Partial Least Squares-linear Discriminant Analysis Research Articles

Multi-block approach for the characterization and discrimination of Italian chickpeas landraces

An untargeted characterization was carried out to attempt the geographical discrimination of three high-valued Italian chickpeas (Cicer arietinum L.), harvested in 2019 in three traditional and relatively close production areas: Navelli (Abruzzo, Central Italy), Cicerale (Campania, South Italy), and Valentano (Lazio, Central Italy). The present study aimed to develop and validate a potentially non-destructive and routine-compatible method for the geographical traceability of chickpea landraces of high traditional value. The outer part of 146 kernels belonging to the three varieties was analysed by Attenuated Total Reflectance-Fourier Transform-Mid Infrared (ATR-FT-MIR) and FT-Near Infrared (NIR) spectroscopies. Eventually, each sample was cut in two, and the cross-sections (the internal and external parts) were analysed by the two spectroscopic techniques. Spectral information was organized in four data blocks (MIRout, MIRin, NIRout, and NIRin), and single-block Partial Least Squares-Linear Discriminant Analysis (PLS-LDA) was applied. Accurate results were obtained from the single-block-processing of the spectroscopic profiles for the outer kernel part (MIRout and NIRout) that, combined with interesting outcomes from a preliminary class modelling approach suggest the real possibility of implementing a non-destructive authentication method. Notwithstanding, Sequential and Orthogonalized (SO)-PLS-LDA and SO-Covariance selection (Covsel)-LDA were applied to interpret better the information in the four collected data blocks. In this context, VIP (Variable Importance in Projection) analysis was performed to identify the significant variables, leading to a direct chemical interpretation of the classification models.

Real-time in vivo cancer staging of nasopharyngeal carcinoma patients with rapid fiberoptic Raman endoscopy

Cancer staging is important to guide treatment and for prognostication. This work aims to demonstrate the ability of rapid fiberoptic Raman endoscopy for real-time in vivo cancer staging of nasopharyngeal cancer (NPC) patients. We interrogate 278 tissue sites on the primary NPC with different cancer stages from 61 NPC patients and 50 healthy volunteers using rapid fiberoptic Raman endoscopy examination. Distinct Raman spectral differences of NPC at different cancer stages are observed through simultaneous fingerprint and high-wavenumber (FP/HW) Raman spectral measurements, reflecting the biomolecular differences of NPC tumor across various cancer stages. Raman staging model is established based on in vivo FP/HW tissue Raman spectra together with partial-least-squares linear-discriminant-analysis (PLS-LDA) and leave-one-tissue-site-out cross-validation (LOOCV). In vivo FP/HW Raman endoscopy provides an overall diagnostic accuracy of 92.81% for identifying different stages of NPC (i.e., NPC stage I&II and NPC stage III&IV) from normal nasopharynx. Specifically, the diagnostic sensitivity of 91.18% is obtained for identifying NPC stage I& II; and the sensitivity of 93.04% is achieved for classifying NPC stage III&IV from normal tissue. The key tissue biomolecular variations responsible for different NPC stages have been identified using biomolecular Raman modeling developed based on non-negative linear regression. The essential biomolecules (chondroitin sulfate, glucose, hemoglobin, oleic acid and triolein) are uncovered from the Raman spectra of NPC tissues through biomolecular modeling with significant variations (p < 0.05) between early-stage NPC (stage I and stage II) and late-stage NPC patients (stage III and stage IV). Our pivotal work demonstrates for the first time that fiberoptic Raman endoscopy is a robust analytical tool for real-time in vivo NPC staging in clinical settings.

A regularised logistic regression model with structured features for classification of geographical origin in olive oils

Geographical origin of extra virgin olive oil is a factor that consumers may take into account when making purchasing decisions. Oils that are labelled to be from regions famous for olive cultivation may be assumed to be of higher quality. However, difficulties in the authentication of the geographical origin of olive oils arise due to the similarity in chemical compositions of the oils involved. Fourier-transform infrared (FTIR) spectroscopy has been found to be a viable technology for the classification of oil samples by geographical origin. However, classical methods involving dimension reduction before model fitting usually yield models that are more challenging to interpret. Sparse fused group lasso logistic regression (SFGL-LR) is used with FTIR spectroscopic data to discriminate between Greek and non-Greek organic extra-virgin olive oils. The prediction performance is also compared with that obtained by partial least squares linear discriminant analysis (PLS-LDA). While both methods give comparable good prediction performance, with more than 90% accuracy in classification, the SFGL-LR model demonstrates improvements in the interpretability of the model coefficients.

