E-nose System Research Articles

A novel multi-odour identification by electronic nose using non-parametric modelling-based feature extraction and time-series classification

The electronic nose (e-nose) is an olfaction system that consists of an array of chemical sensors and effective machine learning algorithms for the detection of various target odours. Feature extraction and classification methods are of great importance in improving the performance of the e-nose system. In this paper, a novel odour identification method is presented. Firstly, we use the kernel-based system modelling approach to extract odour features. Its solution is a series of finite impulse responses which containing discriminant information of different odours. In addition, a parameter optimisation method based on normalised mean square error and information entropy is proposed to optimise the kernel function. The entropy is effective in preventing the finite impulse responses from overfitting. Multi-odour classification is achieved based on Gaussian mixture density hidden Markov model (GMM-HMM) considering the characteristic of the extracted features. Also, parameter selection for GMM-HMM is realised according to BIC index and cross-validation. Then, we validate the performance of the proposed feature extraction method in resistance to noise and compare it with other existed features. The modelling-based feature reached the highest performance even without applying any filtering or smoothing techniques. Finally, we compare the proposed combination of feature extraction and classification algorithms with other approaches. The proposed method outperformed other approaches reaching 93.56% in sensitivity and 98.71% in specificity. The results demonstrate that the proposed method is applicable in e-nose-based odour identification.

Electronic-Nose System for Classification of Fruits and Freshness Measurement using K-NN Algorithm

The main objective of this work is to design electronic nose system (E-Nose) which detects the odour and freshness of different fruits such as mango, pineapple, orange which are mainly used in food manufacturing industries. E-Nose system consists of sensor array of two with each sensor respond to different types of odours. These sensor data is analyzed with the K-nearest neighbour algorithm (K-NN Algorithm) using MATLAB for identification of different fruits. Freshness of fruit juice is determined by the measurement of pH value of juice by using pH electrode

Freshness Classification of Horse Mackerels with E-Nose System Using Hybrid Binary Decision Tree Structure

The aim of this study is to test the freshness of horse mackerels by using a low cost electronic nose system composed of eight different metal oxide sensors. The process of freshness evaluation covers a scala of seven different classes corresponding to 1, 3, 5, 7, 9, 11, and 13 storage days. These seven classes are categorized according to six different classifiers in the proposed binary decision tree structure. Classifiers at each particular node of the tree are individually trained with the training dataset. To increase success in determining the level of fish freshness, one of the k-Nearest Neighbors (k-NN), Support Vector Machines (SVM), Linear Discriminant Analysis (LDA) and Bayes methods is selected for every classifier and the feature spaces change in every node. The significance of this study among the others in the literature is that this proposed decision tree structure has never been applied to determine fish freshness before. Because the freshness of fish is observed under actual market storage conditions, the classification is more difficult. The results show that the electronic nose designed with the proposed decision tree structure is able to determine the freshness of horse mackerels with 85.71% accuracy for the test data obtained one year after the training process. Also, the performances of the proposed methods are compared against conventional methods such as Bayes, k-NN, and LDA.

Detection and Discrimination of Volatile Organic Compounds using a Single Film Bulk Acoustic Wave Resonator with Temperature Modulation as a Multiparameter Virtual Sensor Array.

This paper describes the detection and discrimination of volatile organic compounds (VOCs) using an e-nose system based on a multiparameter virtual sensor array (VSA), which consists of a single-chip temperature-compensated film bulk acoustic wave resonator (TC-FBAR) coated with 20-bilayer self-assembled poly(sodium 4-styrenesulfonate)/poly(diallyldimethylammonium chloride) thin films. The high-frequency and microscale FBAR multiparameter VSA was realized by temperature modulation, which can greatly reduce the cost and complexity compared to those of a traditional e-nose system and can allow it to operate at different temperatures. The discrimination effect depends on the synergy of temperature modulation and the sensing material. For proof-of-concept validation purposes, the TC-FBAR was exposed to six different VOC vapors at six different gas partial pressures by real-time VOC static detection and dynamic detection. The resulting frequency shifts and impedance responses were measured at different temperatures and evaluated using principal component analysis and linear discriminant analysis, which revealed that all analytes can be distinguished and classified with more than 97% accuracy. To the best of our knowledge, this report is the first on an FBAR multiparameter VSA based on temperature modulation, and the proposed novel VSA shows great potential as a compact and promising e-nose system integrated in commercial electronic products.

