Coding smell patterns of crude oil by the electronic nose: A soil pollution case.

In our study, we leveraged an electronic nose to detect the patterns of crude oils and their mixtures, sourced from the oil fields from neighboring regions in pursuit of the task of environmental impact evaluation. The temporal dynamics of oil-related patterns acquired by an electronic nose was tracked to identify the influence of high or low emissions of volatiles that depend on the oil composition. Analyzing the oils by Fourier-transform IR-spectroscopy and GC×GC-MS, we confirmed the correlation between sensor responses and the oil compositions, significantly dependent on the ratio of aromatic compounds/alkanes. Using pattern recognition techniques, Random Forest classifier enabled good accuracy of classification of oil samples and contaminated soils underscoring a high-resolution distinction between the response data. Applying these principles to determine the oil origin, we observed that the studied oil samples and contaminated soil samples corroborate with the dynamic changes in odor patterns based only on volatile and semivolatile compounds. Crude oils from the border of two oil fields facilitate a change in the odor pattern to remain one of the fields depending on the weathering time. These proposed intelligent multisensor systems show great promise as a tool for estimating oil-contaminated soils, thereby potentially enhancing environmental monitoring practices.

A machine learning-based electronic nose for detecting neonatal sepsis: Analysis of volatile organic compound biomarkers in fecal samples.

BACKGROUND: Neonatal sepsis is a global health threat, contributing to high morbidity and mortality rates among newborns. Recognizing the profound impact of neonatal sepsis on long-term health outcomes emphasizes the critical need for timely detection to mitigate its consequences and ensure optimal health for the affected newborns. Currently, various diagnostic approaches have been implemented, but they are limited by their invasiveness, high costs, centralized testing, frequent delays or inaccuracies in results, and the need for sophisticated laboratory equipment. METHODS: We introduced a novel, non-invasive, cost-efficient, and easy-to-use technology that can provide rapid results at a point-of-care. The technology utilized a lab-built metal oxide semiconductor-based electronic nose (cNose) combined with volatile organic compound (VOC) biomarkers identified through gas chromatography-mass spectrometry (GC-MS) analysis. The system was evaluated using fecal profiling tests involving a total of 32 samples, including 17 positive and 15 negative sepsis, confirmed by blood culture. To assess the performance in discriminating patients from healthy controls, four machine learning algorithms were implemented. RESULTS: Based on the cross-validation results, the MLPNN model provided the best results in distinguishing between neonates with positive and negative sepsis, achieving high-performance results of 90.63% accuracy, 88.24% sensitivity, and 93.33% specificity at a 95% confidence interval. Specific VOCs associated with neonatal sepsis, such as alcohols, acids, and esters, were successfully identified through GC-MS analysis, further validating the diagnostic capability of the cNose device. CONCLUSION: The overall observations show the feasibility of using cNose system as a promising tool for a real-time and bedside sepsis detection, potentially improving patient outcomes.

Olfactory Diagnosis Model for Lung Health Evaluation Based on Pyramid Pooling and SHAP-Based Dual Encoders.

This study introduces a novel deep learning framework for lung health evaluation using exhaled gas. The framework synergistically integrates pyramid pooling and a dual-encoder network, leveraging SHapley Additive exPlanations (SHAP) derived feature importance to enhance its predictive capability. The framework is specifically designed to effectively distinguish between smokers, individuals with chronic obstructive pulmonary disease (COPD), and control subjects. The pyramid pooling structure aggregates multilevel global information by pooling features at four scales. SHAP assesses feature importance from the eight sensors. Two encoder architectures handle different feature sets based on their importance, optimizing performance. Besides, the model's robustness is enhanced using the sliding window technique and white noise augmentation on the original data. In 5-fold cross-validation, the model achieved an average accuracy of 96.40%, surpassing that of a single encoder pyramid pooling model by 10.77%. Further optimization of filters in the transformer convolutional layer and pooling size in the pyramid module increased the accuracy to 98.46%. This study offers an efficient tool for identifying the effects of smoking and COPD, as well as a novel approach to utilizing deep learning technology to address complex biomedical issues.

Imaging of Volatile Organic Compounds Using a Single Nanowire-Based Electronic Nose for Future Biomedical Applications.