Optical photothermal infrared spectroscopy and discrete wavenumber imaging for high content screening of single cells.

Major drawbacks of direct mid-infrared spectroscopic imaging of single cells in an aqueous buffer are strong water absorption, low resolution typically above 10 μm, and Mie scattering effects. This study demonstrates how an indirect detection principle can overcome these drawbacks using the optical photothermal infrared (O-PTIR) technique for high-resolution discrete wavenumber imaging and fingerprint spectroscopy of cultivated cells as a model system in a simple liquid sample chamber. The O-PTIR spectra of six leukemia- and cancer-derived cell lines showed main IR bands near 1648, 1547, 1447, 1400, 1220, and 1088 cm-1. Five spectra of approximately 260 single cells per cell type were averaged, the O-PTIR data set was divided into leukemia-derived cells (THP-1, HL 60, Jurkat, and Raji) and cancer cells (HeLa and HepaRG), and partial least squares linear discriminant analysis (PLS-LDA) was applied in the spectral range 800-1800 cm-1 to train three classification models. A leukemia versus cancer cell model showed an accuracy of 90.0%, the HeLa versus HepaRG cell model had an accuracy of 95.4%, and the model for the distinction of leukemia cells had an accuracy of 75.4%. IR bands in linear discriminants (LDs) of the models were correlated with second derivative spectra that resolved more than 25 subbands. The IR and second derivative spectra of proteins, DNA, RNA and lipids were collected as references to confirm band assignments. O-PTIR images of single cells at a 200 nm step size were acquired at 1086, 1548, and 1746 cm-1 to visualize the nucleic acid, protein, and lipid distribution, respectively. Variations in subcellular features and in the lipid-to-protein and nucleic acid-to-protein ratios were identified that were consistent with biomolecular information in LDs. In conclusion, O-PTIR can provide high-quality spectra and images with submicron resolution of single cells in aqueous buffers that offer prospects in high-content screening applications.

Molecular mechanism of the prefrontal cortex involved in electroacupuncture against myocardial ischemia based on metabolomics

To explore the effects of electroacupuncture(EA) on metabolic patterns of the prefrontal cortex in rats with acute myocardial ischemia. Eighteen SD rats were randomly divided into sham group, model group and EA group, with 6 rats in each group. Rats in the model and EA groups were subject to acute myocardial ischemia by ligation of the left anterior descending coronary artery. For the EA group, EA stimulation (1 mA, 2 Hz/15 Hz, 30 min) was applied to "Shenmen"(HT7) -"Tongli"(HT5) once a day for 3 consecutive days. The histopathological changes of myocardial tissue and levels of ischemia modified albumin (IMA) in serum were determined by HE staining and ELISA, respectively. The LC-MS/MS technique was used to characterize the metabolic profiling of the prefrontal cortex. The differentially expressed metabolites were screened by principal component analysis(PCA) and partial least squares-linear discriminant analysis (PLS-LDA), and subsequently Kyoto encyclopedia of genes and genomes (KEGG) metabolic pathway enrichment analysis was performed. Compared with the sham group, the myocardial fibers were disordered and fractured, and content of serum IMA was significantly increased in the model group (P<0.01), which, however, were significantly decreased in the EA group (P<0.01). With PCA and PLS-LDA, there were 18 differential metabolites between the model and sham groups. Forty-eight differential metabolites were emerged between the EA and model groups. Three metabolites associated to the sphingolipid metabolism were reversed by EA stimulation, as indicated by KEGG. The molecular mechanism of EA against myocardial ischemia is partially mediated by regulating sphingolipid-related metabolites in the prefrontal cortex.

Raman Spectral Characterization of Urine for Rapid Diagnosis of Acute Kidney Injury.