A comparison of online and offline measurement of exhaled breath for diabetes pre-screening by graphene-based sensor; from powder processing to clinical monitoring prototype

Several breath analysis studies have suggested a correlation between blood glucose (BG) levels and breath acetone, indicating acetone as a primary biomarker in exhaled breath for diabetes diagnosis. Herein, we have (i) fabricated and validated graphene-based chemi-resistive sensors for selective and sensitive detection of acetone, (ii) performed offline breath analysis by a static gas sensing set-up to acquire olfactory signals, and (iii) developed an LED-based portable on/off binary e-nose system for pre-screening diabetes through online analysis. The fabricated sensors showed selective detection for acetone with high sensitivity (5.66 for 1 ppm acetone vapor) and fast response and recovery times (10 s and 12 s) at low concentrations. The sensor responses of end tidal fractional breath (collected in Tedlar bags) in the fasting and postprandial conditions were compared with BG levels and glycated hemoglobin (HbA1c) levels taken at the same time in 30 volunteers (13 healthy and 17 diabetic subjects). The mean sensor responses of the diabetic subjects as obtained by offline analysis were 1.1 times higher than those of the healthy subjects. The optimal regression equation framed with the significant correlating variables for HbA1c estimation achieved an accuracy of 66.67%. The online breath analysis by on/off binary prototype exhibited an accuracy of 60.51%. Though there exists a minimal uncertainty in classification, the on/off type portable prototype is easy to operate, gives a quicker response with a refresh/recovery rate of 19 s and can be used for preliminary diagnosis, and can be used for preliminary diagnosis. This inexpensive sensor technology may revolutionize personalized medicine in the near future and greatly benefit the underprivileged.

Discrimination of Chinese Liquors Based on Electronic Nose and Fuzzy Discriminant Principal Component Analysis.

The detection of liquor quality is an important process in the liquor industry, and the quality of Chinese liquors is partly determined by the aromas of the liquors. The electronic nose (e-nose) refers to an artificial olfactory technology. The e-nose system can quickly detect different types of Chinese liquors according to their aromas. In this study, an e-nose system was designed to identify six types of Chinese liquors, and a novel feature extraction algorithm, called fuzzy discriminant principal component analysis (FDPCA), was developed for feature extraction from e-nose signals by combining discriminant principal component analysis (DPCA) and fuzzy set theory. In addition, principal component analysis (PCA), DPCA, K-nearest neighbor (KNN) classifier, leave-one-out (LOO) strategy and k-fold cross-validation (k = 5, 10, 20, 25) were employed in the e-nose system. The maximum classification accuracy of feature extraction for Chinese liquors was 98.378% using FDPCA, showing this algorithm to be extremely effective. The experimental results indicate that an e-nose system coupled with FDPCA is a feasible method for classifying Chinese liquors.

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.

Breath Analysis for Detecting Diseases on Respiratory, Metabolic and Digestive System

Recently, biological technology and computer science are of great importance in medical applications. Since one’s breath biomarkers have been proved to be related with diseases, it is possible to detect diseases by analysis of breath samples captured by e-noses. In this paper, a novel medical e-nose system specific to disease diagnosis was used to collect a large-scale breath dataset. Methods for signal processing, feature extracting as well as feature & sensor selection were discussed for detecting diseases on respiratory, metabolic and digestive system. Sequential forward selection is used to select the best combination of sensors and features. The experimental results showed that the proposed system was able to well distinguish healthy samples and samples with different diseases. The results also showed the most significant sensors and features for different tasks, which meets the relationship between diseases and breath biomarkers. By selecting best combination of different sensors and features for different tasks, the e-nose system is shown to be helpful and effective for diseases diagnosis on respiratory, metabolic and digestive system.

An Olfactory Sensor Array for Predicting Chemical Odor Characteristics from Mass Spectra with Deep Learning.

