Detecting cooking state of grilled chicken by electronic nose and computer vision techniques

A Paradigm of Computer Vision and Deep Learning Empowers the Strain Screening and Bioprocess Detection

High-performance strain and corresponding fermentation process are essential for achieving efficient biomanufacturing. However, conventional offline detection methods for products are cumbersome and less stable, hindering the "Test" module in the operation of "Design-Build-Test-Learn" cycle for strain screening and fermentation process optimization. This study proposed and validated an innovative research paradigm combining computer vision with deep learning to facilitate efficient strain selection and effective fermentation process optimization. A practical framework was developed for gentamicin C1a titer as a proof-of-concept, using computer vision to extract different color space components across various cultivation systems. Subsequently, by integrating data preprocessing with algorithm design, a prediction model was developed using 1D-CNN model with Z-score preprocessing, achieving a correlation coefficient (R2) of 0.9862 for gentamicin C1a. Furthermore, this model was successfully applied for high-yield strain screening and real-time monitoring of the fermentation process and extended to rapid detection of fluorescent protein expression in promoter library construction. The visual sensing research paradigm proposed in this study provides a theoretical framework and data support for the standardization and digital monitoring of color-changing bioprocesses.

Textured vegetable protein as a partial replacement for lean meat in salami analogues: Perspectives on physicochemical properties, flavour and proteome changes

Integrating plant proteins into meat products offers a sustainable way to reduce the environmental impact of meat consumption while satisfying the growing flexitarian population. This study explored the effects of textured vegetable proteins (TVPs) on the physico-chemical attributes and flavour profile of hybrid salamis using 4D label-free proteomics. Results showed that hybrid salamis had lower pH, reduced water activity and increased weight loss compared with traditional salamis, along with greater hardness and a slightly rough, porous texture with a filamentous structure. TVPs substantially modified crucial meaty flavour compounds (nitrogen oxides, sulfides and pyrazine), increasing heightening sourness and bitterness while diminishing umami. Proteomic analysis revealed significant upregulation of myosin and actin in hybrid salamis; notably, these proteins were involved in glycerol-3-phosphate dehydrogenase activity and calcineurin-mediated signalling, underscoring their role in flavour enhancement. Therefore, hybrid salamis offer an attractive alternative to traditional salamis by merging meat-like taste and texture with plant protein.

A methodological review in sensory analyses of chicken meat

The sensory characteristics of poultry products are crucial in defining their quality and widely influence consumer choices. Even though the scientific literature clearly indicates that for muscle foods the sensory profile is relevant in purchase decisions and overall acceptability, sensory evaluation has often been underestimated and considered complementary to instrumental and/or chemical assessments. Sensory analysis includes different types of validated tests (discriminative, descriptive, and affective), applied depending on the purpose of the research study, requiring special attention in the sample preparation phase, in particular for nonhomogeneous products such as poultry meat, requiring reproducible cutting, cooking and presentation to the tasters. The aim of this paper is to review, critically assess and discuss sensory methods, standardized procedures and sample preparation tailored for chicken meat, through the literature from 2000 to 2023, with a section dedicated to ethical aspects that must be carefully considered when designing a sensory protocol. The target readers are both the research and the business communities, as the information can be widely applied for quality control, to develop new food products, to understand or drive preferences or, for example, to assess potential sensory differences among chickens fed with different diets. To the best of the authors' knowledge, this review represents a useful first guide for those approaching the sensory analysis of chicken meat.

Does the Formation of a Taylor Cone in a Pulsating Electrospray Directly Impact Mass Spectrometry Signals?

Electrospray ionization (ESI) remains the dominant technique in mass spectrometry (MS)-based analyses. Here, we investigated the relationship between a crucial aspect of ESI, the formation of the Taylor cone, and the MS ion current by utilizing a triple quadrupole (QqQ) mass spectrometer coupled with a streaming high-speed camera and a 3-ring electrode system. In one test, ion current over a 30-s plume gate (a ring electrode) opening was compared with the Taylor cone occurrence analyzed offline, with Spearman's correlation coefficients consistently near 0 despite parameter variations. In another test, real-time detection of Taylor cones was synchronized with QqQ-MS, selectively opening (de-energizing) the plume gate based on the Taylor cone status. This approach enabled matching the ion current with the Taylor cone occurrence. There was no apparent difference between the MS signals recorded in the presence and absence of a Taylor cone. Additionally, a Faraday plate was employed as a detector in offline experiments, revealing agreement between the frequency of liquid meniscus (Taylor cone) oscillation (∼1.92 kHz)-measured by high-speed imaging-and the frequency of spray current (∼1.93 kHz). We suggest that the lack of positive correlations in the MS experiments is due to intrinsic ion carryover during transit from the ion source to the detector and due to the insufficient data acquisition rate of the mass spectrometer, which erases short-term fluctuations of ion current.

