Active Learning on Dynamic Clustering for Drift Compensation in an Electronic Nose System

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.

Electronic Nose Sensor Drift Affects Diagnostic Reliability and Accuracy of Disease-Specific Algorithms

Sensor drift is a well-known disadvantage of electronic nose (eNose) technology and may affect the accuracy of diagnostic algorithms. Correction for this phenomenon is not routinely performed. The aim of this study was to investigate the influence of eNose sensor drift on the development of a disease-specific algorithm in a real-life cohort of inflammatory bowel disease patients (IBD). In this multi-center cohort, patients undergoing colonoscopy collected a fecal sample prior to bowel lavage. Mucosal disease activity was assessed based on endoscopy. Controls underwent colonoscopy for various reasons and had no endoscopic abnormalities. Fecal eNose profiles were measured using Cyranose 320®. Fecal samples of 63 IBD patients and 63 controls were measured on four subsequent days. Sensor data displayed associations with date of measurement, which was reproducible across all samples irrespective of disease state, disease activity state, disease localization and diet of participants. Based on logistic regression, corrections for sensor drift improved accuracy to differentiate between IBD patients and controls based on the significant differences of six sensors (p = 0.004; p < 0.001; p = 0.001; p = 0.028; p < 0.001 and p = 0.005) with an accuracy of 0.68. In this clinical study, short-term sensor drift affected fecal eNose profiles more profoundly than clinical features. These outcomes emphasize the importance of sensor drift correction to improve reliability and repeatability, both within and across eNose studies.

Cross-Domain Active Learning for Electronic Nose Drift Compensation

The problem of drift in the electronic nose (E-nose) is an important factor in the distortion of data. The existing active learning methods do not take into account the misalignment of the data feature distribution between different domains due to drift when selecting samples. For this, we proposed a cross-domain active learning (CDAL) method based on the Hellinger distance (HD) and maximum mean difference (MMD). In this framework, we weighted the HD with the MMD as a criterion for sample selection, which can reflect as much drift information as possible with as few labeled samples as possible. Overall, the CDAL framework has the following advantages: (1) CDAL combines active learning and domain adaptation to better assess the interdomain distribution differences and the amount of information contained in the selected samples. (2) The introduction of a Gaussian kernel function mapping aligns the data distribution between domains as closely as possible. (3) The combination of active learning and domain adaptation can significantly suppress the effects of time drift caused by sensor ageing, thus improving the detection accuracy of the electronic nose system for data collected at different times. The results showed that the proposed CDAL method has a better drift compensation effect compared with several recent methodological frameworks.

Metal oxide semiconducting nanomaterials for air quality gas sensors: operating principles, performance, and synthesis techniques

To meet requirements in air quality monitoring, sensors are required that can measure the concentration of gaseous pollutants at concentrations down to the ppb and ppt levels, while at the same time they exhibiting high sensitivity, selectivity, and short response/recovery times. Among the different sensor types, those employing metal oxide semiconductors (MOSs) offer great promises as they can be manufactured in easy/inexpensive ways, and designed to measure the concentration of  a wide range of target gases. MOS sensors rely on the adsorption of target gas molecules on the surface of the sensing material and the consequent capturing of electrons from the conduction band that in turn affects their conductivity. Despite their simplicity and ease of manufacturing, MOS gas sensors are restricted by high limits of detection (LOD; which are typically in the ppm range) as well as poor sensitivity and selectivity. LOD and sensitivity can in principle be addressed by nanostructuring the MOSs, thereby increasing their porosity and surface-to-volume ratio, whereas selectivity can be tailored through their chemical composition. In this paper we provide a critical review of the available techniques for nanostructuring MOSs using chemiresistive materials, and discuss how these can be used to attribute desired properties to the end gas sensors. We start by describing the operating principles of chemiresistive sensors, and key material properties that define their performance. The main part of the paper focuses on the available methods for synthesizing nanostructured MOSs for use in gas sensors. We close by addressing the current needs and provide perspectives for improving sensor performance in ways that can fulfill requirements for air quality monitoring.

Online Inertial Machine Learning for Sensor Array Long-Term Drift Compensation

The sensor drift problem is objective and inevitable, and drift compensation has essential research significance. For long-term drift, we propose a data preprocessing method, which is different from conventional research methods, and a machine learning framework that supports online self-training and data analysis without additional sensor production costs. The data preprocessing method proposed can effectively solve the problems of sign error, decimal point error, and outliers in data samples. The framework, which we call inertial machine learning, takes advantage of the recent inertia of high classification accuracy to extend the reliability of sensors. We establish a reasonable memory and forgetting mechanism for the framework, and the choice of base classifier is not limited. In this paper, we use a support vector machine as the base classifier and use the gas sensor array drift dataset in the UCI machine learning repository for experiments. By analyzing the experimental results, the classification accuracy is greatly improved, the effective time of the sensor array is extended by 4–10 months, and the time of single response and model adjustment is less than 300 ms, which is well in line with the actual application scenarios. The research ideas and results in this paper have a certain reference value for the research in related fields.

Beverage and Food Fragrance Biotechnology, Novel Applications, Sensory and Sensor Techniques: An Overview

Flavours and fragrances are especially important for the beverage and food industries. Biosynthesis or extraction are the two main ways to obtain these important compounds that have many different chemical structures. Consequently, the search for new compounds is challenging for academic and industrial investigation. This overview aims to present the current state of art of beverage fragrance biotechnology, including recent advances in sensory and sensor methodologies and statistical techniques for data analysis. An overview of all the recent findings in beverage and food fragrance biotechnology, including those obtained from natural sources by extraction processes (natural plants as an important source of flavours) or using enzymatic precursor (hydrolytic enzymes), and those obtained by de novo synthesis (microorganisms' respiration/fermentation of simple substrates such as glucose and sucrose), are reviewed. Recent advances have been made in what concerns "beverage fragrances construction" as also in their application products. Moreover, novel sensory and sensor methodologies, primarily used for fragrances quality evaluation, have been developed, as have statistical techniques for sensory and sensors data treatments, allowing a rapid and objective analysis.