Machine learning techniques are useful for applications such as electronic nose (e-nose) systems to classify or identify the target odor. In recent years, deep learning is regarded as one of the most powerful machine learning methods. It enables researchers to extract useful features automatically from high-dimensional raw data and has been widely applied to computer vision, speech recognition, and natural language processing, though little has been reported in the field of olfaction. In this chapter, we describe the procedure to build a deep neural network to predict odor characteristics of chemicals from their mass spectra.

Pollutant Gases Detection using the Machine learning on Benchmark Research Datasets

In real time environment, different varieties of sensor devices are deployed to collect and transmit the periodic environment data to the base station to monitor the specific tasks. The pollutant gases like methanol, LPG, ammonia are harmful to human beings, hence such vulnerabilities should automatically be detected and safety alarm generated in a particular area. Such systems are often called as Electronic Nose (E-nose) systems which are an automated system that analyze continues periodic data and detect any harmful situations based on various approaches such as threshold-based or machine learning based. In this Paper, we have evaluated the performance metrics Sensitivity, Specificity, and Accuracy for GSAD dataset and Air Quality datasets.

Effects of Different Fermentation Strains on the Flavor Characteristics of Fermented Soybean Curd.

The effects of different fermentation strains on the flavor characteristics of fermented soybean curd (FSC) were investigated in this study. Fresh tofu was fermented by Actinomucor elegans, Rhizopus arrhizus, Mucor racemosus, and Rhizopus chinensis, either alone or in various combinations. The FSC manufacturing process included prefermentation by different strains at 28 °C for 60 hr, followed by salting at 16 °C for 7 days and finally proceeding postfermentation at 25 °C for 35 days. Subsequently, five tested samples were obtained, namely, sample A (fermented by A. elegans alone), R (fermented by R. arrhizus alone), AR (fermented by A. elegans and R. arrhizus at 5:1), AM (fermented by A. elegans and M. racemosus at 1:1), and RR (fermented by R. arrhizus and R. chinensis at 7:3). The flavors of the five samples were determined by E-nose, sensory evaluation, and GC-MS. E-nose system observed significant discriminations by principal component analysis and linear discriminant analysis analysis. Sensory evaluation ranked the overall sensory scores: AR>AM>A>RR>R. As shown in GC-MS results, sample AR also had, on average, the highest level of many volatiles. Out of 10 critical volatiles, the detected frequency of samples AR, AM, RR, A, and R was 10, 9, 9, 8, and 7, respectively. PLS2 regression model was used to explore the influence on flavor quality of different strains. All three analytic methods revealed similar results, with sample AR providing the best flavor quality, while the opposite was the case with sample R. Therefore, it could be concluded that A. elegans and R. arrhizus at 5:1 (v/v) was the optimal combination, and may likely promote the production of critical volatile compounds. PRACTICAL APPLICATION: The flavors of fermented soybean curds are influenced by various factors such as physicochemical and microorganism during the fermentation surroundings. The results of this work not only provide valuable information for FSC flavor studies, but can also guide the FSC industry to improve flavor quality by applying the most appropriate production strains.

Use of Electronic Noses in Seawater Quality Monitoring: A Systematic Review.

Electronic nose (eNose) systems are particularly appreciated for their portability, usability, relative low cost, and real-time or near real-time response. Their application finds space in several domains, including environmental monitoring. Within this field, marine monitoring is of particular scientific relevance due to the fragility of this specific environment, daily threatened by human activities that can potentially bring to catastrophic and irreversible consequences on marine wildlife. Under such considerations, a systematic review, complying with the PRISMA guidelines, was conducted covering the period up to 15 October 2018, in PubMed, ScienceDirect, and Google Scholar. Despite the relatively low number of articles published on this specific topic and the heterogeneity of the technological approaches employed, the results obtained by the various groups highlight the positive contribution eNose has given and can provide in near future for the monitoring and safeguarding of this delicate environment.

A Set of Platforms with Combinatorial and High-Throughput Technique for Gas Sensing, from Material to Device and to System.