Effects of Variable-Temperature Roasting on the Flavor Compounds of Xinjiang Tannur-Roasted Mutton

This study investigates the effect of variable-temperature roasting on the flavor compounds of Xinjiang tannur-roasted mutton. Gas chromatography coupled with ion mobility spectroscopy (GC-IMS) was used to compare and analyze the volatile components and flavor fingerprints of Xinjiang tannur-roasted mutton using variable-temperature electrically heated air roasting (VTR), constant-temperature electrically heated air roasting (EHAR), and constant-burning charcoal roasting (BCR) techniques. The changes in fatty acids and free amino acids in Xinjiang tannur-roasted mutton under different roasting conditions were compared. By using GC-IMS analysis, 11 flavor compounds, including 4-methyl-3-penten-2-one, isoamyl propionate, trans-2-heptenal, trans-2-heptenal, 2-hexanone, n-hexanol, 2-hexenal, 2-ethylfuran, and ethyl 2-methylbutanoate, were identified as characteristic volatile compounds in the temperature-controlled electrothermal roasting of Xinjiang tannur-roasted mutton using the following conditions: 0-4 min, 300 °C; 5-10 min, 220 °C; and 11-17 min, 130 °C (VTR3). Through principal component analysis, it was found that the substances with the highest positive correlation with PC1 and PC2 were n-hexanol and 3-methylbutanol. The sensory evaluation showed that VTR3 had high acceptability (p < 0.05) and a fat flavor (p < 0.05). There was no significant difference in the total fatty acid (TFA) content between the VTR3 and burning charcoal roast for 1-17 min at 300 °C (BCR3) (p > 0.05), but it was lower than that in the other experimental groups (p < 0.05). The lowest proportion of glutamic acid content in VTR3 was 22.44%, and the total free amino acid content in the electric thermostatic roasting for the 1-17 min, 300 °C (EHAR3) group (347.05 mg/100 g) was significantly higher than that in the other experimental groups (p < 0.05). By using Spearman correlation analysis, the roasting loss rate showed a highly significant negative correlation with essential amino acids (EAAs), non-essential amino acids (NEAAs), and total free amino acids (TAAs) (the correlation coefficients (r) were 0.82, 0.87, and 0.87, respectively) with p < 0.01. There was no correlation between changes in the free amino acid content and fatty acid content (p > 0.05). By using Differential scanning calorimetry (DSC) analysis, we also found that there was no significant difference in peak temperature (Tp) between the VTR3 and EHAR experimental groups (p > 0.05). Variable temperature electric heating can affect the flavor of lamb, and there are significant differences in the content of flavor precursors such as fatty acids and amino acids in Xinjiang tannur-roasted mutton.

Sensing signal augmentation by flow rate modulation of carrier gas for accurate differentiation of complex odours

For olfactory sensors, clear differentiation of complex odour samples requires diverse information. To obtain such information, hardware modifications, such as introducing additional channels with different physical/chemical properties, are usually needed. In this study, we present a new approach to augmenting the sensing signals of an olfactory sensor by modulating the flow rate of the carrier gas. The headspace vapour of complex odours is measured using a sensing system of nanomechanical sensor (Membrane-type Surface stress Sensor, MSS). The resulting data set is quantitatively evaluated using the Davies-Bouldin index (DBI) of principal component analysis (PCA). The increasing number of sensing signals obtained at different gas flow rates leads to a decrease in the DBI, achieving better cluster separation between different odours. Such gas flow effects can be attributed to several factors, including the sample evaporation and the equilibrium of the gas-liquid and gas-solid interfaces. Proton-transfer-reaction time-of-flight mass spectrometry (PTR-TOF-MS) experiments reveal that the compositions of odour samples vary with the different gas flow rates. It is demonstrated that a simple technique for modulating gas flow rates can significantly improve the differentiation performance of complex odours, providing an additional degree of freedom in olfactory sensing.

