Journal of Food Engineering 144 (2015) 103–111

Contents lists available at ScienceDirect

Journal of Food Engineering

journal homepage: www.elsevier.com/locate/jfoodeng

Review

Electronic noses for food quality: A review

Amy Loutfi a, Silvia Coradeschi a, Ganesh Kumar Mani b, Prabakaran Shankar b,
John Bosco Balaguru Rayappan b,⇑
a
Center for Applied Autonomous Sensors Systems, Fakultetsgatan 1, Orebro, Sweden
b
Centre for Nanotechnology & Advanced Biomaterials (CeNTAB), School of Electrical & Electronics Engineering (SEEE), SASTRA University, Thanjavur 613 401, India

a r t i c l e i n f o a b s t r a c t

Article history: This paper provides a review of the most recent works in electronic noses used in the food industry. Focus
Received 21 March 2014 is placed on the applications within food quality monitoring that is, meat, milk, fish, tea, coffee and wines.
Received in revised form 24 July 2014 This paper demonstrates that there is a strong commonality between the different application area in
Accepted 25 July 2014
terms of the sensors used and the data processing algorithms applied. Further, this paper provides a
Available online 4 August 2014
critical outlook on the developments needed in this field for transitioning from research platforms to
The authors dedicate this review article to industrial instruments applied in real contexts.
Prof. Silvia Coradeschi who sadly passed Ó 2014 Elsevier Ltd. All rights reserved.
away in February 2014. Prof. Silvia was
known for her unique leadership quality
and dedication with unparalleled energy
and enthusiasm. Her absence has made an
unmatchable vacuum in our team. It is a
great loss to us all and she is greatly missed.

Keywords:
Food quality
e-Nose
Metal oxide
Pattern recognition

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
2. Electronic noses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
3. Application areas in food . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
3.1. Milk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
3.2. Wine. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
3.3. Tea and coffee . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
3.4. Fish and meat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
4. Common methodologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
4.1. Electronic nose comparison with sensor panels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
4.2. Hybrid electronic nose technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
4.3. New pattern recognition methods applied to food analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
4.4. Toward new sensor materials for electronic nose instruments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108

Abbreviations: ANN, artificial neural network; APLSR, analysis of variance partial least squares regression; BP-MLP, back propagation multilayer perceptron; BPNN,
back-propagation neural network; CA, cluster analysis; DA, discriminant analysis; DBN, deep belief network; DFA, discriminant factorial analysis; FNN, fuzzy neural network;
LDA, linear discriminant analysis; LVQ, learning vector quantization; MGLH, multivariate general linear hypothesis; MLP, multi-layer perceptron; MOS, metal oxide
semiconductor; MOSFET, metal oxide semiconductor field effect transistor; m-TDNN, multiple-time-delay neural networks; PCA, principal component analysis; PLS, partial
least squares regression; PNN, probabilistic neural network; QDA, quadratic discriminant analysis; RBF, radial basis function; SQC, statistical quality control; SVM, support
vector machines; VOCs, volatile organic compounds; KAMINA, Karlsruhe Micro. Nose.
⇑ Corresponding author. Tel.: +91 4362 264 101x2255, mobile: +91 9944468389; fax: +91 4362 264120.
E-mail address: [email protected] (J.B.B. Rayappan).

http://dx.doi.org/10.1016/j.jfoodeng.2014.07.019
0260-8774/Ó 2014 Elsevier Ltd. All rights reserved.
104 A. Loutfi et al. / Journal of Food Engineering 144 (2015) 103–111