Acute kidney injury (AKI) is a common syndrome characterized by various etiologies and pathophysiologic processes that deteriorate kidney function. The aim of this study is to identify potential biomarkers in the urine of non-acute kidney injury (non-AKI) and AKI patients through Raman spectroscopy (RS) to predict the advancement in complications and kidney failure. Selected spectral regions containing prominent peaks of renal biomarkers were subjected to partial least squares linear discriminant analysis (PLS-LDA). This discriminant analysis classified the AKI patients from non-AKI subjects with a sensitivity and specificity of 97% and 100%, respectively. In this study, the RS measurements of urine specimens demonstrated that AKI had significantly higher nitrogenous compounds, porphyrin, tryptophan and neopterin when compared with non-AKI. This study’s specific spectral information can be used to design an in vivo RS approach for the detection of AKI diseases.

Remote sensing of invasive alien wattle using image texture ratios in the low-lying Midlands of KwaZulu-Natal, South Africa

The presence of invasive alien plants (IAPs) cause dynamic structural and ecological change, since they encroach on the productivity of ecosystems with their high-water consumption, ability to rapidly spread and reduce natural biodiversity. Remote sensing techniques utilised for early detection and mapping of invasive alien wattle (Acacia mearnsii) are pivotal towards understanding the spatial distribution of IAPs for control and management. In this study, we assessed the utility of image texture ratios computed from a 6 m SPOT-6 image combined with the sparse partial least squares discriminant analysis (SPLS-DA) algorithm to detect and map invasive alien wattle, based on field information related to the presence of these invasive species. We specifically tested the performance of single band and multi-band texture metrices, as well as vegetation indices in detecting and mapping invasive alien wattle. The results revealed that the image texture ratio model (overall accuracy = 74%; kappa statistic = 0.70) outcompeted the single image texture model (overall accuracy = 68%; kappa statistic = 0.65) and vegetation indices model (overall accuracy = 62%; kappa statistic = 0.59). The most significant texture filters selected by the SPLS-DA model were correlation, second moment and homogeneity, which were predominantly computed from the NIR and red bands. In addition, the 5 × 5 moving window was the most optimal window-size for detecting the alien invasive. Moreover, the SPLS-DA model outcompeted the partial least squares linear discriminant analysis (PLS-LDA) model in detecting invasive alien wattle over the entire study area. Overall, this study underscores the potential of image texture ratios in concert with SPLS-DA in effectively detecting and mapping invasive alien wattle.

Study on anti-osteoporosis effect of Eucommiae Cortex based on GC-MS metabonomics

Based on GC-MS metabolomics and biochemical index analysis, the mechanism of bone mass loss in osteoporosis and the evaluation of anti-osteoporosis in Eucommiae Cortex were studied. The OVX rats model was established by bilateral ovariectomized. The routine indexes such as BMC, BMD, BGP and TRAP5 b were determined. The GC-MS technique was used to analyze the metabolism profile of serum samples between the control group, model group and medicine group, and multiple statistical analysis methods including principal component analysis(PCA), partial least squares-linear discriminant analysis(PLS-LDA) and subwindow rearrangement analysis(SPA) were used to screen and identify biomarkers. Five metabolites were selected as potential biomarkers, glycine, lysine, tryptophan, docosahexaenoic acid and glucose. Except for the significant increase of tryptophan in serum of OVX rats, the other four metabolites were significantly decreased. Moreover, the five biomarkers of the medicine group had a trend of returning to rats in control group. The significantly altered metabolite levels indicated that Eucommiae Cortex may relieve the symptoms of osteoporosis by regulating amino acid metabolism and oxidative stress.

Classification approaches for sorting maize (Zea mays subsp. mays) haploids using single‐kernel near‐infrared spectroscopy

AbstractDoubled haploids (DHs) are an important breeding tool for creating maize inbred lines. One bottleneck in the DH process is the manual separation of haploids from among the much larger pool of hybrid siblings in a haploid induction cross. Here, we demonstrate the ability of single‐kernel near‐infrared reflectance spectroscopy (skNIR) to identify haploid kernels. The skNIR is a high‐throughput device that acquires an NIR spectrum to predict individual kernel traits. We collected skNIR data from haploid and hybrid kernels in 15 haploid induction crosses and found significant differences in multiple traits such as percent oil, seed weight, or volume, within each cross. The two kernel classes were separated by their NIR profile using Partial Least Squares Linear Discriminant Analysis (PLS‐LDA). A general classification model, in which all induction crosses were used in the discrimination model, and a specific model, in which only kernels within a specific induction cross, were compared. Specific models outperformed the general model and were able to enrich a haploid selection pool to above 50% haploids. Applications for the instrument are discussed.