Monitoring of meat quality and change-point detection by a sensor array and profiling of bacterial communities

BACKGROUND

Real-time monitoring of food consumer quality remains challenging due to diverse bio-chemical processes taking place in the food matrices, and hence it requires accurate analytical methods. Thresholds to determine spoiled food are often difficult to set. The existing analytical methods are too complicated for rapid in situ screening of foodstuff.

RESULTS

We have studied the dynamics of meat spoilage by electronic nose (e-nose) for digitizing the smell associated with volatile spoilage markers of meat, comparing the results with changes in the microbiome composition of the spoiling meat samples. We apply the time series analysis to follow dynamic changes in the gas profile extracted from the e-nose responses and to identify the change-point window of the meat state. The obtained e-nose features correlate with changes in the microbiome composition such as increase in the proportion of Brochothrix and Pseudomonas spp. and disappearance of Mycoplasma spp., and with representative gas sensors towards hydrogen, ammonia, and alcohol vapors with R2 values of 0.98, 0.93, and 0.91, respectively. Integration of e-nose and computer vision into a single analytical panel improved the meat state identification accuracy up to 0.85, allowing for more reliable meat state assessment.

SIGNIFICANCE

Accurate identification of the change-point in the meat state achieved by digitalizing volatile spoilage markers from the e-nose unit holds promises for application of smart miniaturized devices in food industry.

Hydrogen Gas-Grilling in Meat: Impact on Odor Profile and Contents of Polycyclic Aromatic Hydrocarbons and Volatile Organic Compounds

The effect of fuel (hydrogen vs. butane) on the formation of volatile organic compounds (VOCs) and polycyclic aromatic hydrocarbons (PAHs) was evaluated for grilled horse meat (very low-fat and low-fat) cooking vertically. Gas chromatography-mass spectrometry was used to analyze PAHs and VOCs. An electronic nose was used to evaluate the odor profile. Total high-molecular-weight PAHs ranged from 19.59 to 28.65 µg/kg with butane and from 1.83 to 1.61 µg/kg with hydrogen. Conversely, total low-molecular-weight PAHs went from 184.41 to 286.03 µg/kg with butane and from 36.88 to 41.63 µg/kg with hydrogen. Aldehydes and alkanes were the predominant family in a total of 59 VOCs. Hydrogen gas-grilling reduced significantly (p < 0.05) the generation of VOCs related to lipid oxidation. The odor profile was not modified significantly despite the change of PAHs and VOCs. The findings indicate that hydrogen is a viable alternative to butane for grilling horse meat. Hydrogen gas-grilling may be regarded as a safe cooking procedure of meat from a PAH contamination point and perhaps sustainable environmentally compared to a conventional technique. The present study provides the basis for the use of hydrogen gas in grilled meat.

Unraveling the flavor profiles of chicken meat: Classes, biosynthesis, influencing factors in flavor development, and sensory evaluation

Chicken is renowned as the most affordable meat option, prized by consumers worldwide for its unique flavor, and universally recognized for its essential savory flavor. Current research endeavors are increasingly dedicated to exploring the flavor profile of chicken meat. However, there is a noticeable gap in comprehensive reviews dedicated specifically to the flavor quality of chicken meat, although existing reviews cover meat flavor profiles of various animal species. This review aims to fill this gap by synthesizing knowledge from published literature to describe the compounds, chemistry reaction, influencing factors, and sensory evaluation associated with chicken meat flavor. The flavor compounds in chicken meat mainly included water-soluble low-molecular-weight substances and lipids, as well as volatile compounds such as aldehydes, ketones, alcohols, acids, esters, hydrocarbons, furans, nitrogen, and sulfur-containing compounds. The significant synthesis pathways of flavor components were Maillard reaction, Strecker degradation, lipid oxidation, lipid-Maillard interaction, and thiamine degradation. Preslaughter factors, including age, breed/strain, rearing management, muscle type, and sex of chicken, as well as postmortem conditions such as aging, cooking conditions, and low-temperature storage, were closely linked to flavor development and accounted for the significant differences observed in flavor components. Moreover, the sensory methods used to evaluate the chicken meat flavor were elaborated. This review contributes to a more comprehensive understanding of the flavor profile of chicken meat. It can serve as a guide for enhancing chicken meat flavor quality and provide a foundation for developing customized chicken products.

