Metal Oxide-based Gas Sensor Array for the VOCs Analysis in Complex Mixtures using Machine Learning

Aug 4, 2026

Upload by Patrick

Detection of Volatile Organic Compounds (VOCs) from exhaled breath enables non-invasive early disease identification. The study presents a metal oxide sensor array with three electrodes and applies machine learning to distinguish four VOCs—ethanol, acetone, toluene, and chloroform—within complex mixtures. Training and evaluation use datasets from individual gases and mixed conditions. Multiple algorithms are compared, with KNN and Random Forest achieving over 99% classification accuracy. Regression results show KNN delivering R2 above 0.99 and low LOD values, supporting simultaneous concentration prediction for monitoring and diagnostic use.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages45

Reading Time1 h 53 min

AccessFree to download

Abstract

Introduction

Exhaled breath as a non-invasive source for VOC biomarkers

Variables influencing exhaled breath composition

VOC indicators and disease relevance

Sensor array and machine learning approach (overview)

Metal Oxide-based Gas Sensor Array for the VOCs Analysis in Complex Mixtures using Machine Learning Shivam Singh1, Sajana S1, Poornima2, Gajje Sreelekha3 Chandranath Adak3,*, Rajendra P. Shukla4,*, Vinayak Kamble1,* 1 School of Physics, Indian Institute of Science Education and Research Thiruvananthapuram, 695551 India. 2Dept. ofCSE, Indian Institute of Information Technology Lucknow, Uttar Pradesh 226002, India. 3Dept. ofCSE, Indian Institute of Technology Patna, Bihar 801106, India. 4BIOS Lab-on-a-Chip Group, MESA+ Institute for Nanotechnology, Max Planck Center for Complex Fluid Dynamics, University of Twente, P.O. Box 217, 7500 AE Enschede, The Netherlands. Abstract Detection of Volatile Organic Compounds (VOCs) from the breath is becoming a viable route for the early detection of diseases non-invasively. This paper presents a sensor array with three metal oxide electrodes that can use machine learning methods to identify four distinct VOCsin a mixture. The metal oxide sensor array was subjected to various VOC concentrations, including ethanol, acetone, toluene and chloroform. The dataset obtained from individual gases and their mixtures were analyzed using multiple machine learning algorithms, such as Random Forest (RF), K-Nearest Neighbor (KNN), Decision Tree, Linear Regression, Logistic Regression, Naive Bayes, Linear Discriminant Analysis, Artificial Neural Network, and Support Vector Machine. KNN and RF have shown more than 99% accuracy in classifying different varying chemicals in the gas mixtures. In regression analysis, KNN has delivered the best results with R2 value of more than 0.99 and LOD of 0 .012, 0.015, 0.014 and 0.025 PPM for predicting the concentrations of varying chemicals Acetone, Toluene, Ethanol, and Chloroform, respectively in complex mixtures. Therefore, it is demonstrated that the array utilizing the provided algorithms can classify and predict the concentrations of the four gases simultaneously for disease diagnosis and treatment monitoring. Key words: Gas sensor array, Metal oxide, Volatile Organic Compound, Complex mixture, Machine learning. 1. Introduction Modern technology is becoming even more essential for applications relating to healthcare. Consequently, there is much interest in reducing surgical involvement and enhancing illness early identification. Since it is quicker, less intrusive, and more accessible than a traditional clinical assessment, identifying certain illnesses employing human exhaled air has garnered great interest[1-3] . In this context, exhaled breath is the ideal non-invasive approach since it accurately captures the metabolic processes occurring within the human body[4, 5] . Compared to standard urine or serum tests, disease identification utilizing expiratory VOCs has emerged as the preferable approach for early screening. Besides, it has another excellent relevance for continuous breath monitoring for knowing health anomalies that appear transient or periodic[6] . Breath monitoring has several benefits, the most significant among them being a simple, quick, and straightforward sampling collection method provided by its non-invasive approach[7]. Many (about hundreds) of volatile chemical molecules are found in an individual's breath[4] . Some Volatile Organic Compounds (VOC) chemicals, notably isoprene (heart disease), acetone (diabetes), toluene (lung cancer), nitrogen monoxide (asthma), pentane (heart disease) and ammonia (kidney dysfunction) are established indicators that anticipate underlying disorders[8-10] . However, as shown in Fig. 1(a), several variables affect the constitution of exhaled breath and can be broadly classified as lifestyle-based, health-based and environment based. The usual range for toxicants in a person's exhaled breath is between parts per billion (PPB) to parts per trillion (PPT)[11] . The number of VOCs and their relative proportions are specific to the health of individuals, or unexpected VOCs may be released

FAQs

What is the purpose of the metal oxide sensor array in this work?

It is designed to analyze volatile organic compounds (VOCs) in complex gas mixtures using machine learning for identifying multiple VOCs and predicting their concentrations.

Which VOCs are considered in the sensor experiments?

The work evaluates ethanol, acetone, toluene, and chloroform, both as individual gases and as components of mixtures.

How accurate are the machine learning models for classifying VOCs?

KNN and Random Forest achieve more than 99% accuracy for classifying different chemicals in the gas mixtures.

Aug 4, 2026

Upload by Patrick