This study introduces an array of semiconductor oxide single nanowires fabricated using advanced semiconductor processing techniques, including electron beam lithography and thin-film deposition, which is well-suited for large-scale nanowire integration. A four-channel nanowire array consisting of tin oxide (SnO2), indium oxide (In2O3), ferric oxide (Fe3O4), and titanium oxide (TiO2) was developed. As a proof of concept, we converted the response curves of the sensor array to heat maps, enabling comprehensive feature representation. The fabricated electronic nose (E-nose) was utilized to detect three types of volatile organic compounds (VOCs), with the results visualized in a heat map format. Additionally, the performance of each individual sensor was quantitatively studied, highlighting the array's potential for enhanced gas detection and analysis. To further illustrate the interaction between gas molecules and the nanowires, we visualized the gas response results by mapping the sensor's signal changes. These visualizations provide a clear representation of how different gas molecules interact with specific nanowires. For example, the heat maps reveal distinct response patterns for each type of VOC, allowing for the identification and differentiation of gases based on their unique signatures. This visualization technique not only enhances the understanding of gas-nanowire interactions but also demonstrates the effectiveness of the E-nose in distinguishing between various VOCs. The SnO2 nanowire gas sensor showed enhanced gas response compared to other materials. The SnO2 and TiO2 gas sensors showed enhanced response (62 and 56 s) and recovery times (100 and 37 s).

Data set for Gambung green tea aroma using on electronic nose.

OBJECTIVES: In recent years, there has been much discussion and research on electronic nose (e-nose). This topic has developed mainly in the medical and food fields. Typically, e-nose is combined with machine learning algorithms to predict or detect multiple sensory classes in each tea sample. Therefore, in e-nose systems, e-nose signal processing is an important part. In many situations, a comprehensive set of experiments is required to ensure the prediction model can be generalized well. This data set specifically focuses on two main goals such as classification of green tea quality and prediction of organoleptic score. In this experiment, Gambung dry green tea samples were used. The challenge is that dry tea does not emit as strong an aroma as tea infusions, making it more difficult for the e-nose system to detect and identify the aromas. This data set offers a valuable resource for researchers and developers to conduct investigations and experiments by classifying and detecting organoleptic scores that aim to categorize and identify organoleptic ratings. This enables a deeper understanding of the quality of dry green tea and encourages further integration of e-nose technology in the tea industry. DATA DESCRIPTION: This experiment focused on analyzing green tea aroma using six gas sensors. Seventy-eight green tea samples were tested, each observed three times, using a tea chamber connected to a sensor chamber via a hose and an intake micro air pump. Air flowed from the tea chamber to the sensor chamber for 60 s, followed by 60 s of aroma data recording. This data was saved into CSV files and labeled according to the Indonesian National Standard (SNI) 3945:2016, which includes special and general requirements for green tea quality. An organoleptic test by a tea tester further labeled the data set into "good" or "quality defect" for classification and provided organoleptic scores based on dry appearance, brew color, taste, aroma, and dregs of brewing for continuous label.

Change in volatile profiles of wheat flour during maturation.

The volatile profiles of wheat flour during maturation were examined through headspace solid-phase micro-extraction gas chromatography-mass spectrometry (HS-SPME-GC/MS) combined with electronic nose (E-nose) and electronic tongue (E-tongue) analyses. The wheat flour underwent maturation under three distinct conditions for predetermined durations. While GC/MS coupled with E-tongue exhibited discernment capability among wheat flour samples subjected to varying maturation conditions, E-nose analysis solely relying on principal component analysis failed to achieve discrimination. 83 volatile compounds were identified in wheat flour, with the highest abundance observed in samples matured for 50 d at 25 °C. Notably, trans-2-Nonenal, decanal, and nonanal were the main contributors to the characteristic flavor profile of wheat flour. Integration of HS-SPME-GC/MS with E-tongue indicated superior flavor development and practical viability in wheat flour matured for 50 d at 25 °C. This study furnishes a theoretical groundwork for enhancing the flavor profiles of wheat flour and its derivative products.

[Development of an Odor Evaluation Method for Citrus Unshiu Peel Using an Electronic Nose Based on the Intensity of per Unit Length (INPULTH)].

