Calibration Transfer of Partial Least Squares Regression Models between Desktop Nuclear Magnetic Resonance Spectrometers

Strategies for the quality control of Chrysanthemi Flos: Rapid quantification and end-to-end fingerprint conversion based on FT-NIR spectroscopy

INTRODUCTION

Chrysanthemi Flos (CF) is widely used as a natural medicine or tea. Due to its diverse cultivation regions, CF exhibits varying quality. Therefore, the quality and swiftness in evaluation holds paramount significance for CF.

OBJECTIVE

The aim of the study was to construct a comprehensive evaluation strategy for assessing CF quality using HPLC, near-infrared (NIR) spectroscopy, and chemometrics, which included the rapid quantification analyses of chemical components and the Fourier transform (FT)-NIR to HPLC conversion of fingerprints.

MATERIALS AND METHODS

A total of 145 CF samples were utilised for data collection via NIR spectroscopy and HPLC. The partial least squares regression (PLSR) models were optimised using various spectral preprocessing and variable selection methods to predict the chemical composition content in CF. Both direct standardisation (DS) and PLSR algorithms were employed to establish the fingerprint conversion model from the FT-NIR spectrum to HPLC, and the model's performance was assessed through similarity and cluster analysis.

RESULTS

The optimised PLSR quantitative models can effectively predict the content of eight chemical components in CF. Both DS and PLSR algorithms achieve the calibration conversion of CF fingerprints from FT-NIR to HPLC, and the predicted and measured HPLC fingerprints are highly similar. Notably, the best model relies on CF powder FT-NIR spectra and DS algorithm [root mean square error of prediction (RMSEP) = 2.7590, R2 = 0.8558]. A high average similarity (0.9184) prevails between predicted and measured fingerprints of test set samples, and the results of the clustering analysis exhibit a high level of consistency.

CONCLUSION

This comprehensive strategy provides a novel and dependable approach for the rapid quality evaluation of CF.

Efficient Near-Infrared Spectrum Detection in Nondestructive Wood Testing via Transfer Network Redesign

This study systematically developed a deep transfer network for near-infrared spectrum detection using convolutional neural network modules as key components. Through meticulous evaluation, specific modules and structures suitable for constructing the near-infrared spectrum detection model were identified, ensuring its effectiveness. This study extensively analyzed the basic network components and explored three unsupervised domain adaptation structures, highlighting their applications in the nondestructive testing of wood. Additionally, five transfer networks were strategically redesigned to substantially enhance their performance. The experimental results showed that the Conditional Domain Adversarial Network and Globalized Loss Optimization Transfer network outperformed the Direct Standardization, Piecewise Direct Standardization, and Spectral Space Transformation models. The coefficients of determination for the Conditional Domain Adversarial Network and Globalized Loss Optimization Transfer network are 82.11% and 83.59%, respectively, with root mean square error prediction values of 12.237 and 11.582, respectively. These achievements represent considerable advancements toward the practical implementation of an efficient and reliable near-infrared spectrum detection system using a deep transfer network.

Successful combination of benchtop nuclear magnetic resonance spectroscopy and chemometric tools: A review

Low-field nuclear magnetic resonance (NMR) has three general modalities: spectroscopy, imaging, and relaxometry. In the last twelve years, the modality of spectroscopy, also known as benchtop NMR, compact NMR, or just low-field NMR, has undergone instrumental development due to new permanent magnetic materials and design. As a result, benchtop NMR has emerged as a powerful analytical tool for use in process analytical control (PAC). Nevertheless, the successful application of NMR devices as an analytical tool in several areas is intrinsically linked to its coupling with different chemometric methods. This review focuses on the evolution of benchtop NMR and chemometrics in chemical analysis, including applications in fuels, foods, pharmaceuticals, biochemicals, drugs, metabolomics, and polymers. The review also presents different low-resolution NMR methods for spectrum acquisition and chemometric techniques for calibration, classification, discrimination, data fusion, calibration transfer, multi-block and multi-way.

