Noninvasive detection of COPD and Lung Cancer through breath analysis using MOS Sensor array based e-nose

Area of Science:

• Biomedical Engineering
• Respiratory Medicine
• Analytical Chemistry

Background:

• Pulmonary diseases, including lung cancer and chronic obstructive pulmonary disease (COPD), pose significant health challenges.
• Current diagnostic methods can be invasive or lack specificity.
• Volatile organic compounds (VOCs) in exhaled breath are potential biomarkers for respiratory diseases.

Purpose of the Study:

• To develop and evaluate a breath-analyzing electronic nose (e-nose) for noninvasive diagnosis of lung cancer and COPD.
• To assess the e-nose's ability to differentiate between patients with lung cancer, COPD, and healthy controls based on breath VOC profiles.

Main Methods:

• Design and implementation of a metal-oxide semiconductor (MOS) sensor array-based e-nose system.
• Collection of breath samples from 40 lung cancer patients, 48 COPD patients, and 90 healthy controls.
• Application of pattern recognition algorithms, including Support Vector Machine (SVM) and k-nearest neighbors (k-NN), for data analysis.

Main Results:

• Support Vector Machine (SVM) achieved 92.3% accuracy in cross-validation for differentiating lung cancer and COPD from controls.
• k-nearest neighbors (k-NN) demonstrated 75.0% accuracy in external validation for the same discrimination task.
• The e-nose system successfully identified variations in VOCs indicative of disease states.

Conclusions:

• Breath analysis using an e-nose system offers a promising noninvasive approach for diagnosing respiratory diseases.
• The developed e-nose technology has significant potential for clinical application in disease screening and diagnosis.
• Further research and validation are warranted to optimize the e-nose for widespread clinical use.