In a new E-nose development, the sensor array needs to be optimized to have enough sensitivity and selectivity for gas/odor classification in the application. The development process includes the preparation of gas sensitive materials, gas sensor fabrication, array optimization, sensor array package and E-nose system integration, which would take a long time to complete. A set of platforms including a gas sensing film parallel synthesis platform, high-throughput gas sensing unmanned testing platform and a handheld wireless E-nose system were presented in this paper to improve the efficiency of a new E-nose development. Inkjet printing was used to parallel synthesize sensor libraries (400 sensors can be prepared each time). For gas sensor selection and array optimization, a high-throughput unmanned testing platform was designed and fabricated for gas sensing measurements of more than 1000 materials synchronously. The structures of a handheld wireless E-nose system with low power were presented in detail. Using the proposed hardware platforms, a new E-nose development might only take one week.

Gas-Sensor Drift Counteraction with Adaptive Active Learning for an Electronic Nose.

Gas sensors are the key components of an electronic nose (E-nose) in violated odour analysis. Gas-sensor drift is a kind of physical change on a sensor surface once an E-nose works. The perturbation of gas-sensor responses caused by drift would deteriorate the performance of the E-nose system over time. In this study, we intend to explore a suitable approach to deal with the drift effect in an online situation. Considering that the conventional drift calibration is difficult to implement online, we use active learning (AL) to provide reliable labels for online instances. Common AL learning methods tend to select and label instances with low confidence or massive information. Although this action clarifies the ambiguity near the classification boundary, it is inadequate under the influence of gas-sensor drift. We still need the samples away from the classification plane to represent drift variations comprehensively in the entire data space. Thus, a novel drift counteraction method named AL on adaptive confidence rule (AL-ACR) is proposed to deal with online drift data dynamically. By contrast with conventional AL methods selecting instances near the classification boundary of a certain category, AL-ACR collects instances distributed evenly in different categories. This action implements on an adjustable rule according to the outputs of classifiers. Compared with other reference methods, we adopt two drift databases of E-noses to evaluate the performance of the proposed method. The experimental results indicate that the AL-ACR reaches higher accuracy than references on two E-nose databases, respectively. Furthermore, the impact of the labelling number is discussed to show the trend of performance for the AL-type methods. Additionally, we define the labelling efficiency index (LEI) to assess the contribution of certain labelling numerically. According to the results of LEI, we believe AL-ACR can achieve the best effect with the lowest cost among the AL-type methods in this work.

On the feasibility of metal oxide gas sensor based electronic nose software modification to characterize rice ageing during storage

The aroma change of aromatic and non-aromatic rice was traced during storage using electronic nose (E-nose). Various steady-state and transient features were derived from the adsorption and desorption phases of the E-nose responses to characterize the ageing process and classify the storage durations. Principal component analysis was utilized to analyse the ageing in terms of seven classes related to the storage duration of 6 months. The aromatic samples followed a specific pathway in its timeline with separately distinctive grouping, revealing the reduction of ageing indices. The aromatic rice underwent crucial changes in its volatile compounds, mainly in the early stages of storage. While, confused grouping of the non-aromatic rice proved its stability due to less diverse range of the aroma compounds. Several artificial neural networks, namely back propagation (BP), radial basis function (RBF), and learning vector quantization (LVQ) were used to classify the storage durations. For the aromatic samples, full classification was achieved by using all the networks. For the non-aromatic rice, the developed RBF and LVQ networks represented full classification by using the features of rise time and polynomial curve fitting parameters. It is concluded that the E-nose system along with the developing methods could be utilized to control the rice ageing.

Fire Resistance Detection of Flammable Materials Based on Electronic Nose Technology

The electronic nose system can identify and detect the toxic gases released by flammable materials during combustion. This paper studies the application of the e-nose technology in the early fire detection of flammable materials. Based on the brief introduction of the e-nose system and the associated pattern recognition technology, this paper builds a high-throughput flammable material fire simulation platform, and then, with the PVC material of the wires and cables in the electrical equipment which are prone to fire as an example, it detects the volatile substances within the safe operating temperature range (100-180°C) and at the warning temperature point (200°C) during the combustion process. In order to verify the anti-interference ability of the platform, this paper selects liquor and cigarette as interferences, which are also subject to detection in the experiment. It uses discriminant function analysis (DFA) and the BP neural network method to perform statistical analysis of the collected data and the results show that both methods have good discriminant effects. At the same time, it also optimizes the sensor array by the load analysis method. Through comparison and analysis, it is found that the eight-sensor array has a better discriminant effect. The research results show that the electronic nose technology can realize the early detection of flammable materials.

