ABSTRACT
Attenuated Total Reflectance-Fourier transform infrared (ATR-FTIR) spectroscopy is an emerging technology in the medical field. Blood D-dimer was initially studied as a marker of the activation of coagulation and fibrinolysis. It is mainly used as a potential diagnosis screening test for pulmonary embolism or deep vein thrombosis but was recently associated with COVID-19 severity. This study aimed to evaluate the use of ATR-FTIR spectroscopy with machine learning to classify plasma D-dimer concentrations. The plasma ATR-FTIR spectra from 100 patients were studied through principal component analysis (PCA) and two supervised approaches: genetic algorithm with linear discriminant analysis (GA-LDA) and partial least squares with linear discriminant (PLS-DA). The spectra were truncated to the fingerprint region (1800-1000 cm-1). The GA-LDA method effectively classified patients according to D-dimer cutoff (≤0.5 µg/mL and >0.5 µg/mL) with 87.5 % specificity and 100 % sensitivity on the training set, and 85.7 % specificity, and 95.6 % sensitivity on the test set. Thus, we demonstrate that ATR-FTIR spectroscopy might be an important additional tool for classifying patients according to D-dimer values. ATR-FTIR spectral analyses associated with clinical evidence can contribute to a faster and more accurate medical diagnosis, reduce patient morbidity, and save resources and demand for professionals.
Subject(s)
Spectroscopy, Fourier Transform Infrared , Humans , Spectroscopy, Fourier Transform Infrared/methods , Fourier Analysis , Discriminant Analysis , Principal Component Analysis , Ataxia Telangiectasia Mutated ProteinsABSTRACT
The Standard Practices for Infrared Multivariate Quantitative Analysis (ASTM E1655) provide a guide for determining physicochemical properties of materials using multivariate calibration techniques applied to chemical sources that have high multicollinearity and correlated information. Partial least squares (PLS) is the most widely used multivariate regression method due to its excellent prediction capabilities and easy optimization. Initially applied to chromatographic data, PLS has also shown great results in near-infrared (NIR) and mid-infrared (MIR) spectroscopies. However, complex chemical matrices with low correlation may not be efficiently modeled using PLS or other multivariate analyses limited by grouping similar information (such as latent variables or principal components). Therefore, this study aims to evaluate the multicollinearity of different analytical techniques, such as high-temperature gas chromatography (HTGC), NIR, MIR, hydrogen nuclear magnetic resonance (1H NMR), carbon-13 nuclear magnetic resonance (13C NMR), and Fourier transform ion cyclotron resonance mass spectrometry coupled to the electrospray source in positive and negative ionization modes (ESI(±)FT-ICR). Descriptive statistics (coefficient of determination, R2) and principal component analysis (PCA) were used to identify the distribution of correlated information. Results showed that NIR and MIR spectroscopies exhibited a higher percentage of correlated variables, while 13C NMR and ESI(±)FT-ICR MS had more discrete profiles. Therefore, PLS development may be more effectively applied to NIR, MIR, and 1H NMR data, while 13C NMR and mass spectra may require other algorithms or variable selection methods in combination with PLS.
ABSTRACT
The use of drugs of abuse has grown significantly in recent decades. In forensic chemistry, methods of identifying and characterizing illicit drugs contribute to the interests of researchers, experts, and public security authorities. Among existing methods, portable Raman spectroscopy is notable for performing rapid, non-destructive, and highly selective analysis in the laboratory or on-site. When the resulting spectral data are paired with chemometric tools, methods of exploratory analysis and multivariate calibration can be developed. Thus, this work describes the application of Raman spectroscopy associated with principal component analysis (PCA) and interval principal component analysis (iPCA) to assessing trends in samples of cocaine (n = 40), crack (n = 33), and their main adulterants (n = 5) and diluents (n = 5), tablets of ecstasy (n = 14), designer drugs papers (n = 27), and alcoholic solutions adulterated with benzodiazepines (alprazolam and diazepam). In addition, competitive adaptive reweighted sampling (CARS) combined with partial least squares (PLS) regression (CARSPLS) was used to quantify adulterants (benzocaine, lidocaine, and procaine) in binary mixtures with crack (n = 21) and solutions of cachaça adulterated with bromazepam (n = 11).
Subject(s)
Illicit Drugs , Spectrum Analysis, Raman , Illicit Drugs/analysis , Least-Squares Analysis , Principal Component Analysis , Spectrum Analysis, Raman/methods , TabletsABSTRACT
Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has caused the worst global health crisis in living memory. The reverse transcription polymerase chain reaction (RT-qPCR) is considered the gold standard diagnostic method, but it exhibits limitations in the face of enormous demands. We evaluated a mid-infrared (MIR) data set of 237 saliva samples obtained from symptomatic patients (138 COVID-19 infections diagnosed via RT-qPCR). MIR spectra were evaluated via unsupervised random forest (URF) and classification models. Linear discriminant analysis (LDA) was applied following the genetic algorithm (GA-LDA), successive projection algorithm (SPA-LDA), partial least squares (PLS-DA), and a combination of dimension reduction and variable selection methods by particle swarm optimization (PSO-PLS-DA). Additionally, a consensus class was used. URF models can identify structures even in highly complex data. Individual models performed well, but the consensus class improved the validation performance to 85% accuracy, 93% sensitivity, 83% specificity, and a Matthew's correlation coefficient value of 0.69, with information at different spectral regions. Therefore, through this unsupervised and supervised framework methodology, it is possible to better highlight the spectral regions associated with positive samples, including lipid (â¼1700 cm-1), protein (â¼1400 cm-1), and nucleic acid (â¼1200-950 cm-1) regions. This methodology presents an important tool for a fast, noninvasive diagnostic technique, reducing costs and allowing for risk reduction strategies.