Although odor is an important indicator of herbal medicine quality, an objective odor evaluation method remains undiscovered. Quantitative measurement using previous methods is complicated as Citrus Unshiu Peel (Chimpi) emits an odor when broken. To establish odor evaluation methods for herbal medicines using chimpi as an example, we developed a reproducible method for breaking samples and an objective odor evaluation method using an electronic nose (e-nose). First, an odor-emitting device (OED) was fabricated by modifying a pill cutter, which suppressed the spread of odor components into the room air while cutting the samples. The odor was emitted from chimpi with an OED and measured using an e-nose. The cut length was then measured. The sensor intensity was positively correlated with the cut length (r = 0.840-0.927) in the same sample, and the intensity per unit length (INPULTH) calculated from the sensor intensity and cut length enables the comparison of the sensor intensity among different samples. In addition, average d-limonene emission level measured by GC-MS was positively correlated with average INPULTH (r = 0.999), which suggests that this OED and e-nose method enables the comparison of the sensor intensity and d-limonene emissions. INPULTH also positively correlated with other seven monoterpenes such as p-cymene, ß-myrcene, ß-phellandrene, α-pinene, ß-pinene, γ-terpinene, and α-terpinolene (r = 0.701-0.865). Therefore, monoterpene content can be evaluated by measuring the odor in the same way as d-limonene. In conclusion, we developed a simple odor intensity evaluation method optimized for chimpi to establish an odor evaluation method for herbal medicines.

A comprehensive review on advancements in sensors for air pollution applications.

Air pollution, originating from both natural and human-made sources, presents significant threats to human health and the environment. This review explores the latest technological advancements in air quality sensors focusing on their applications in monitoring a wide range of pollution sources from volcanic eruptions and wildfires to industrial emissions, transportation, agricultural activities and indoor air quality. The review categorizes these sources and examines the operational principles, system architectures, and effectiveness of various air quality monitoring instruments including low-cost sensors, gas analyzers, weather stations, passive sampling devices and remote sensing technologies such as satellite and LiDAR. Key insights include the rapid evolution of sensor technology driven by the need for more accurate, real-time monitoring solutions that are both cost-effective and widely accessible. Despite significant advancements, challenges such as sensor calibration, standardization, and data integration remain critical for ensuring reliable air quality assessments. The manuscript concludes by emphasizing the need for continued innovation and the integration of advanced sensor technologies with regulatory frameworks to enhance environmental management and public health protection.

Application of electronic nose technology in the diagnosis of gastrointestinal diseases: a review.

Electronic noses (eNoses) are electronic bionic olfactory systems that use sensor arrays to produce response patterns to different odors, thereby enabling the identification of various scents. Gastrointestinal diseases have a high incidence rate and occur in 9 out of 10 people in China. Gastrointestinal diseases are characterized by a long course of symptoms and are associated with treatment difficulties and recurrence. This review offers a comprehensive overview of volatile organic compounds, with a specific emphasis on those detected via the eNose system. Furthermore, this review describes the application of bionic eNose technology in the diagnosis and screening of gastrointestinal diseases based on recent local and international research progress and advancements. Moreover, the prospects of bionic eNose technology in the field of gastrointestinal disease diagnostics are discussed.

Electronic Noses: From Gas-Sensitive Components and Practical Applications to Data Processing.

Artificial olfaction, also known as an electronic nose, is a gas identification device that replicates the human olfactory organ. This system integrates sensor arrays to detect gases, data acquisition for signal processing, and data analysis for precise identification, enabling it to assess gases both qualitatively and quantitatively in complex settings. This article provides a brief overview of the research progress in electronic nose technology, which is divided into three main elements, focusing on gas-sensitive materials, electronic nose applications, and data analysis methods. Furthermore, the review explores both traditional MOS materials and the newer porous materials like MOFs for gas sensors, summarizing the applications of electronic noses across diverse fields including disease diagnosis, environmental monitoring, food safety, and agricultural production. Additionally, it covers electronic nose pattern recognition and signal drift suppression algorithms. Ultimately, the summary identifies challenges faced by current systems and offers innovative solutions for future advancements. Overall, this endeavor forges a solid foundation and establishes a conceptual framework for ongoing research in the field.

An Exploratory Study on the Rapid Detection of Volatile Organic Compounds in Gardenia Fruit Using the Heracles NEO Ultra-Fast Gas Phase Electronic Nose.

