Performance Of Electronic Nose Research Articles

Multipressure Sampling for Improving the Performance of MOF-based Electronic Noses.

Metal-organic frameworks (MOFs) are a promising class of porous materials for the design of gas sensing arrays, which are often called electronic noses. Due to their chemical and structural tunability, MOFs are a highly diverse class of materials that align well with the similarly diverse class of volatile organic compounds (VOCs) of interest in many gas detection applications. In principle, by choosing the right combination of cross-sensitive MOFs, layered on appropriate signal transducers, one can design an array that yields detailed information about the composition of a complex gas mixture. However, despite the vast number of MOFs from which one can choose, gas sensing arrays that rely too heavily on distinct chemistries can be impractical from the cost and complexity perspective. On the other hand, it is difficult for small arrays to have the desired selectivity and sensitivity for challenging sensing applications, such as detecting weakly adsorbing gases with weak signals, or conversely, strongly adsorbing gases that readily saturate MOF pores. In this work, we employed gas adsorption simulations to explore the use of a variable pressure sensing array as a means of improving both sensitivity and selectivity as well as increasing the information content provided by each array. We studied nine different MOFs (HKUST-1, IRMOF-1, MgMOF-74, MOF-177, MOF-801, NU-100, NU-125, UiO-66, and ZIF-8) and four different gas mixtures, each containing nitrogen, oxygen, carbon dioxide, and exactly one of the hydrogen, methane, hydrogen sulfide, or benzene. We found that by lowering the pressure, we can limit the saturation of MOFs, and by raising the pressure, we can concentrate weakly adsorbing gases, in both cases, improving gas detection with the resulting arrays. In many cases, changing the system pressure yielded a better improvement in performance (as measured by the Kullback-Liebler divergence of gas composition probability distributions) than including additional MOFs. We thus demonstrated and quantified how sensing at multiple pressures can increase information content and cross-sensitivity in MOF-based arrays while limiting the number of unique materials needed in the device.

Contrastive domain generalization convolution neural network correcting the drift of gas sensors

The drift problems of gas sensors caused by aging ingredients and unexpected environmental impacts, which can severely decrease the performance and service life of electronic noses, have been widely discussed. Many scholars believe that the essential reason for sensor drift can be attributed to the different data distributions in the latent feature space. However, traditional drift compensation methods like OSC, GasNet, SniffMultinose, and SniffConv are costly and complex. This paper has introduced an algorithm called contrastive domain generalization convolution neural network (CDCNN) to resolve this problem. For the first time, domain generalization and contrastive learning were used as the drift compensation methods for the gas sensor. A novel data augmentation was proposed to enrich the datasets. Feature generation module is used to simulate the drift of gas sensors. Contrastive learning adapts to the unseen areas in the latent feature space. The artificially generated features and the original features are drawn closer in the feature space to improve the algorithm's generalization ability. Experiments on long-term drift data show that CDCNN achieves high accuracy (0.7230). The experimental results show that the CDCNN algorithm is more suitable for practical applications due to less resource consumption and looser constraints.

How well does your e-nose detect cancer? Application of artificial breath analysis for performance assessment

Comparing electronic nose (e-nose) performance is a challenging task because of a lack of standardised method. This paper proposes a method for defining and quantifying an indicator of the effectiveness of multi-sensor systems in detecting cancers by artificial breath analysis. To build this method, an evaluation of the performances of an array of metal oxide sensors built for use as a lung cancer screening tool was conducted. Breath from 20 healthy volunteers has been sampled in fluorinated ethylene propylene sampling bags. These healthy samples were analysed with and without the addition of nine volatile organic compound (VOC) cancer biomarkers, chosen from literature. The concentration of the VOC added was done in increasing amounts. The more VOC were added, the better the discrimination between ‘healthy’ samples (breath without additives) and ‘cancer’ samples (breath with additives) was. By determining at which level of concentration the e-nose fails to reliably discriminate between the two groups, we estimate its ability to well predict the presence of the disease or not in a realistic situation. In this work, a home-made e-nose is put to the test. The results underline that the biomarkers need to be about 5.3 times higher in concentration than in real breath for the home-made nose to tell the difference between groups with a sufficient confidence.