Prediction and visualization of moisture content in Tencha drying processes by computer vision and deep learning

BACKGROUND

It is important to monitor and control the moisture content throughout the Tencha drying processing procedure so that its quality is ensured. Workers often rely on their senses to perceive the moisture content, leading to relative subjectivity and low reproducibility. Traditional drying methods, which are used for measuring moisture content, are destructive to samples. This research was conducted using computer vision combined with deep learning to detect moisture content during the Tencha drying process. Different color space components of Tencha drying sample images were first extracted by computer vision. The color components were preprocessed using MinMax and Z score. Subsequently, one-dimensional convolutional neural networks (1D-CNN), partial least squares, and backpropagation artificial neural networks models were built and compared.

RESULTS

The 1D-CNN model and Z score preprocessing achieved superior predictive accuracy, with correlation coefficient of prediction (Rp) = 0.9548 for moisture content. The migration of moisture content during the Tencha drying process was eventually visualized by mapping its spatial and temporal distributions.

CONCLUSION

The results indicated that computer vision combined with 1D-CNN was feasible for moisture prediction during the Tencha drying process. This study provides technical support for the industrial and intelligent production of Tencha. © 2024 Society of Chemical Industry.

Geographical traceability of soybean: An electronic nose coupled with an effective deep learning method

The quality of soybeans is correlated with their geographical origin. It is a common phenomenon to replace low-quality soybeans from substandard origins with superior ones. This paper proposes the adaptive convolutional kernel channel attention network (AKCA-Net) combined with an electronic nose (e-nose) to achieve soybean quality traceability. First, the e-nose system is used to collect soybean gas information from different origins. Second, depending on the characteristics of the gas information, we propose the adaptive convolutional kernel channel attention (AKCA) module, which focuses on key gas channel features adaptively. Finally, the AKCA-Net is proposed, which is capable of modeling deep gas channel interdependency efficiently, realizing high-precision recognition of soybean quality. In comparative experiments with other attention mechanisms, AKCA-Net demonstrated superior performance, achieving an accuracy of 98.21%, precision of 98.57%, and recall of 98.60%. In conclusion, the combination of the AKCA-Net and e-nose provides an effective strategy for soybean quality traceability.

Mechanism underlying the tenderness evolution of stir-fried pork slices with heating rate revealed by infrared thermal imaging assistance

This study aimed to explore the mechanism of cooking intensity on the tenderness of stir-fried pork slices from the perspective of the changes in temperature distribution. Infrared thermal imaging was used to monitor the distribution of temperature. Results showed that the high-level heat (HH) treatment could improve tenderness. When the center temperature increased to 100 °C, the shear force of samples from the low-level heat (LH) group increased by around 3-fold, and HH reduced this upward trend. This result was mainly attributed to the shorter heating time undergone by the HH-treated samples compared to the LH treatment, which resulted in less structural shrinkage and faster passing through the protein denaturation interval of the samples. These changes alleviated temperature fluctuations caused by water loss. This explanation could be confirmed by the results of T2 relaxation time and Fourier transform-infrared spectroscopy (FT-IR). However, the LH treatment caused a slower rise in oil temperature due to more moisture migration, which required the samples to undergo longer thermal denaturation, leading to a deterioration in tenderness. Moreover, histological analysis revealed that the greater integrity of endomysium in the HH group inhibited water loss and oil absorption, which contributed to obtain low-fat meat products with higher tenderness. This study provides support for the industrialization of traditional pork cuisines using oil as the heating medium.

