Rapid Quality Discrimination and Amino Nitrogen Quantitative Evaluation of Soy Sauces by Tri-Step IR and E-nose

Effects of Novel Preparation Technology on Flavor of Vegetable-Soy Sauce Compound Condiment

Vegetables contain important bioactive substances which have unique tastes and aromas and provide beneficial effects to human health. In this study, multiflavor blended soy sauce was prepared with the juice of eight kinds of vegetables, dried shrimp boiled stock, and six kinds of commercial soy sauce as raw materials, and thermal ultrasound was used as the sterilization method. The effects of adding different formulas of vegetable and seafood stock on the basic physical and chemical parameters, nutrition, antioxidant activity, flavor, and taste of soy sauce were investigated. The results showed that the basic physicochemical indices such as pH, total acid, color, soluble solids, and amino acid nitrogen of the product with a ratio of soy sauce to vegetable-seafood stock of 1:0.5 (v/v) could meet the production standards of soy sauce, and its flavor, taste, and sensory scores were relatively good, with the highest likeability (overall acceptability). The mixed soy sauce with a ratio of 1:2 (v/v) had higher vegetable and seafood flavors, and different vegetable flavors (celery, carrot, and onion) were more obvious, but its nutritional index was relatively low. Multiflavor vegetable-soy sauce can be used for quick cooking by chefs of catering enterprises, and may be used as a seasoning bag for prefabricated dishes and convenient foods, attracting increasing attention from manufacturers and consumers.

Recent Progress in Smart Electronic Nose Technologies Enabled with Machine Learning Methods

Machine learning methods enable the electronic nose (E-Nose) for precise odor identification with both qualitative and quantitative analysis. Advanced machine learning methods are crucial for the E-Nose to gain high performance and strengthen its capability in many applications, including robotics, food engineering, environment monitoring, and medical diagnosis. Recently, many machine learning techniques have been studied, developed, and integrated into feature extraction, modeling, and gas sensor drift compensation. The purpose of feature extraction is to keep robust pattern information in raw signals while removing redundancy and noise. With the extracted feature, a proper modeling method can effectively use the information for prediction. In addition, drift compensation is adopted to relieve the model accuracy degradation due to the gas sensor drifting. These recent advances have significantly promoted the prediction accuracy and stability of the E-Nose. This review is engaged to provide a summary of recent progress in advanced machine learning methods in E-Nose technologies and give an insight into new research directions in feature extraction, modeling, and sensor drift compensation.

A novel method based on infrared spectroscopic inception-resnet networks for the detection of the major fish allergen parvalbumin

We have developed a novel approach that involves inception-resnet network (IRN) modeling based on infrared spectroscopy (IR) for rapid and specific detection of the fish allergen parvalbumin. SDS-PAGE and ELISA were used to validate the new method. Through training and learning with parvalbumin IR spectra from 16 fish species, IRN, support vector machine (SVM), and random forest (RF) models were successfully established and compared. The IRN model extracted highly representative features from the IR spectra, leading to high accuracy in recognizing parvalbumin (up to 97.3%) in a variety of seafood matrices. The proposed infrared spectroscopic IRN (IR-IRN) method was rapid (~20 min, cf. ELISA ~4 h) and required minimal expert knowledge for application. Thus, it could be extended for large-scale field screening and identification of parvalbumin or other potential allergens in complex food matrices.

Brackish water improves the taste quality in meat of adult male Eriocheir sinensis during the postharvest temporary rearing

We investigated the effect of temporary rearing in brackish water on the taste quality in meat of crab cooked. The main salinity-responsive factors included 5'-nucleotides and free amino acids (FAAs) in crab meat that were identified using tri-step infrared spectroscopy. Compared to the fresh water group, the contents of 5'-adenosine monophosphate and 5'-inosine monophosphate in the brackish water group significantly increased in the 2nd week and decreased in the 6th week, respectively. The contribution ratio of umami FAAs increased from 8.1 to 13.5% in the 4th week in the brackish water group, showing maximum value of equivalent umami concentration. Moreover, Ca²⁺ and Cl^- contents significantly increased in the 4th and 6th weeks, respectively (P < 0.05). Infrared spectroscopy was an effective method to identify the taste components. With respect to the taste quality, four weeks were determined as the best period for temporary rearing of the crab in brackish water.

Geographical origin traceability of Cabernet Sauvignon wines based on Infrared fingerprint technology combined with chemometrics

Mid-infrared (MIR) and near-infrared (NIR) spectroscopy combined with chemometrics were explored to classify Cabernet Sauvignon wines from different countries (Australia, Chile and China). Commercial wines (n = 540) were scanned in transmission mode using MIR and NIR, and their characteristic fingerprint bands were extracted at 1750-1000 cm⁻¹ and 4555-4353 cm⁻¹. Through the identification system of Tri-step infrared spectroscopy, the correlation between macroscopic chemical fingerprints and geographical regions was explored more deeply. Furthermore, Principal component analysis (PCA), soft independent modelling of class analogy (SIMCA) and discriminant analysis (DA) based on MIR and NIR spectra were used to visualize or discriminate differences between samples and to realize geographical origin traceability of Cabernet Sauvignon wines. Through “external test set (n = 157)” validation, SIMCA models correctly classified 97%, 97% and 92% of Australian, Chilean and Chinese Cabernet Sauvignon wines, while the DA models correctly classified 86%, 85% and 77%, respectively. Based on unique digital fingerprints of spectroscopy (FT-MIR and FT-NIR) associated with chemometrics, geographical origin traceability was achieved in a more comprehensive, effective and rapid manner. The developed database models based on IR fingerprint spectroscopy with chemometrics could provide scientific basis and reference for geographical origin traceability of Cabernet Sauvignon wines (Australia, Chile and China).

Integrated recognition and quantitative detection of starch in surimi by infrared spectroscopy and spectroscopic imaging

Surimi products have become increasingly-consumed food with prominent characteristics of high nutrition and convenience and its supply falls short of demand. However, due to exhausted fishery resource in recent years, surimi adulteration, such as addition of plant proteins, starch and other animal origin meat, is becoming serious, so recognition of these exogenous substances has become an urgent issue. In this study, Fourier transform infrared spectroscopy (FT-IR) combined with infrared spectroscopic imaging could distinguish heterogeneity in surimi qualitatively and quantitatively and obtain integral chemical images so that spatial distribution of each component in surimi could be visually displayed, thus a rapid recognition method and a prediction model were developed. The different starch contents in surimi had been primarily identified through intensity change of infrared absorption peaks at 1045cm^-1 and 988cm^-1, specifically with peak shifts to 1041cm^-1 and to 992cm^-1, respectively. In infrared imaging analysis, principal components (PCs) were separated and one key PC was confirmed as starch by characteristic peaks comparison at 1147cm^-1, 1075cm^-1, 997cm^-1 and 930cm^-1. Meanwhile, an established statistic model could predict starch content in surimi correctly with a reliable correlation coefficient (R=0.9856) and root mean square error of prediction (RMSEP=5.64). Therefore, FT-IR combined with infrared spectroscopic imaging could be applicable to integrally recognize and quantitatively detect starch in surimi.