Effect of ZnO morphologies on its sensor response and corresponding E-nose performance

Understanding the intricate relationship among the gas sensing data produced from different ZnO morphologies is crucial for optimizing electronic nose (e-nose) performance. In the present work, three different morphologies of ZnO namely nanoflowers (NFs), nanogranules (NGs) and nanowires (NWs) have been successfully synthesised and used for e-nose application. Incorporation of Pt in the reaction mixture has been found to be responsible for change in the morphology from NWs to NGs. The partial specificity of the developed sensors towards H2S, NO and their mixture has been used effectively for discrimination of gases. For this, the data repository was created utilizing labelled data as well as feature data. Elaborate study on principal component analysis shows well discrimination of target gas mixture. Importantly, using support vector machine it has been demonstrated that a wise selection of combination of sensors can give a very good accuracy on the prediction of target gas mixture.

Evaluation of performance of metal oxide electronic nose for detection of aflatoxin in artificially and naturally contaminated maize

Aflatoxins are of great concern for food safety and security due to their impact on human health and the agriculture economy in developing countries. This study aimed to evaluate the potential use of a field portable metal oxide sensors based electronic nose to detect aflatoxin contamination in Kenyan maize varieties that were artificially and naturally infected with Aspergillus flavus. Mutual information was used to select features from the electronic nose sensor signals for classification of the samples. The effectiveness of selected features to discriminate between the different classes of samples was evaluated by support vector machines and k-nearest neighbour with leave-one-out cross-validation. External validation was also conducted by analysing samples naturally contaminated with A. flavus using the classification model generated with samples that had been artificially inoculated with the aflatoxigenic A. flavus. Cross-validated classification accuracies ranged from 72% to 88% for maize samples artificially inoculated with A. flavus and 61–86% for samples naturally infected with A. flavus. Classification accuracies achieved with external validation for maize samples naturally contaminated with aflatoxins ranged from 58% to 78% and were relatively consistent with accuracies obtained from internal validation. Results suggest that the electronic nose could be a promising cost-effective screening method to detect aflatoxin contamination in maize.

Chemical selection for the calibration of general-purpose electronic noses based on Silhouette coefficients

Sensor drift is often application-dependent and results in a reduction in the overall long-term performance of electronic noses. Even with drift compensation it is remains challenging to transfer these models to other application scenarios. In order to remedy this deficiency, different/generic chemicals are needed to provide a general-purpose calibration approach that can be applied to a wide range of electronic noses. In this article, we investigated a method to identify these chemicals based on four criteria (universality, safety, sensibility, and differentiation). This concept was tested on an in-house electronic nose comprising 37 gas sensors and four environmental sensors combined with an automatic gas acquisition system. The 14 different volatile compounds were tested over four months. The silhouette coefficient was used to evaluate the extent of the sensor drift. Six different chemicals (acetone, alcohol, ethyl acetate, tetrahydrofuran, acetaldehyde, and n-hexane) were finally selected as the most appropriate to calibrate our electronic nose (E-nose). We believe our research may motivate the design of a reasonable chemical selection method for the calibration of general-purpose E-noses.

Prediction of CO concentration in different conditions based on Gaussian-TCN

Accurate prediction of gas concentrations can help people to know if there is hazardous gas in the environment, thus protecting people's health. The electronic nose (E-nose) performance in predicting gas concentrations mostly depends on its algorithm. In recent years, recurrent neural networks (RNNs) have demonstrated their suitability for processing time series data. Typical models are long short-term memory (LSTM) neural networks and gated recursive units (GRUs). However, traditional recurrent neural networks use a gate structure to drop some information randomly, which can cause important features to be discarded. The temporal convolutional neural network (TCN) based on convolutional structures is more suitable for time series prediction. To enhance the accurate model forecasts, we first optimized the critical parameters in the model. Then, we adapted the residual structure of TCN and replaced the activation function of TCN with Gaussian error linear units (GELUs), And the modified model is named Gaussian-TCN. Finally, the experimental results show that the proposed Gaussian-TCN outperforms traditional recurrent networks regarding prediction accuracy.

Performance of electronic nose based on gas sensor-partition column for synthetic flavor classification

Electronic nose (e-nose) has been developed and implemented in a wide area, included in food industries. This study was conducted to investigate the performance of an e-nose that utilizes a packed gas chromatography column and a gas sensor for classification of synthetic flavor products. There were six aroma variants of synthetic flavor evaluated, namely durian, jackfruit, ambonese banana, melon, orange and lemon. The e-nose was designed with four main parts, namely aroma provider, column and detector room, microcontroller, and data acquisition system. The device was operated automatically at a stable temperature of 60 °C. Collected data consisted of ten data of each sample was preprocessed by baseline equalization and normalization, extracted its distinctive feature and then were analyzed through pattern recognition analysis. There were two kinds of methods used to analyzed the patterns of the data, namely a fuzzy c-means clustering and an artificial neural network (ANN). With the fuzzy c-means clustering, the result was six data clusters with an unbalanced number of members, indicated that this analysis could not classify samples properly. Meanwhile, analysis with the ANN could classify properly the samples with the level of accuracy of 70%.

