A portable electronic nose as an expert system for aroma-based classification of saffron Sajad Kiani a , Saeid Minaei a, ⁎, Mahdi Ghasemi-Varnamkhasti b a Biosystems Engineering Department, Tarbiat Modares University, Tehran, Iran b Department of Mechanical Engineering of Biosystems, Shahrekord University, Shahrekord, Iran abstract article info Article history: Received 29 January 2016 Received in revised form 17 April 2016 Accepted 18 May 2016 Available online 20 May 2016 This study focuses on the development and evaluation of a portable electronic nose (e-nose) system for identifi- cation of different types of saffron, stigma of Crocus sativus L. (Iridaceae), based on their Volatile Organic Com- pounds (VOCs). The system utilizes metal oxide semiconductor gas sensors and direct head space sampling. Real-time data acquisition system, microcontroller devices and a laptop computer along with multivariate com- putational tools were used for development of an expert system. Eleven saffron samples from different regions were prepared for the experiments. Principal Component Analysis (PCA) and Hierarchical Clustering Analysis (HCA) as unsupervised models and Multilayer Perceptron (MLP) neural networks and Partial Least Squares (PLS) as supervised models were utilized to develop the e-nose discrimination capability. Based on the results, PCA of volatile compounds fingerprints revealed eleven distinct groups corresponding to the eleven different saf- fron samples. This was further confirmed by HCA which classified the groups into five distinct Quality Classes (QCs) (excellent, very good, good, medium, and poor quality) which were used as the MLP and PLS classification goals. Results of analysis showed that performance of the MLP model for prediction of saffron samples QC was better than the PLS model, with 100% success rate and high correlation coefficients of cross validation (R 2 = 0.989 and relatively low RMSE value of 0.0141). These results show that the developed system is capable of dis- criminating saffron samples based on their aroma and can be utilized as an aroma quality control system. © 2016 Published by Elsevier B.V. Keywords: Artificial senses Electronic nose MOS Saffron volatile compounds Aroma quality 1. Introduction Dried flowers of saffron, Crocus sativus L., are highly valued for their flavor, color and health-promoting properties [42]. With respect to saf- fron chemistry, the three main characterizing constituents of saffron which are the compounds responsible for its attributes are Crocins (C 44 H 64 O 24 ), Picrocrocin (C 16 H 26 O 7 ), and Safranal (C 10 H 14 O) [25,29,32, 39,47]. Product quality depends on the concentration of these major metabolites responsible for the unique color, taste and aroma, respec- tively [33,49]. Since conventional quality-analysis techniques such as chromatography and spectroscopy are not suitable for on-line quality control of food products [27], new “green” methods such as e-nose, electronic tongue (e-tongue), electronic eye (e-eye) and their fusion for in-process and real-time assessment of these products have been proposed [6,11,12,13,51,53]. E-nose is an instrument designed to imi- tate the sense of smell and discriminate among complex odors by means of an array of gas sensors (which respond to gases and vapors generated by the sample) and multivariate data analysis methods. More details on e-nose systems are given in Bhattacharyya and Bandhopadhyay [8], Ghasemi-Varnamkhasti et al. [18,19],and Alcaniz et al. [1,28]. There are several studies reporting the use of e-nose for quality assessment of medicinal and aromatic plants such as coffee [15,16,36,37,43], Tea [7,48,52], and white pepper [31]. Other studies re- port on the identification of aroma compounds in cocoa powder [26], in- vestigation of changes in aroma of ginsengs [14], discrimination between ten different species of Chinese herbal medicines [30], quality control of Lonicera japonica during several months of storage [50] and characterization of the VOCs of Jasmine flowers grown in India [40]. To date, limited research has been conducted concerning the applica- tion of e-nose for saffron volatiles analysis. An e-nose based on Metal Oxide Semiconductor (MOS) gas sensors was first used by Carmona et al. [10] to discriminate between volatile profiles of saffron samples native to four countries, namely, Iran, Morocco, Greece and Spain. Heidarbeigi et al. [24] described an e-nose technique to differentiate non-adulterated and adulterated saffron (more than 10% impurities). Aroma is a prominent quality factor in saffron grading. The higher the saffron aroma, the higher the saffron quality and price. Thus, the main aim of this research was to develop and evaluate an expert system for the identification of Iranian saffron with different aromas based on their VOCs and geographical origin. Iran is the most prominent saffron producing country in the world. Detailed objectives of this work were Chemometrics and Intelligent Laboratory Systems 156 (2016) 148–156 ⁎ Corresponding author. E-mail address: minaee@modares.ac.ir (S. Minaei). http://dx.doi.org/10.1016/j.chemolab.2016.05.013 0169-7439/© 2016 Published by Elsevier B.V. Contents lists available at ScienceDirect Chemometrics and Intelligent Laboratory Systems journal homepage: www.elsevier.com/locate/chemolab