Understanding the learning mechanism of convolutional neural networks in spectral analysis

Deep learning approaches, especially convolutional neural network (CNN) models, have achieved excellent performances in vibrational spectral analysis. The critical drawback of the CNN approach is the lack of interpretation, and it is regarded as a black box. Interpreting the learning mechanism of chemometric models is critical for intuitive understanding and further application. In this study, an interpretable CNN model with a global average pooling layer is presented for Raman and mid-infrared spectral data analysis. A class activation mapping (CAM)-based approach is leveraged to visualize the active variables in the whole spectrum. The visualization of active variables shows a discriminative pattern in which the most contributed variables peaked around theoretical chemical characteristic bands. The visualization of the feature maps by three convolutional layers demonstrates the data transformation pipeline and how the CNN model hierarchically extracts informative spectral features. The first layer acts as a Savitzky-Golay filter and learns spectral shape characteristics, while the second layer learns enhanced patterns from typical spectral peaks on a few correlated variables. The third layer shows stable activations on critical spectral peaks. A partial least squares - linear discriminant analysis (PLS-LDA) model is presented for comparison on classification accuracy and model interpretation. The CNN model yields mean classification accuracies of 99.01 and 100% for E. coli and meat datasets on the test set, while the PLS-LDA models obtain accuracies of 98.83 and 100%. Both the CNN and PLS-LDA models demonstrate stable patterns on active variables while CNN models are more stable than PLS-LDA models on classification performances for various dataset partitions with Monte-Carlo cross-validation.

On-line species identification of green hem-fir timber mix based on near infrared spectroscopy and chemometrics

In this study, a method suitable for on-line rapid species classification of western hemlock and amabilis fir (hem-fir) green mix of timber was developed and tested with the use of near infrared spectroscopy (NIRS) and chemometrics. The spectra of 600 wood specimens obtained from each species were collected over the wavelength range of 350–2500 nm. They were thereafter pretreated by smoothing; first derivative, second derivative and standard normal variate and calibration models were developed using the wavelength range of 800–1800 nm by partial least squares-linear discriminant analysis (PLS-LDA) and least squares-support vector machines (LS-SVM). The effects of wood surface (transverse, tangential and radial), wood zone (heartwood and sapwood) and sample moving speed (0, 0.5 and 1 m/s) were also explored. LS-SVM is superior to PLS-LDA in terms of classification performance at moving conditions. The best results were obtained using the LS-SVM method when spectra were collected on the transverse surface at a speed of 1 m/s and pretreated by smoothing. The sensitivity, specificity, and accuracy for both calibration and validation sets were all 100%. Collectively, NIR spectroscopy combined with LS-SVM is capable of on-line species separation of green hem-fir mix prior to wood drying.

Rapid detection and visualization of slight bruise on apples using hyperspectral imaging

ABSTRACTDuring harvesting and storage, slight bruises on apple surface caused by impact, compression, vibration, or abrasion are inevitable. To find an appropriate method to identify the bruised apples at five stages (1 min, 1 day, 2 days, 3 days and 4 days after bruising), 108 Fuji apples were collected as samples. Hyperspectral images of apples covering the wavelength between 400 and 1000 nm were acquired by the SOC710-VP hyperspectral imaging system. The standard normal variate (SNV) method was utilized for smoothing and denoising of the original hyperspectral data. Classification models, including Extreme Learning Machine (ELM), Partial Least Squares Linear Discriminant Analysis (PLS-DA) and Classification and Regression Tree (CART), coupled with a variable selection method named Minimum Redundancy Maximum Relevance (mRMR), were built to identify the bruised apples. The results showed that the ELM models exhibited the best classification capability, with the mean correct classification rate of 95.97%. The bruised samples are easier to be identified over time. Minimum noise fraction (MNF) method was implemented to classify the bruised region of apples based on the selected wavelengths. The overall classification accuracy of MNF is 92.9%, which indicates that MNF is an effective method for identifying bruised regions of apples.

