricotta whey cheese made from sheep, goat, cow, or water buffalo milk were analyzed by gas-chromatography (GC). Then, sequential preprocessing through orthogonalization (SPORT) was used in order to classify samples according to the nature of the milk they were made from. This strategy achieved excellent results, correctly classifying 77 (out of 80) validation samples. Eventually, since 36 (over 114) sheep ricotta whey cheeses were PDO products, a second classification problem, finalizing the discrimination of PDO and Non-PDO dairies, was faced. In this case, two classifiers were used, SPORT and soft independent modelling by class analogy (SIMCA). Both approaches provided more than satisfying results; in fact, SPORT properly assigned 63 (of 65) test samples, whereas the SIMCA model accepted 14 PDO individuals over 15 (93.3% sensitivity) and correctly rejected all the other samples (100.0% specificity). In conclusion, all the tested approaches resulted as suitable for the two fixed purposes. Eventually, variable importance in projection (VIP) analysis was used to understand which FAs characterize the different categories of ricotta. Among the 22 analyzed compounds, about 10 are considered the most relevant for the solution of the investigated problems.

Optimizing the soft independent modeling of class analogy (SIMCA) using statistical prediction regions.

The ultimate goal of a one-class classifier like the "rigorous" soft independent modeling of class analogy (SIMCA) is to predict with a certain confidence probability, the conformity of future objects with a given reference class. However, the SIMCA model, as currently implemented often suffers from an undercoverage problem, meaning that its observed sensitivity often falls far below the desired theoretical confidence probability, hence undermining its intended use as a predictive tool. To overcome the issue, the most reported strategy in the literature, involves incrementing the nominal confidence probability until the desired sensitivity is obtained in cross-validation. This article proposes a statistical prediction interval-based strategy as an alternative strategy to properly overcome this undercoverage issue. The strategy uses the concept of predictive distributions sensu stricto to construct statistical prediction regions for the metrics. Firstly, a procedure based on goodness-of-fit criteria is used to select the best-fitting family of probability models for each metric or its monotonic transformation, among several plausible candidate families of right-skewed probability distributions for positive random variables, including the gamma and the lognormal families. Secondly, assuming the best-fitting distribution, a generalized linear model is fitted to each metric data using the Bayesian method. This method enables to conveniently estimate uncertainties about the parameters of the selected distribution. Propagating these uncertainties to the best-fitting probability model of the metric enables to derive its so-called posterior predictive distribution, which is then used to set its critical limit. Overall, the evaluation of the proposed approach on a diversity of real datasets shows that it yields unbiased and more accurate sensitivities than existing methods which are not based on predictive densities. It can even yield better specificities than the strategy that attempts to improve sensitivities of existing methods by "optimizing" the type 1 error, especially in low sample sizes' contexts.

Multivariate Prognostic Model for Predicting the Outcome of Critically Ill Patients Using the Aromatic Metabolites Detected by Gas Chromatography-Mass Spectrometry.

A number of aromatic metabolites of tyrosine and phenylalanine have been investigated as new perspective markers of infectious complications in the critically ill patients of intensive care units (ICUs). The goal of our research was to build a multivariate model for predicting the outcome of critically ill patients regardless of the main pathology on the day of admission to the ICU. Eight aromatic metabolites were detected in serum using gas chromatography-mass spectrometry. The samples were obtained from the critically ill patients (n = 79), including survivors (n = 44) and non-survivors (n = 35), and healthy volunteers (n = 52). The concentrations of aromatic metabolites were statistically different in the critically ill patients and healthy volunteers. A univariate model for predicting the outcome of the critically ill patients was based on 3-(4-hydroxyphenyl)lactic acid (p-HPhLA). Two multivariate classification models were built based on aromatic metabolites using SIMCA method. The predictive models were compared with the clinical APACHE II scale using ROC analysis. For all of the predictive models the areas under the ROC curve were close to one. The aromatic metabolites (one or a number of them) can be used in clinical practice for the prognosis of the outcome of critically ill patients on the day of admission to the ICU.