Development of a Dual MOS Electronic Nose/Camera System for Improving Fruit Ripeness Classification.

Electronic nose (E-nose) systems have become popular in food and fruit quality evaluation because of their rapid and repeatable availability and robustness. In this paper, we propose an E-nose system that has potential as a non-destructive system for monitoring variation in the volatile organic compounds produced by fruit during the maturing process. In addition to the E-nose system, we also propose a camera system to monitor the peel color of fruit as another feature for identification. By incorporating E-nose and camera systems together, we propose a non-destructive solution for fruit maturity monitoring. The dual E-nose/camera system presents the best Fisher class separability measure and shows a perfect classification of the four maturity stages of a banana: Unripe, half-ripe, fully ripe, and overripe.

Domain Correction Based on Kernel Transformation for Drift Compensation in the E-Nose System.

This paper proposes a way for drift compensation in electronic noses (e-nose) that often suffers from uncertain and unpredictable sensor drift. Traditional machine learning methods for odor recognition require consistent data distribution, which makes the model trained with previous data less generalized. In the actual application scenario, the data collected previously and the data collected later may have different data distributions due to the sensor drift. If the dataset without sensor drift is treated as a source domain and the dataset with sensor drift as a target domain, a domain correction based on kernel transformation (DCKT) method is proposed to compensate the sensor drift. The proposed method makes the distribution consistency of two domains greatly improved through mapping to a high-dimensional reproducing kernel space and reducing the domain distance. A public benchmark sensor drift dataset is used to verify the effectiveness and efficiency of the proposed DCKT method. The experimental result shows that the proposed method yields the highest average accuracies compared to other considered methods.

Development of Novel Electronic Nose Applied for Strawberry Freshness Detection during Storage

Abstract A newly self-developed electronic nose (E-nose) system for the detection of “Hongyan” strawberry freshness at different storage periods was studied. The system consisted of six metal oxide semiconductor sensors connected to a data acquisition system and a computer with pattern recognition software. The aroma emitted by “Hongyan” strawberry samples was detected during post-harvesting storage, and stable E-nose response values were used to develop cluster analysis and classification models. The successive projections algorithm was employed to optimize the sensors array, and the results obtained by gas chromatography–mass spectrometry analysis proved that the optimized sensor array was feasible to differentiate decayed strawberries from fresh ones. Partial least squares discriminant analysis and support vector machine (SVM) models were built. Accuracy of 94.9 % on the testing set was obtained based on the optimized sensor array, and this result was satisfactory compared to that of commercial PEN3 E-nose.

Lab-made electronic-nose with polyaniline sensor array used in classification of different aromas in gummy candies

Aroma is closely related to the food product acceptability and an important product quality indicator. Electronic-nose (E-nose) systems are an interesting alternative to traditional methods of aroma analyses. A lab-made E-nose system equipped with an array of sensing units comprised by gold interdigitated microelectrodes (IDEs) using polyaniline (Pani) as sensitive layers deposited by the in situ and Layer-by-layer (LbL) methods was used to analyze aromas in gummy candies. Different concentrations from artificial aromas (apple, strawberry and grape), added to the gummy candies were evaluated. Our system presented 21.6 mV.ppb−1 sensitivity, ppb range detection limit, and good reversibility, around 97.6%. The sensitive layers of Pani films was adequate deposited on IDEs observed by the Attenuated Total Reflection/Fourier-transform infrared spectroscopy (ATR/FTIR). Linear Discriminant Analysis (LDA) was able to classify apple, strawberry, and grape aromas added to gummy candies using saturation potential values from the E-nose system, demonstrating its applicability in food matrices.