Fourier transform infrared spectroscopy and chemometrics for the discrimination of paper relic types.

The paper relic identification is a pending issue to be resolved in the field of cultural heritage. As we all known, heritage paper has significant importance in archaeological research. Nowadays, there are a variety of research methodologies focuses on the analysis of inks for dating documents. While the paper analysis attained little attention. This work is to explore the non-destructive application of ATR-FTIR technique in discrimination of paper relics. 15 types of paper spectra were collected by ATR-FTIR, which wavenumber range were range from 4000 to 650 cm-1. And the moving average smoothing and normalization was used for pretreatment analysis. Five different classification algorithms, principal component analysis-linear discriminant analysis (PCA-LDA), partial least squares discriminant analysis (PLS-DA), soft independent modeling of class analogy (SIMCA), least squares-support vector machine (LS-SVM), partial least squares-linear discriminant analysis (PLS-LDA) were selected to classify the types of paper. PLS-LDA and LS-SVM are effective techniques with 100% classification accuracy. PCA-LDA, PLS-DA and SIMCA give accuracy of 98.67%, 97.33% and 95.56%, respectively. The present experiment suggested that ATR-FTIR combining with chemometrics will be highly useful in paper identification of cultural heritage.

Subcutaneous Metabolites Involving Acupoint Sensitization Induced by Myocardial Ischemia and Acupuncture Stimulation in Rabbits

The present study aimed at observing the profile of metabolites in sensitized acupoints induced by myocardial ischemia (MI) and the effect of acupuncture intervention on the changes of the metabolites so as to explore the material basis of acupoint sensitization. A total of 20 New Zealand rabbits were randomly and equally divided into a control group and a model group. The MI model was established by occlusion of the anterior descending branch of the left coronary artery with a controllable air balloon inflation method for 5 min/time, twice a day (4-hours' interval) for continuous 5 days (the first stage of MI). After one day's rest, another 5 days' occlusion was conducted again (the second stage of MI) in the same way. The rabbits of the control group were treated with the same procedures but without occlusion. Subcutaneous microdialysis fluid samples were collected from "Neiguan" (PC 6), "Shenmen" (HT 7), "Xinshu" (BL 15), and "Taixi" (KI 3) regions on day 8(after recovery from operation), 14 (the first stage of MI), and 20 (the second stage of MI), as well as collected from PC 6 region during and post-acupuncture stimulation of PC 6, respectively. Manual acupuncture stimulation was applied to the right PC 6 for 30 min. Partial least squares -linear discriminant analysis (PLS-DA) was used to identify different metabolism patterns of the microdialysis fluid sample between groups and at different time-points in the same one group, and the distinct metabolites as the potential markers between groups were weighted via the values of variable importance in the projection (VIP) in combination with t-test analysis. An area under the curve (AUC) >1.0 indicated a test exhibiting good discrimination between groups. Six metabolites identified to be significantly different between the control and model groups were L-glutamic acid, phenylalanine and 3-hydroxyisobutyric acid (which were significantly increased relevant to the control group), and L-histidine, octadecanedioic acid and 9-keto palmitic acid (significantly decreased relevant to the control group) in the microdialysate of PC 6, HT 7 and BL 15 regions. In the microdialysate of PC 6, 4 metabolites including L-glutamic acid, octadecanedioic acid and 8-isohydroxy PGF 2 α (significantly increased), as well as L-histidine (markedly decreased) were identified to be considerably different between the model and control groups. After acupuncture for 30 min, the AUC level of L-glutamic acid was further significantly increased (P<0.05), that of L-histidine obviously decreased, and those of octadecanedioic acid and 8-isohydroxy PGF 2 α turned back nearly to the level of pre-MI. L-glutamic acid, phenylalanine, 3-hydroxyisobutyric acid, L-histidine, octadecanedioic acid and 9-keto palmitic acid from PC 6, HT 7 and BL 15 regions may be used as the material biomarker for MI-induced sensitization of these acupoints. Manual acupuncture intervention of PC 6 induces a significant change of L-histidine and L-glutamic acid in the local subcutaneous tissues.