Near infrared spectroscopy quantification based on Bi-LSTM and transfer learning for new scenarios

This study proposed a deep transfer learning methodology based on an improved Bi-directional Long Short-Term Memory (Bi-LSTM) network for the first time to address the near infrared spectroscopy (NIR) model transfer issue between samples. We tested its effectiveness on two datasets of manure and polyglutamic acid (γ-PGA) solution, respectively. First, the optimal primary Bi-LSTM networks for cattle manure and the first batch of γ-PGA were developed by ablation experiments and both proved to outperform one-dimensional convolutional neural network (1D-CNN), Partial Least Square (PLS) and Extreme Learning Machine (ELM) models. Then, two types of transfer learning approaches were carried out to determine model transferability to non-homologous samples. For poultry manure and the second batch of γ-PGA, the obtained predicting results verified that the second approach of fine-tuning Bi-LSTM layers and re-training FC layers transcended the first approach of fixing Bi-LSTM layers and only re-training FC layers by reducing the RMSEP_target of 23.4275% and 50.7343%, respectively. Finally, comparisons with fine-tuning 1D-CNN and other traditional model transfer methods further identified the superiority of the proposed methodology with exceeding accuracy and smaller variation, which decreased RMSEP_target of poultry manure and the second batch of γ-PGA of 7.2832% and 48.1256%, 67.1117% and 80.6924% when compared to that acquired by fine-tuning 1D-CNN, Tradaboost-ELM and CCA-PLS which were the best of five traditional methods, respectively. The study demonstrates the potential of the Fine-tuning-Bi-LSTM enabled NIR technology to be used as a simple, cost effective and reliable detection tool for a wide range of applications under various new scenarios.

Parameter Visualization of Benchtop Nuclear Magnetic Resonance Spectra toward Food Process Monitoring

Low-cost and user-friendly benchtop low-field nuclear magnetic resonance (NMR) spectrometers are typically used to monitor food processes in the food industry. Because of excessive spectral overlap, it is difficult to characterize food mixtures using low-field NMR spectroscopy. In addition, for standard compounds, low-field benchtop NMR data are typically unavailable compared to high-field NMR data, which have been accumulated and are reusable in public databases. This work focused on NMR parameter visualization of the chemical structure and mobility of mixtures and the use of high-field NMR data to analyze benchtop NMR data to characterize food process samples. We developed a tool to easily process benchtop NMR data and obtain chemical shifts and T2 relaxation times of peaks, as well as transform high-field NMR data into low-field NMR data. Line broadening and time–frequency analysis methods were adopted for data processing. This tool can visualize NMR parameters to characterize changes in the components and mobilities of food process samples using benchtop NMR data. In addition, assignment errors were smaller when the spectra of standard compounds were identified by transferring the high-field NMR data to low-field NMR data rather than directly using experimentally obtained low-field NMR spectra.

Is the Calibration Transfer of Multivariate Calibration Models between High- and Low-Field NMR Instruments Possible? A Case Study of Lignin Molecular Weight

Although several successful applications of benchtop nuclear magnetic resonance (NMR) spectroscopy in quantitative mixture analysis exist, the possibility of calibration transfer remains mostly unexplored, especially between high- and low-field NMR. This study investigates for the first time the calibration transfer of partial least squares regressions [weight average molecular weight (M_w) of lignin] between high-field (600 MHz) NMR and benchtop NMR devices (43 and 60 MHz). For the transfer, piecewise direct standardization, calibration transfer based on canonical correlation analysis, and transfer via the extreme learning machine auto-encoder method are employed. Despite the immense resolution difference between high-field and low-field NMR instruments, the results demonstrate that the calibration transfer from high- to low-field is feasible in the case of a physical property, namely, the molecular weight, achieving validation errors close to the original calibration (down to only 1.2 times higher root mean square errors). These results introduce new perspectives for applications of benchtop NMR, in which existing calibrations from expensive high-field instruments can be transferred to cheaper benchtop instruments to economize.