Gardenia fruit is a popular functional food and raw material for natural pigments. It comes from a wide range of sources, and different products sharing the same name are very common. Volatile organic compounds (VOCs) are important factors that affect the flavor and quality of gardenia fruit. This study used the Heracles NEO ultra-fast gas phase electronic nose with advanced odor analysis performance and high sensitivity to analyze six batches of gardenia fruit from different sources. This study analyzed the VOCs to find a way to quickly identify gardenia fruit. The results show that this method can accurately distinguish the odor characteristics of various gardenia fruit samples. The VOCs in gardenia fruit are mainly organic acid esters, ketones, and aldehyde compounds. By combining principal component analysis (PCA) and discriminant factor analysis (DFA), this study found that the hexanal content varied the most in different gardenia fruit samples. The VOCs allowed for the fruit samples to be grouped into two main categories. One fruit sample was quite different from the fruits of other origins. The results provide theoretical support for feasibility of rapid identification and quality control of gardenia fruit and related products in the future.

Quality characterization of tobacco flavor and tobacco leaf position identification based on homemade electronic nose.

A set of nine unique tobacco extract samples was analyzed using a self-developed electronic nose (E-nose) system, a commercial E-nose, and gas chromatography-mass spectrometry (GC-MS). The evaluation employed principal component analysis, statistical quality control, and soft independent modeling of class analogies (SIMCA). These multifaceted statistical methods scrutinized the collected data. Subsequently, a quality control model was devised to assess the stability of the sample quality. The results showed that the custom E-nose system could successfully distinguish between tobacco extracts with similar odors. After further training and the development of a quality control model for accepted tobacco extracts, it was possible to identify samples with normal and abnormal quality. To further validate our E-nose and extend its use within the tobacco industry, we collected and accurately classified the flavors of different tobacco leaf positions, with a remarkable accuracy rate of 0.9744. This finding facilitates the practical application of our E-nose system for the efficient identification of tobacco leaf positions.

Firewood as a tool to valorize meat: A sensory and instrumental approach to grilled veal, lamb, and pork meat.

Two firewood species (beech and olive) were used for grilling three meat types (lamb, pork, and veal) to assess their influence on the sensorial properties of meat. A multimethod approach was adopted, including sensory evaluation with consumers and two analytical techniques to characterize the volatile fraction (Solid-Phase Micro-Extraction Gas Chromatography-Mass Spectrometry [SPME-GC/MS] and electronic nose [e-nose]). The sensory session included three pairwise preference tests (one for each type of meat), an overall liking test, a Rate-All-That-Apply test, and a questionnaire on the interest and perceived value of using sustainably certified firewood in food preparation. The firewood species significantly affected the perception of a few crucial attributes. In particular, olive wood increased the roasted meat flavor perception in lamb and veal, while beech wood increased the perceived intensity of a vegetable/herbaceous flavor in veal. No effect of firewood was observed on preference within each pair of meat samples. Lamb was the significantly most liked meat by consumers, followed by pork; veal was the least liked meat type. Positive and negative drivers of preference were discussed. 36 volatile organic compounds were identified from SPME-GC/MS in meats. Congruently with sensory data, the two veal samples showed a greater distance in terms of volatile composition. Relative distances among samples on maps obtained from SPME-GC/MS and the e-nose were similar. This multi-method approach innovatively showed the potential of using firewood as a 'gastronomic' tool to sensorially characterize and valorize cooked meat.

The Effect of Phytogenic Additive in Broiler's Diet on Production Results, Physicochemical Parameters, and the Composition of Volatile Organic Compounds of Broiler Meat Assessed by an Electronic Nose System.

The primary objective of this study was to investigate the impact of a phytogenic additive (PA) in broiler chickens' diet on production, physiochemical parameters, and the profile of volatile organic compounds present in broiler chickens' meat. The experiment was conducted in a commercial chicken house, where Ross 308 broiler chickens were divided into two groups, each consisting of 65,000 broilers. One group was fed a diet supplemented with 100 ppm of PA throughout the rearing period. The primary chemical composition of the meat and its physicochemical parameters were determined. A visual assessment of breast muscles for defects and volatile organic compounds were evaluated using an electronic nose system. No statistically significant differences were shown in the production performance of the chickens; while summarizing all production parameters, a higher EPEF index of 31 points in the experimental group was highlighted. Breast muscle quality showed differences in drip loss and WHC (p ≤ 0.01) in favor of the experimental group, and a lower cutting force value (p ≤ 0.05) was found for breast muscles from the experimental group. The group also had a lower proportion of muscles with a white striping defect, and the results of volatile organic compound profiling showed the most aroma units.