Detection of wheat powdery mildew based on hyperspectral reflectance through SPA and PLS-LDA

Abstract: The accurate recognition and quantitative assessment at the early stage of wheat powdery mildew ( Blumeria graminis f. sp. tritici ) are vital for precision crop management for spraying the fungicides, reducing the cost, protecting the environment and enhancing the quality of crop. However, early disease detection remained highly difficult due to the subtle changes in the physiology and phenology of the plants at early infection stage. In this study, two wheat cultivars with different disease resistances were inoculated by the powdery mildew, hyperspectral reflectance and physiological parameters of leaves were obtained after inoculation at early stem elongation stage. The major contribution of this study is to extract sensitive wavebands and vegetation indices using sub-window permutation analysis (SPA) by fully exploiting the hyperspectral data for early disease identification. Extracted sensitive features by SPA were then used as input in partial least squares-linear discriminant analysis (PLS-LDA) recognition model to classify the healthy and diseased wheat leaves. Finally, validation was carried out with independent data to verify the accuracy of the recognition model. The results indicated that (1) the pigment contents and photosynthetic capacity were changed slightly at the early infection stage but decreased rapidly with the aggravation of the disease severity; (2) the visible and the near infrared bands were the most sensitive to the disease at early infection stage; (3) the overall accuracy of the PLS-LDA model constructed with the sensitive features extracted by SPA method performed better than features selected conventionally by correlation analysis. The calibration and validation accuracies at 5% disease severity were 85.12 and 84.43% for model based on wavelength features and were 82.14 and 85.63 for model constructed with spectral indices features extracted by SPA, respectively.  In conclusion, SPA is a new effective strategy for feature selection which has not been yet used in plant disease research, having the benefit of considering cooperative effect among different variables and demonstrated the potential of early disease detection. Such a technique can be an efficient and economical substitute to conventional methods, especially in case of high throughput hyperspectral crop sensing. Keywords: wheat powdery mildew, early detection, hyperspectral reflectance, fungicide spraying, sub-window permutation analysis (SPA), PLS-LDA DOI:  10.33440/j.ijpaa.20200301.67  Citation: Khan I H, Liu H Y, Cheng T, Tian Y C, Cao Q, Zhu Y, Cao W X, Yao X. Detection of wheat powdery mildew based on hyperspectral reflectance through SPA and PLS-LDA.  Int J Precis Agric Aviat, 2020; 3(1): 13–22.

Characterization of serum metabolite profile in patients with rheumatoid arthritis

Objective To investigate and characterize the serum metabolite profile of patients with rheumatoid arthritis (RA), a common autoimmune disease, which will be of help for early diagnosis in clinic. Methods Serum specimens from 26 patients with RA and 19 age-matched healthy volunteers were collected from the First Affiliated Hospital of Hunan University of Chinese Medicine from January through April 2015. Samples were detected on the gas chromatography-mass spectrometry (GC-MS), and serum metabolites were identified by the chemometric methods. Discriminative model of RA and healthy volunteers was established using partial least squares-linear discriminant analysis (PLS-LDA), and furthermore the established model was evaluated with double cross validation (DCV) for predicting ability. Finally, differential metabolites with clinical diagnostic potential were screened out by using subwindow permutation analysis (SPA). Results The established PLS-LDA discriminative model identified 48 metabolites. The total accuracy of the model approached 97.73%, in which the accuracy of the model for predicting RA and healthy volunteers was 96.15% and 100%, respectively. Our studies screened out 3-hydroxy butyric acid, phosphoric acid, isoleucine, mannose and hexadecanoic acid with clinical diagnostic potential based on the SPA. Conclusions Metabolomics with the application of GC-MS combined with the PLS-LDA method can distinguish patients with RA and healthy volunteers, and can aid in potential early diagnosis of RA. Key words: Arthritis, rheumatoid/ME; Gas chromatography-mass spectrometry

Raman and Infrared Spectroscopy Distinguishing Replicative Senescent from Proliferating Primary Human Fibroblast Cells by Detecting Spectral Differences Mainly Due to Biomolecular Alterations.