Feasibility of Tenebrio molitor larvae protein to partially replace lean meat in the processing of hybrid frankfurters: Perspectives on quality profiles and in vitro digestibility

The aim of this study was to investigate the effect of replacing different amounts (5 %, 10 %, 15 %, 20 % and 25 %) of lean meat with Tenebrio molitor larvae protein (TMLP) on the quality profiles of hybrid frankfurters. The results showed that there were no obvious differences in moisture, protein or fat content of all the hybrid frankfurters (P > 0.05), only a higher substitution rate (from 10 % to 25 %) resulted in a higher ash content than the control group (P < 0.05). With the increasing replacement rate (5 %, 10 % and 15 %), the cooking loss of the hybrid frankfurters showed the similar effects as the control group (P > 0.05), whereas the higher replacement rates of 20 % and 25 % obviously decreased the emulsion stability of the hybrid frankfurters. Moreover, with lower substitution rate (5 %, 10 % and 15 %) there were no significant differences in cooking loss between the hybrid frankfurters and the control group (P > 0.05), whereas the higher substitution rates (20 % and 25 %) obviously increased the cooking loss of the hybrid frankfurters (P < 0.05). Meanwhile, as the level of substitution increased, the hybrid frankfurters had higher digestibility, poorer texture than the standard frankfurters, as well as the rheological behaviour of hybrid meat batters (P < 0.05). The results showed that a moderate level (15 %) of TMLP was used to replace lean pork could be potentially and successfully be used to produce hybrid frankfurters.

Characterization of non-volatile and volatile flavor profiles of Coregonus peled meat cooked by different methods

This study investigated the effects of different cooking methods on non-volatile flavor (free amino acids, 5'-nucleotides, and organic acids, etc.) of Coregonus peled meat. The volatile flavor characteristics were also analyzed by electric nose and gas chromatography-ion migration spectrometry (GC-IMS). The results indicated that the content of flavor substances in C. peled meat varied significantly. The electronic tongue results indicated that the richness and umami aftertaste of roasting were significantly greater. The content of sweet free amino acids, 5'-nucleotides, and organic acids was also higher in roasting group. Electronic nose principal component analysis can distinguish C. peled meat cooked (the first two components accounted for 98.50% and 0.97%, respectively). A total of 36 volatile flavor compounds were identified among different groups, including 16 aldehydes, 7 olefine aldehydes, 6 alcohols, 4 ketones, and 3 furans. In general, roasting was recommended and gave more flavor substances in C. peled meat.

Formation and Analysis of Volatile and Odor Compounds in Meat-A Review

The volatile composition and odor of meat and meat products is based on the precursors present in the raw meat. These are influenced by various pre-slaughter factors (species, breed, sex, age, feed, muscle type). Furthermore, post-mortem conditions (chiller aging, cooking conditions, curing, fermentation, etc.) determine the development of meat volatile organic compounds (VOCs). In this review, the main reactions leading to the development of meat VOCs such as the Maillard reaction; Strecker degradation; lipid oxidation; and thiamine, carbohydrate, and nucleotide degradation are described. The important pre-slaughter factors and post-mortem conditions influencing meat VOCs are discussed. Finally, the pros, cons, and future perspectives of the most commonly used sample preparation techniques (solid-phase microextraction, stir bar sorptive extraction, dynamic headspace extraction) and analytical methods (gas chromatography mass spectrometry and olfactometry, as well as electronic noses) for the analysis of meat VOCs are discussed, and the continued importance of sensorial analysis is pinpointed.

Progress of Research on the Application of Nanoelectronic Smelling in the Field of Food

In the past 20 years, the development of an artificial olfactory system has made great progress and improvements. In recent years, as a new type of sensor, nanoelectronic smelling has been widely used in the food and drug industry because of its advantages of accurate sensitivity and good selectivity. This paper reviews the latest applications and progress of nanoelectronic smelling in animal-, plant-, and microbial-based foods. This includes an analysis of the status of nanoelectronic smelling in animal-based foods, an analysis of its harmful composition in plant-based foods, and an analysis of the microorganism quantity in microbial-based foods. We also conduct a flavor component analysis and an assessment of the advantages of nanoelectronic smelling. On this basis, the principles and structures of nanoelectronic smelling are also analyzed. Finally, the limitations and challenges of nanoelectronic smelling are summarized, and the future development of nanoelectronic smelling is proposed.