A global extended extreme learning machine combined with electronic nose for identifying tea gas information

To enhance the detection performance of electronic nose (e-nose), a recognition method of gas feature based on a global extended extreme learning machine (GEELM) is proposed, which combines the expansion factor and global balance coefficient to expand and balance the difference between categories, and improve the classification performance. Then this method is applied to identify the quality of tea. Firstly, the dragging factor and following matrix are introduced to increase the distance between classes. Secondly, the global identification coefficient is introduced further to increase the feature differences among different types of tea, and improve the classification stability. Finally, under different feature sets, the classification performance of multi-pattern recognition methods is compared to prove the effectiveness of GEELM in e-nose gas feature recognition. The results show that GEELM has the best classification accuracy of 98.20%, F1-score of 0.9871, and Kappa coefficient of 0.9775. In conclusion, GEELM can be an effective technique to identify gas features, and it also provides a new method for tea quality measurement.

Diagnostic Performance of Electronic Noses in Cancer Diagnoses Using Exhaled Breath

There has been a growing interest in the use of electronic noses (e-noses) in detecting volatile organic compounds in exhaled breath for the diagnosis of cancer. However, no systematic evaluation has been performed of the overall diagnostic accuracy and methodologic challenges of using e-noses for cancer detection in exhaled breath. To provide an overview of the diagnostic accuracy and methodologic challenges of using e-noses for the detection of cancer. An electronic search was performed in the PubMed and Embase databases (January 1, 2000, to July 1, 2021). Inclusion criteria were the following: (1) use of e-nose technology, (2) detection of cancer, and (3) analysis of exhaled breath. Exclusion criteria were (1) studies published before 2000; (2) studies not performed in humans; (3) studies not performed in adults; (4) studies that only analyzed biofluids; and (5) studies that exclusively used gas chromatography-mass spectrometry to analyze exhaled breath samples. PRISMA guidelines were used for the identification, screening, eligibility, and selection process. Quality assessment was performed using Quality Assessment of Diagnostic Accuracy Studies 2. Generalized mixed-effects bivariate meta-analysis was performed. Main outcomes were sensitivity, specificity, and mean area under the receiver operating characteristic curve. This review identified 52 articles with a total of 3677 patients with cancer. All studies were feasibility studies. The sensitivity of e-noses ranged from 48.3% to 95.8% and the specificity from 10.0% to 100.0%. Pooled analysis resulted in a mean (SE) area under the receiver operating characteristic curve of 94% (95% CI, 92%-96%), a sensitivity of 90% (95% CI, 88%-92%), and a specificity of 87% (95% CI, 81%-92%). Considerable heterogeneity existed among the studies because of differences in the selection of patients, endogenous and exogenous factors, and collection of exhaled breath. Results of this review indicate that e-noses have a high diagnostic accuracy for the detection of cancer in exhaled breath. However, most studies were feasibility studies with small sample sizes, a lack of standardization, and a high risk of bias. The lack of standardization and reproducibility of e-nose research should be addressed in future research.

Accuracy of a new electronic nose for prostate cancer diagnosis in urine samples.

ObjectiveTo evaluate the accuracy of a new electronic nose to recognize prostate cancer in urine samples.MethodsA blind, prospective study on consecutive patients was designed. Overall, 174 subjects were included in the study: 88 (50.6%) in prostate cancer group, and 86 (49.4%) in control group. Electronic nose performance for prostate cancer was assessed using sensitivity and specificity. The diagnostic accuracy of electronic nose was reported as area under the receiver operating characteristic curve.ResultsThe electronic nose in the study population reached a sensitivity 85.2% (95% confidence interval 76.1–91.9; 13 false negatives out of 88), a specificity 79.1% (95% confidence interval 69.0–87.1; 18 false positives out of 86). The accuracy of the electronic nose represented as area under the receiver operating characteristic curve 0.821 (95% confidence interval 0.764–0.879).ConclusionsThe diagnostic accuracy of electronic nose for recognizing prostate cancer in urine samples is high, promising and susceptible to supplemental improvement. Additionally, further studies will be necessary to design a clinical trial to validate electronic nose application in diagnostic prostate cancer nomograms.

