An electronic nose in the discrimination of breath from smokers and non-smokers: a model for toxin exposure

Exhaled breath contains hundreds of volatile organic compounds (VOCs) that may be used as non-invasive markers of lung disease. Electronic noses (e-noses) can analyse VOCs by composite nanosensor arrays with learning algorithms. This study investigated the use of an e-nose (Cyranose C320) to distinguish the breath of smokers from that of non-smokers. Smoking and non-smoking subjects exhaled from total lung capacity into a 2 L Tedlar bag and these samples were introduced offline to the e-nose in a random order. Two classes of breath, 'smoker' and 'non-smoker', were established and this model was then cross-validated. Principal component analysis then identified the maximal point of difference between classes. Smellprints of breath from smokers were separated from those of non-smokers (cross-validation value, 95%; Mahalanobis distance, 3.96). Subsequently, 15 smokers (mean age 37.9 +/- 4.78 years, FEV1 3.15 +/- 0.21 L), and 24 non-smokers (add mean age and FEV1 as for smokers) were sampled to revalidate the model. The e-nose correctly identified the smoking status in 37 of the 39 subjects. This demonstrates that the e-nose is simple to use in clinical practice and can differentiate the breath of smokers from that of non-smokers. It may prove to be a useful, non-invasive tool for further breath assessment of exposure to other inhaled noxious substances as well as disease monitoring.