Effects of Wickerhamomyces anomalus Co-Fermented with Saccharomyces cerevisiae on Volatile Flavor Profiles during Steamed Bread Making Using Electronic Nose and HS-SPME-GC-MS.

Steamed bread is a traditional staple food in China, and it has gradually become loved by people all over the world because of its healthy production methods. With the improvement in people's living standards, the light flavor of steamed bread fermented by single yeast cannot meet people's needs. Multi-strain co-fermentation is a feasible way to improve the flavor of steamed bread. Here, the dynamic change profiles of volatile substances in steamed bread co-fermented by Saccharomyces cerevisiae SQJ20 and Wickerhamomyces anomalus GZJ2 were analyzed using the electronic nose (E-nose) and headspace solid-phase microextraction combined with gas chromatography-mass spectrometry (HS-SPME-GC-MS). The five detectors of the E-nose rapidly detected the changes in volatile substances in different dough or steamed bread with the highest response value in co-fermented dough. A total of 236 volatile substances were detected in all the samples using HS-SPME-GC-MS, and alcohols were the most variable component, especially Phenylethyl alcohol. Significantly, more alcohols and esters were upregulated in co-fermented dough, and the addition of W. anomalus GZJ2 improved the key volatile aroma compounds of steamed bread using the relative odor activity value method (ROAV), especially the aldehydes and alcohols. Moreover, these key volatile aroma compounds can be quickly distinguished using the W2S detector of the E-nose, which can be used for the rapid detection of aroma components in steamed bread.

Assessment of 'Golden Delicious' Apples Using an Electronic Nose and Machine Learning to Determine Ripening Stages.

Consumers often face a lack of information regarding the quality of apples available in supermarkets. General appearance factors, such as color, mechanical damage, or microbial attack, influence consumer decisions on whether to purchase or reject the apples. Recently, devices known as electronic noses provide an easy-to-use and non-destructive assessment of ripening stages based on Volatile Organic Compounds (VOCs) emitted by the fruit. In this study, the 'Golden Delicious' apples, stored and monitored at the ambient temperature, were analyzed in the years 2022 and 2023 to collect data from four Metal Oxide Semiconductor (MOS) sensors (MQ3, MQ135, MQ136, and MQ138). Three ripening stages (less ripe, ripe, and overripe) were identified using Principal Component Analysis (PCA) and the K-means clustering approach from various datasets based on sensor measurements in four experiments. After applying the K-Nearest Neighbors (KNN) model, the results showed successful classification of apples for specific datasets, achieving an accuracy higher than 75%. For the dataset with measurements from all experiments, an impressive accuracy of 100% was achieved on specific test sets and on the evaluation set from new, completely independent experiments. Additionally, correlation and PCA analysis showed that choosing two or three sensors can provide equally successful results. Overall, the e-nose results highlight the importance of analyzing data from several experiments performed over a longer period after the harvest of apples. There are similarities and differences in investigated VOC parameters (ethylene, esters, alcohols, and aldehydes) for less or more mature apples analyzed during autumn or spring, which can improve the determination of the ripening stage with higher predicting success for apples investigated in the spring.

Effects of Different Fermentation Methods on Flavor Quality of Liupao Tea Using GC-Q-TOF-MS and Electronic Nose Analyses.

To further develop Liupao tea products and enhance their flavor, this study investigated the effects of different fermentation methods on the aroma quality of Liupao tea. The aroma quality of Liupao tea was comprehensively analyzed using HS-SPME in combination with GC-Q-TOF-MS, electronic nose, and sensory evaluations. Electronic nose detection showed that the aroma fingerprints of Liupao tea samples with different fermentation methods were different. Sulfides, alcohols, ketones, and methyls were the main aroma categories affecting the aroma of the four groups of Liupao tea samples. GC-Q-TOF-MS analysis revealed significant differences in the composition of aroma components among the four fermentation methods of Liupao tea (p < 0.05). Furthermore, the total amount of aroma compounds was found to be highest in the group subjected to hot fermentation combined with the inoculation of Monascus purpureus (DMl group). Based on the OPLS-DA model, candidate differential aroma components with VIP > 1 were identified, and characteristic aroma compounds were selected based on OAV > 10. The key characteristic aroma compounds shared by the four groups of samples were 1,2,3-Trimethoxybenzene with a stale aroma and nonanal with floral and fruity aromas. The best sensory evaluation results were obtained for the DMl group, and its key characteristic aroma compounds mainly included 1,2,3-Trimethoxybenzene, nonanal, and cedrol. The results of this study can guide the development of Liupao tea products and process optimization.