Mitigating instrument effects in 60 MHz 1H NMR spectroscopy for authenticity screening of edible oils

Low field (60 MHz) ¹H NMR spectroscopy was used to analyse a large (n = 410) collection of edible oils, including olive and argan, in an authenticity screening scenario. Experimental work was carried out on multiple spectrometers at two different laboratories, aiming to explore multivariate model stability and transfer between instruments. Three modelling methods were employed: Partial Least Squares Discriminant Analysis, Random Forests, and a One Class Classification approach. Clear inter-instrument differences were observed between replicated data collections, sufficient to compromise effective transfer of models based on raw data between instruments. As mitigations to this issue, various data pre-treatments were investigated: Piecewise Direct Standardisation, Standard Normal Variates, and Rank Transformation. Datasets comprised both phase corrected and magnitude spectra, and it was found that that the latter spectral form may offer some advantages in the context of pattern recognition and classification modelling, particularly when used in combination with the Rank Transformation pre-treatment.

A One-Dimensional U-Net-Based Calibration-Transfer Method for Low-Field Nuclear Magnetic Resonance Signals

The reconstruction of the statistical analysis model of an instrument is a time-consuming and expensive process. Herein, the feasibility of spectral model calibration-transfer application to the same type of low-field nuclear magnetic resonance (LF-NMR) instrument was investigated using a one-dimensional U-net (1D U-net). Unlike conventional calibration-transfer algorithms such as direct standardization (DS), the 1D U-net network can reduce the error between the master and slave instruments through iterative cycles. The calibration-transfer ability was verified; three experiments that entailed the use of edible oil and copper sulfate (CuSO₄) samples were implemented. The analysis of the spectral responses and feature analysis of the edible oil samples revealed that the signal of the slave instrument calibrated using the 1D U-net most resembled the signal of the master instrument, and its relative residual value was reduced to 0.0045. Further analysis of the CuSO₄ concentration prediction showed that on the support vector regression (SVR) model constructed using the master instrument, the signal of the slave instrument calibrated by the 1D U-net was more similar to the response of the master instrument, and its root mean square error (RMSE) was only 0.01606 mmol/L. Thus, 1D U-net is a viable candidate for calibration-transfer applications to LF-NMR instruments.

Compact low-field NMR spectroscopy and chemometrics applied to the analysis of edible oils

Compact nuclear magnetic resonance (NMR) spectroscopy combined with chemometric tools opens new perspectives for NMR use. This work compares the potential of 43, 60 and 400 MHz NMR spectroscopy for quality control of edible oils. Partial least squares regression (PLSR) and support vector regression (SVR) models built on the three NMR devices had equivalent performances for fatty acids and iodine value, and the models built with the low field spectra were equivalent to the high field. Moreover, performances for calibration indicated that most of the models built with medium/or high-resolution fields presented reproducibility values lower than the minimum accepted by the American Oil Chemists' Society (AOCS). Compared to classical methods, this new approach allows the application of medium resolution devices as a sample screening tool in analytical laboratories since it allows the spectrum obtention in a few seconds, without the need for sample preparation or the use of deuterated solvents.

E-sensing and nanoscale-sensing devices associated with data processing algorithms applied to food quality control: a systematic review

Devices of human-based senses such as e-noses, e-tongues and e-eyes can be used to analyze different compounds in several food matrices. These sensors allow the detection of one or more compounds present in complex food samples, and the responses obtained can be used for several goals when different chemometric tools are applied. In this systematic review, we used Preferred Reporting Items for Systematic Reviews and Meta-Analysis guidelines, to address issues such as e-sensing with chemometric methods for food quality control (FQC). A total of 109 eligible articles were selected from PubMed, Scopus and Web of Science. Thus, we predicted that the association between e-sensing and chemometric tools is essential for FQC. Most studies have applied preliminary approaches like exploratory analysis, while the classification/regression methods have been less investigated. It is worth mentioning that non-linear methods based on artificial intelligence/machine learning, in most cases, had classification/regression performances superior to non-liner, although their applications were seen less often. Another approach that has generated promising results is the data fusion between e-sensing devices or in conjunction with other analytical techniques. Furthermore, some future trends in the application of miniaturized devices and nanoscale sensors are also discussed.