Effect of different cooking water on flavor characteristics of mutton soup

The mutton flavor is affected by cooking water significantly, and the flavor of mutton is delicious and widely loved by consumers through an extremely simple processing in northwest China, such as Inner Mongolia, Ningxia, and Xinjiang. The flavor shows obvious changes if get out of these areas even use the same raw meat, which may be caused by different cooking water. To determine whether and how the cooking water affect the mutton soup flavor, the elements in water, the flavor was studied by inductively coupled plasma mass spectrometry (ICP-MS), amino acid analyzer, and thermal desorption (TDS)-gas chromatography-mass spectrometry (GC-MS). Specifically, three water samples from different sources, Ningxia (NXW), Beijing (BJW), and ultrapure water from the laboratory (PUW), were used for cooking with Tan sheep's ribs to get different mutton soups. The inductively coupled plasma mass spectrometry (ICP-MS) results showed that the elements and the concentration of solutes in different water sources were significantly different. The NXW batch had the highest Na, Mg, K, and Sr concentrations, and Na in NXW water reached to 50.60 mg/L, which existed as Na+, significantly (p < .05) higher than BJW (8.63 mg/L) and PUW, which were important to the flavor of mutton soup. The PUW batch had the highest content of free amino acids, and the content of glutamic acid (Glu) reached to 17.89 μg/mL. The NXW batch had the highest content of taste nucleotides, and the content of 5´-IMP reached to 68.68 μg/ml. The volatiles of the three batches had significant differences, and only 40 volatiles were detected in all batches. Further flavor studies using electronic nose and electronic tongue showed significant differences in overall aroma and overall taste, especially bitterness, saltiness, and astringency. The results could provide a basis for improving the flavor quality for the mutton soup.

Effects of different cooking methods on free fatty acid profile, water-soluble compounds and flavor compounds in Chinese Piao chicken meat

Piao chicken breast meat was cooked by three different methods (boiling, frying and roasting). Non-volatile and volatile substances in the three cooked chicken were analyzed by GC-MS, UPLC-Q-Exactive-MS and GC-IMS, respectively. Arachidonic acid was the highest in boiled chicken, oleic acid was the highest in roasted chicken, linoleic acid, EPA and DHA were the highest in fried chicken. Compared with the control group, the total content of small molecular metabolites of chicken in each treatment group decreased. The total amount of amino acids in roasted chicken was 2.90 times of that in boiled chicken (P < 0.05), and 2.23 times of that in fried chicken (P < 0.05). A total of 26 volatile flavor compounds were detected. Phenylacetaldehyde etc. were the main volatile flavor compounds in boiled chicken, 3-butanedione etc. were the main volatile flavor compounds in fried chicken, while 3-methylbutyraldehyde etc. were the main volatile flavor compounds in roasted chicken.

A novel method based on machine vision system and deep learning to detect fraud in turmeric powder

Assessing the quality of food and spices is particularly important in ensuring proper human nutrition. The use of computer vision method as a non-destructive technique in measuring the quality of food and spices has always been taken into consideration by researchers. Due to the high nutritional value of turmeric among the spices as well as the fraudulent motives to gain economic profit from the selling of this product, its quality assessment is very important. The lack of marketability of grade 3 chickpeas (small and broken chickpeas) and their very low price have made them a good choice to be mixed with turmeric in powder form and sold in the market. In this study, an improved convolutional neural network (CNN) was used to classify turmeric powder images to detect fraud. CNN was improved through the use of gated pooling functions. We also show with a combined approach based on the integration of average pooling and max pooling that the accuracy and performance of the proposed CNN has increased. In this study, 6240 image samples were prepared in 13 categories (pure turmeric powder, chickpea powder, chickpea powder mixed with food coloring, 10, 20, 30, 40 and 50% fraud in turmeric). In the preprocessing step, unwanted parts of the image were removed. The data augmentation (DA) was used to reduce the overfitting problem on CNN. Also in this research, MLP, Fuzzy, SVM, GBT and EDT algorithms were used to compare the proposed CNN results with other classifiers. The results showed that prevention of the overfitting problem using gated pooling, the proposed CNN was able to grade the images of turmeric powder with 99.36% accuracy compared to other classifiers. The results of this study also showed that computer vision, especially when used with deep learning (DL), can be a valuable method in evaluating the quality and detecting fraud in turmeric powder.