Cellular senescence is a terminal cell cycle arrested state, assumed to be involved in tumor suppression. We studied four human fibroblast cell strains (BJ, MRC-5, IMR-90, and WI-38) from proliferation into senescence. Cells were investigated by label-free vibrational Raman and infrared spectroscopy, following their transition into replicative senescence. During the transition into senescence, we observed rather similar biomolecular abundances in all four cell strains and between proliferating and senescent cells; however, in the four aging cell strains, we found common molecular differences dominated by protein and lipid modifications. Hence, aging induces a change in the appearance of biomolecules (including degradation and storage of waste) rather than in their amount present in the cells. For all fibroblast strains combined, the partial least squares-linear discriminant analysis (PLS-LDA) model resulted in 75% and 81% accuracy for the Raman and infrared (IR) data, respectively. Within this validation, senescent cells were recognized with 93% sensitivity and 90% specificity for the Raman and 84% sensitivity and 97% specificity for the IR data. Thus, Raman and infrared spectroscopy can identify replicative senescence on the single cell level, suggesting that vibrational spectroscopy may be suitable for identifying and distinguishing different cellular states in vivo, e.g., in skin.

Qualitative and quantitative detection of honey adulterated with high-fructose corn syrup and maltose syrup by using near-infrared spectroscopy

Near-infrared spectroscopy (NIR) was used for qualitative and quantitative detection of honey adulterated with high-fructose corn syrup (HFCS) or maltose syrup (MS). Competitive adaptive reweighted sampling (CARS) was employed to select key variables. Partial least squares linear discriminant analysis (PLS-LDA) was adopted to classify the adulterated honey samples. The CARS-PLS-LDA models showed an accuracy of 86.3% (honey vs. adulterated honey with HFCS) and 96.1% (honey vs. adulterated honey with MS), respectively. PLS regression (PLSR) was used to predict the extent of adulteration in the honeys. The results showed that NIR combined with PLSR could not be used to quantify adulteration with HFCS, but could be used to quantify adulteration with MS: coefficient (Rp2) and root mean square of prediction (RMSEP) were 0.901 and 4.041 for MS-adulterated samples from different floral origins, and 0.981 and 1.786 for MS-adulterated samples from the same floral origin (Brassica spp.), respectively.

Development and validation of an APCI-MS/GC–MS approach for the classification and prediction of Cheddar cheese maturity

Headspace techniques have been extensively employed in food analysis to measure volatile compounds, which play a central role in the perceived quality of food. In this study atmospheric pressure chemical ionisation-mass spectrometry (APCI-MS), coupled with gas chromatography–mass spectrometry (GC–MS), was used to investigate the complex mix of volatile compounds present in Cheddar cheeses of different maturity, processing and recipes to enable characterisation of the cheeses based on their ripening stages. Partial least squares-linear discriminant analysis (PLS-DA) provided a 70% success rate in correct prediction of the age of the cheeses based on their key headspace volatile profiles. In addition to predicting maturity, the analytical results coupled with chemometrics offered a rapid and detailed profiling of the volatile component of Cheddar cheeses, which could offer a new tool for quality assessment and accelerate product development.

Nondestructive Detection of Blackheart in Potato by Visible/Near Infrared Transmittance Spectroscopy

The possibility of using visible/near infrared (Vis/NIR) transmission spectroscopic technique in the 513–850 nm region coupled with partial least squares-linear discriminant analysis (PLS-LDA) and other chemometric methods to classify potatoes with blackheart was investigated. The discrimination performance of different morphological correction methods, including weight correction, height correction, and volume correction, was compared. The results showed that height corrected transmittance has the best performance, with both calibration and validation sets having a success rate of 97.11%. Out of 1800 wavelengths, only six wavelengths (711, 817, 741, 839, 678, and 698 nm) were selected as the optimum wavelengths for the discrimination of blackheart tubers based on principal component analysis (PCA). The data analysis showed that the overall classification rate by PLS-LDA method decreased from 97.11% to 96.82% in calibration set and from 97.11% to 96.53% in validation set, which was acceptable. The importance of these conclusions may be helpful to transfer Vis/NIR transmission technology from laboratory to industrial application in nondestructive, real-time, or portable measurement of potatoes quality.