Oregano authentication by mid-level data fusion of chemical fingerprint signatures acquired by ambient mass spectrometry

Abstract

Economically motivated adulteration (EMA) of herbs and spices is very frequent and a major cause of concern to consumers, manufacturers and legislators. Rapid, non-targeted methods capable of distinguishing authentic herbs from those mixed with low cost materials are desirable. To this aim, ambient mass spectrometry (AMS) was applied to authenticate dried oregano leaves. Authentic and adulterated oregano samples were submitted to two extraction procedures and analysed in positive and negative ion modes by direct analysis in real time-high resolution mass spectrometry (DART-HRMS). Mid-level data fusion of the four blocks of DART-HRMS data was performed, and the resultant unique dataset was submitted to supervised statistical analysis to ascertain the signals that discriminate authentic from adulterated oregano. The fourteen most informative signals of authenticity were tentatively assigned and validated by support vector machine (SVM). In the final model, accuracy, sensitivity and specificity of >90% were obtained. Independent authentic and adulterated oregano samples were used to validate the discriminative m/z values and evaluate the classification capability of the model. To the best of our knowledge this is the first application of AMS coupled to mid-level data fusion for oregano authentication.