Design and Construction of Electronic Nose for Multi-purpose Applications by Sensor Array Arrangement Using IBGSA

A Hybrid Machine Learning Approach to Screen Optimal Predictors for the Classification of Primary Breast Tumors from Gene Expression Microarray Data

The high dimensionality and sparsity of the microarray gene expression data make it challenging to analyze and screen the optimal subset of genes as predictors of breast cancer (BC). The authors in the present study propose a novel hybrid Feature Selection (FS) sequential framework involving minimum Redundancy-Maximum Relevance (mRMR), a two-tailed unpaired t-test, and meta-heuristics to screen the most optimal set of gene biomarkers as predictors for BC. The proposed framework identified a set of three most optimal gene biomarkers, namely, MAPK 1, APOBEC3B, and ENAH. In addition, the state-of-the-art supervised Machine Learning (ML) algorithms, namely Support Vector Machine (SVM), K-Nearest Neighbors (KNN), Neural Net (NN), Naïve Bayes (NB), Decision Tree (DT), eXtreme Gradient Boosting (XGBoost), and Logistic Regression (LR) were used to test the predictive capability of the selected gene biomarkers and select the most effective breast cancer diagnostic model with higher values of performance matrices. Our study found that the XGBoost-based model was the superior performer with an accuracy of 0.976 ± 0.027, an F1-Score of 0.974 ± 0.030, and an AUC value of 0.961 ± 0.035 when tested on an independent test dataset. The screened gene biomarkers-based classification system efficiently detects primary breast tumors from normal breast samples.

Enose Lab Made with Vacuum Sampling: Quantitative Applications

A lab-made electronic nose (Enose) with vacuum sampling and a sensor array, comprising nine metal oxide semiconductor Figaro gas sensors, was tested for the quantitative analysis of vapor–liquid equilibrium, described by Henry’s law, of aqueous solutions of organic compounds: three alcohols (i.e., methanol, ethanol, and propanol) or three chemical compounds with different functional groups (i.e., acetaldehyde, ethanol, and ethyl acetate). These solutions followed a fractional factorial design to guarantee orthogonal concentrations. Acceptable predictive ridge regression models were obtained for training, with RSEs lower than 7.9, R2 values greater than 0.95, slopes varying between 0.84 and 1.00, and intercept values close to the theoretical value of zero. Similar results were obtained for the test data set: RSEs lower than 8.0, R2 values greater than 0.96, slopes varying between 0.72 and 1.10, and some intercepts equal to the theoretical value of zero. In addition, the total mass of the organic compounds of each aqueous solution could be predicted, pointing out that the sensors measured mainly the global contents of the vapor phases. The satisfactory quantitative results allowed to conclude that the Enose could be a useful tool for the analysis of volatiles from aqueous solutions containing organic compounds for which Henry’s law is applicable.

Effect of Supplementation of Flour with Fruit Fiber on the Volatile Compound Profile in Bread

This paper presents the analyses of the effect of fiber additives on volatile organic compounds in bread. The bread was baked from wheat flour with the addition of 3% of fruit fiber, following common procedures. After baking, volatile organic compounds contained in the control bread and breads supplemented with cranberry, apple, and chokeberry fiber were determined. The SPME/GC-MS technique was used for the identification of the odor profile, and the electronic nose Agrinose (e-nose) was used to assess the intensity of the aroma. The results of the analyses revealed the profile of volatile organic compounds in each experimental variant, which was correlated with responses of the electronic nose. The results indicate that the volatile compound profile depends on the bread additives used and influences the intensity of bread aroma. Moreover, the profile of volatile organic compounds in terms of their amount and type, as well as the intensity of their interaction with the active surface of the electrochemical sensors, was specific exclusively for the additive in each case.

A Machine Learning Method for the Fine-Grained Classification of Green Tea with Geographical Indication Using a MOS-Based Electronic Nose

Chinese green tea is known for its health-functional properties. There are many green tea categories, which have sub-categories with geographical indications (GTSGI). Several high-quality GTSGI planted in specific areas are labeled as famous GTSGI (FGTSGI) and are expensive. However, the subtle differences between the categories complicate the fine-grained classification of the GTSGI. This study proposes a novel framework consisting of a convolutional neural network backbone (CNN backbone) and a support vector machine classifier (SVM classifier), namely, CNN-SVM for the classification of Maofeng green tea categories (six sub-categories) and Maojian green tea categories (six sub-categories) using electronic nose data. A multi-channel input matrix was constructed for the CNN backbone to extract deep features from different sensor signals. An SVM classifier was employed to improve the classification performance due to its high discrimination ability for small sample sizes. The effectiveness of this framework was verified by comparing it with four other machine learning models (SVM, CNN-Shi, CNN-SVM-Shi, and CNN). The proposed framework had the best performance for classifying the GTSGI and identifying the FGTSGI. The high accuracy and strong robustness of the CNN-SVM show its potential for the fine-grained classification of multiple highly similar teas.

Gas Sensors Based on Chemi-Resistive Hybrid Functional Nanomaterials

Chemi-resistive sensors based on hybrid functional materials are promising candidates for gas sensing with high responsivity, good selectivity, fast response/recovery, great stability/repeatability, room-working temperature, low cost, and easy-to-fabricate, for versatile applications. This progress report reviews the advantages and advances of these sensing structures compared with the single constituent, according to five main sensing forms: manipulating/constructing heterojunctions, catalytic reaction, charge transfer, charge carrier transport, molecular binding/sieving, and their combinations. Promises and challenges of the advances of each form are presented and discussed. Critical thinking and ideas regarding the orientation of the development of hybrid material-based gas sensor in the future are discussed.

Design and development of instrumentation for the measurement of sensor array responses

Indigenous instrumentation has been designed and developed for the measurement of the concentration of analytes from eight conductometric metal oxide sensors. The hardware scheme of instrumentation is based on the astable multivibrator configuration. The hardware measures the resistance output from the sensors, conditions, processes, and displays the data on the liquid crystal display. An 8051 based processor averages the data, converts them into engineering units, and sends them to remote PC through ethernet communication for post-data analysis. A graphical user interface (GUI) is developed to acquire, monitor, and display the eight channels' sensor output. GUI plots the online data and offline data as a popup window. The hardware and software of the instrument were tested with standard resistors for calibration and found that in-house developed instrumentation is able to measure with an accuracy of ±0.5% with a resolution of 500 Ω. The instrument has been tested with a semiconductor metal oxide sensor, viz., chromium niobate (CrNbO₄).