Evaluation of Performance of Metal Oxide Electronic Nose for Detection of Aflatoxin in Artificially and Naturally Contaminated Maize

Evaluation of Performance of Metal Oxide Electronic Nose for Detection of Aflatoxin in Artificially and Naturally Contaminated Maize

P26 - A large pilot study on the diagnostic performance of electronic nose in detecting Bladder Cancer

P26 - A large pilot study on the diagnostic performance of electronic nose in detecting Bladder Cancer

A novel WWH problem-based semi-supervised online method for sensor drift compensation in E-nose

Sensor drift caused by the aging components and unsuspected environmental factors is an urgent problem to be solved, as it seriously affects the detection performance and service life of electronic nose (E-nose). Existing researches mainly resorted to offline compensation techniques. Nevertheless, due to the dynamic and uncertainty of sensor drift, the offline techniques are not suitable for practical application scenarios. For this reason, the methods of compensating sensor drift online have been attracting more and more attention. To achieve the online compensation, three problems about the prediction model updating should be addressed first: ① When to update the prediction model (When); ② Which samples are used to update the prediction model (Which); ③ How to update the prediction model (How), that is, a WWH-problem. For addressing the three problems, a WWH problem-based semi-supervised online (WWH-SSO) method is proposed in this paper. The proposed WWH-SSO uses the unlabeled samples collected in the work process of E-nose to update the prediction model for realizing the unsupervised and online drift compensation. The sensor drift benchmark dataset collected by A. Vergara is used to verify the effectiveness of proposed method. The experiment results show that the sensor drift can be satisfactorily compensated online.

Reducing the destructive effect of ambient humidity variations on gas detection capability of a temperature modulated gas sensor by calcium chloride

Despite the successful detection performance of electronic nose in laboratories, they face challenges for use in the industry due to their accuracy reduction resulted from variable ambient properties especially relative humidity (RH) variation which is studied here. Responses of a single temperature modulated metal oxide gas sensor have been analyzed by SVM and k-NN methods to achieve a detector for low concentration levels of acetone, ethanol, 1-propanol and 1-butanol in air. The classifier models were designed and tested under different train-test conditions which showed that studied gases can be detected by the classifier if only they were measured in the same train and test environmental conditions and deviation of humidity level from train condition, reduces the detection accuracy to less than 60%. The accuracy increases by expanding the training dataset and training the system with responses carried out for gas with various RH contents. It was also shown that by using CaCl2 at the rout of gas flow, the destructive effect of RH variation is reduced and the detection accuracy increases to above 90%, while to achieve this accuracy, it is not necessary to train the system in all humidity conditions. By this method, the number of required test for system training reduces drastically. The method can be generalized to other electronic nose and gas detectors which suffer from humidity variations.

Sensory and Physicochemical Evaluation of Acacia and Linden Honey Adulterated with Sugar Syrup.

Honey is produced by honeybees and is used as a food and medical product. Adulteration of honey has been a problem for several years now because of the relatively high price of honey on the market according to its valuable composition. The aim of our study is to determine the physicochemical properties of authentic Hungarian linden and acacia honeys (pure samples or manipulated ones blended with sugar syrup) as well as commercially available blends of European Union (EU) non-European Union (non-EU) honeys. Authentic linden and acacia were blended with sugar syrup at 10%, 20% and 50% concentration levels, and physicochemical properties were determined according to the methods of the International Honey Commission. Our objectives also included testing of the performance of electronic sensory techniques (electronic tongue (ET) and electronic nose (EN)) in the detection of adulteration, and the results are compared to the sensory profile analysis. The results provide good average recognition and prediction abilities for the classification of adulterated and authentic honeys (>90% for ET and higher than >80 for EN). Misclassifications were found only in the case of honey with 10% added sugar syrup. The methods were also able to reveal adulteration of independently predicted samples.

Enhancing Electronic Nose Performance by Feature Selection Using an Improved Grey Wolf Optimization Based Algorithm.