Comparison of Different Deodorizing Treatments on the Flavor of Paddy Field Carp, Analyzed by the E-Nose, E-Tongue and Gas Chromatography-Ion Mobility Spectrometry.

Changes in the flavor and taste profiles of Paddy Field Carp after deodorization with perilla juice (PJ), cooking wine (CW) and a mixture of the two (PJ-CW) were analyzed using the E-nose, E-tongue, gas chromatography-ion mobility spectrometry (GC-IMS), free amino acid analysis and taste nucleotide analysis. The E-nose and E-tongue revealed that deodorization reduced the content of sulfur-containing compounds, enhanced umami, bitterness, sourness and astringency, and decreased saltiness. PCA and OPLS-DA analysis successfully distinguished between the effects of the treatments. Free amino acids increased from 8777.67 to 11,125.98 mg/100 g and umami amino acids increased from 128.24 to 150.37 mg/100 g after PJ-CW deodorization (p < 0.05). Equivalent umami concentration (EUC) comparisons showed that PJ-CW treatment produced the greatest synergistic umami enhancement (to 3.15 g MSG equiv./100 g). GC-IMS detected 52 aroma compounds; PJ treatment produced the greatest diversity of aldehydes, including heptanal, nonanal, hexanal, 3-methylbutanal, (E)-2-heptenal and (E,E)-2,4-heptadienal. The total content of volatile flavor compounds was the highest after PJ-CW treatment, and the content of many characteristic flavor substances (3-hydroxy-2-butanone, benzaldehyde, 5-methyl-2(3H)-furanone) increased. These findings provided a theoretical basis for the further development of deodorization methods for Paddy Field Carp.

Novel Insights into the Effects of Different Cooking Methods on Salted Egg Yolks: Physicochemical and Sensory Analysis.

In this study, the flavor characteristics and physicochemical properties of salted egg yolk (SEY) under different cooking methods (steaming/baking/microwaving) were investigated. The microwave-treated SEY exhibited the highest levels of salt content, cooking loss, lightness, and b* value, as well as the highest content of flavor amino acids. A total of 31, 27, and 29 volatile compounds were detected after steaming, baking, and microwave treatments, respectively, covering 10 chemical families. The partial least squares discriminant analysis confirmed that 21 compounds, including octanol, pyrazine, 2-pentyl-furan, and 1-octen-3-ol, were the key volatile compounds affecting the classification of SEY aroma. The electronic nose revealed a sharp distinction in the overall flavor profile of SEY with varying heat treatments. However, no dramatic differences were observed in terms of fatty acid composition. Microwave treatment was identified as presenting a promising approach for enhancing the aroma profile of SEY. These findings contribute novel insights into flavor evaluation and the development of egg products as ingredients for thermal processing.

E-Nose: Time-Frequency Attention Convolutional Neural Network for Gas Classification and Concentration Prediction.

In the electronic nose (E-nose) systems, gas type recognition and accurate concentration prediction are some of the most challenging issues. This study introduced an innovative pattern recognition method of time-frequency attention convolutional neural network (TFA-CNN). A time-frequency attention block was designed in the network, aiming to excavate and effectively integrate the temporal and frequency domain information in the E-nose signals to enhance the performance of gas classification and concentration prediction tasks. Additionally, a novel data augmentation strategy was developed, manipulating the feature channels and time dimensions to reduce the interference of sensor drift and redundant information, thereby enhancing the model's robustness and adaptability. Utilizing two types of metal-oxide-semiconductor gas sensors, this research conducted qualitative and quantitative analysis on five target gases. The evaluation results showed that the classification accuracy could reach 100%, and the coefficient of the determination (R2) score of the regression task was up to 0.99. The Pearson correlation coefficient (r) was 0.99, and the mean absolute error (MAE) was 1.54 ppm. The experimental test results were almost consistent with the system predictions, and the MAE was 1.39 ppm. This study provides a method of network learning that combines time-frequency domain information, exhibiting high performance in gas classification and concentration prediction within the E-nose system.