Electronic nose is a kind of widely-used artificial olfactory system for the detection and classification of volatile organic compounds. The high dimensionality of data collected by electronic noses can hinder the process of pattern recognition. Thus, the feature selection is an essential stage in building a robust and accurate model for gas recognition. This paper proposed an improved grey wolf optimizer (GWO) based algorithm for feature selection and applied it on electronic nose data for the first time. Two mechanisms are employed for the proposed algorithm. The first mechanism contains two novel binary transform approaches, which are used for searching feature subset from electronic nose data that maximizing the classification accuracy while minimizing the number of features. The second mechanism is based on the adaptive restart approach, which attempts to further enhance the search capability and stability of the algorithm. The proposed algorithm is compared with five efficient feature selection algorithms on three electronic nose data sets. Three classifiers and multiple assessment indicators are used to evaluate the performance of algorithm. The experimental results show that the proposed algorithm can effectively select the feature subsets that are conducive to gas recognition, which can improve the performance of the electronic nose.

An Electronic Nose Based on Copper Oxide Heterojunctions for Rapid Assessment of Liquor

Abstract The novel methods for rapid detection and discrimination of aroma have gained huge attentions not only because of the growing demands for food safety supervision, but also due to their high potentials in monitoring and regulating fermentation process in food production. In this work, a prototype of electronic nose was developed by eight metal oxide semiconductor gas sensors which were fabricated based on two different types of CuO heterojunction, namely ZnO-CuO (n-p) and NiO-CuO (p-p). The sensing performance of as-fabricated electronic nose towards 7 volatile organic compounds (VOCs) gases samples and 4 real samples of Chinese Jing Wine was conducted. The responses were analyzed with standard statistical methods, e.g., hierarchical clustering analysis (HCA), and were distinguished accurately against one another with no errors or misclassifications.

Domain Transfer Broad Learning System for Long-Term Drift Compensation in Electronic Nose Systems

The long-term sensor drift phenomenon seriously restricts the performance of Electronic Nose (E-nose) systems in their various applications. Due to frequent recalibrations, traditional drift compensation methods are costly and laborious, and their performance are limited due to the nonlinear dynamic properties of the drift. The latest proposed Broad Learning System (BLS) has been confirmed to be an efficient and effective learning technique for many machine learning problems. However, BLS with cross-domain learning capability has rarely been studied. In this paper, a novel unified framework called Domain Transfer Broad Learning System (DTBLS) is proposed based on BLS, to address the issue of drift via adaptive compensation. For the case where there is no labeled target sample, with simultaneous considerations of the empirical loss of source data, marginal distribution adaptation, conditional distribution adaptation and manifold regularization, the DTBLS framework learns a robust target classifier by using labeled source data and unlabeled target data to compensate the drift of sensor response adaptively. To the best of our knowledge, DTBLS is the first BLS-based transfer learning framework for the problem of dataset shift existing in E-nose systems. Like the basic BLS, high computation efficiency is achieved due to the existence of analytical solution. Parameter sensitivity analysis is also conducted to show that the optimal solution can be obtained in a wide range. Experiments on a public gas sensor drift dataset demonstrate that the proposed method outperforms the state-of-the-art methods well.

On Line Monitoring of Odour Nuisance at a Sanitary Landfill for Non-Hazardous Waste

A field study was carried out to assess the possibility of monitoring online malodours from a sanitary landfill for non-hazardous waste. To that aim, the performance of a multisensory electronic nose and a H2S continuous analyser used in situ was evaluated to identify odorous compounds in ambient air emitted from the landfill as perceived at the nearest municipal receptor (the city of Statte, Southern Italy) and compared with the results obtained in laboratory by dynamic olfactometry. Alert situations were conventionally established when two consecutive H2S concentration measurements at 5 min intervals exceeded 20 ppb or when odour emissions measured by the e-nose exceeded 500 OU/m3 for more than 5 min. The study comprised two main activities: 1) training the e-nose to recognize the odorous compounds emitted from the landfill; 2) evaluating the efficiency of the H2S continuous analyser to provide useful information related to biogas emission. Activity 1 demonstrated the e-nose effectiveness for on-line monitoring, quantifying and identifying the odours produced by the landfill. Data processing showed that CDA was the most suitable algorithm to achieve best odours’ pattern recognition. Activity 2 showed that H2S measurements, depending on site-specific factors identified by Kolmogorov-Smirnov test, were poorly correlated with odour concentration in ambient air measured by dynamic olfactometry, as dynamic olfactometry takes into account the effect of other emitted compounds or the masking/enhancing effect of other compounds over H2S. Accordingly, further site-specific investigation is deemed necessary in order to achieve full on-line control of air quality and to adopt proper management measures at sanitary landfills and other odour emitting sources.