5. Discussion and future perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108

Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110

1. Introduction A broad list of e-nose reviews can be found in the literature that
are structured and focused on mass spectrometry based electronic
Foodborne illnesses cause about 76 million cases of illnesses, noses (Peris and Escuder-Gilabert, 2009), biomedical and health
325,000 hospitalizations, and 5000 deaths in the United States care applications (Wilson and Baietto, 2011), agriculture and
each year (Hedberg, 1999). Common symptoms of foodborne forestry applications (Wilson, 2013), microbial quality control of
illness include diarrhea, nausea, abdominal cramps, headache, food products (Falasconi et al., 2012), pharmaceutical applications
dizziness and fever. In the developed/developing countries, sur- (Alam et al., 2012), for developing chemical sensor arrays (James
veillance of foodborne disease is a fundamental component of food et al., 2005). Our review is distinctive in a way that we focus on
safety systems (WHO). According to the estimates of US Centers for the methodologies which are common across the various applica-
Disease Control and Prevention (CDC) in the year 2011, roughly 1 tions. Therefore the motivation of this review is to make accessible
out of 6 Americans or 48 million people get sick, 128,000 are hos- to the various research groups not only the progress within their
pitalized and 3000 die of foodborne illnesses (CDC, 2011). Hence specific area but also in adjacent areas and most importantly com-
researchers started exploring better way for quality discrimination mon methodologies that can be used to solve related challenges in
of perishable foods. different fields. A secondary motivation is to promote the use of
Highly perishable, muscle foods like fish, meat and poultry have electronic noses for food quality monitoring fin an industrial set-
become an integral part of human diet over many decades. How- ting by summarizing the extensive work in the past few years that
ever, in the past two decades awareness about the food safety from indicates promising results with respect to the applicability of
the point of specific pathogenic bacteria has exemplified the e-noses to a vast number of areas.
requirement for a rapid and accurate detection system for micro-
bial spoilage in fish and meat (Frost, 2001; Haugen et al., 2006).
In general fish and meat quality will be assessed either by examin- 2. Electronic noses
ing the structure (texture, tenderness, flavor, juiciness, and color)
or by detecting the microorganism and its count or by detecting During the 1980s research on machine olfaction lead to a gen-
the gas/VOCs generated by these microorganisms. erally accepted definition of an electronic nose as an instrument
The practical application of human nose as a smell assessment that comprises an array of heterogeneous electrochemical gas sen-
instrument is severely limited by the fact that our sense of smell sors with partial specificity and a pattern recognition system
is subjective, gets tired easily, and is therefore difficult to use. Con- (Gardner and Bartlett, 1999; Persaud and Dodd, 1982). However,
sequently, there is considerable need for an instrument that could in more recent years, the term electronic nose has been used in a
mimic the human sense of smell and its use in routine industrial broader sense to refer to gas sensors that measure the ambient
applications. To promote this technology to industrial application, gas atmosphere based on the general principle that changes in
metal oxide gas/odour sensors became exemplary candidates in the gaseous atmosphere alter the sensor properties in a character-
areas like food industry, environment control, automobile industry, istic way. A variety of different sensor types have been developed,
indoor air quality check and monitoring, industrial production, to which three types of materials are commonly used: metal oxi-
medicine and in safety aspects, to name a few – Scientific groups des, conducting polymers composites and intrinsically conducting
worldwide are investigating them giving due importance to the polymers. Apart from conductive sensors, gas detection has also
various aspects of gas/odour sensing properties. been done using optical sensors, surface acoustic wave sensors,
Electronic nose instruments are attractive for a number of sig- gas sensitive field effect transistors and quartz microbalance
nificant features: the relatively fast assessment of headspace, a (QMB) sensors. Micro-electro-mechanical systems (MEMS) plus
quantitative representation or signature of a gas and cheap sensors nanotechnologies are the most promising emerging technologies
which can be easily integrated in current production processes. in the area. The term electronic nose has also been used to charac-
Despite these features, there are still relatively few applications terize systems where ultra-fast gas chromatography or mass spec-
of electronic noses adopted in industry. This could be attributed trometry is employed in the detection process. Once the data from
to difficulties in robustness, selectivity and reproducibility of the the individual sensors from the array is collected, the electronic
sensors and to the need for pattern recognition algorithms which nose systems require a suitable post processing procedure to ana-
can cope with the complex signal analysis. Nonetheless, the use lyze and classify the data. Pre-processing of multivariate signals in
of electronic noses is rapidly expanding and there have been sensor arrays represents an essential part of the measuring system.
notable achievements relevant for the food industry, particularly Data processing techniques used in post processing of pattern
in the past few years. Furthermore, this progress coincides with recognition routines include principal component analysis (PCA),
an increased understanding of the biological mechanisms behind linear discriminate analysis (LDA), partial least squares (PLS), func-
the human olfactory system. Specifically, we now have a greater tional discriminate analysis (FDA), cluster analysis (CA), fuzzy logic
understanding not only of the genetics behind the olfactory recep- or artificial neural network (ANN) such as probabilistic neural net-
tors but also of the relationships between an odorant’s molecular work (PNN). Among these techniques, PCA, PLS, LDA, FDA and CA
property and the quality of an odor. This paper focuses on the lat- are based on a linear approach while fuzzy logic, ANN and PNN
est developments within key areas related to foodstuff where a are regarded as nonlinear methods (Scott et al., 2006).
quantitative approach to quality estimation is important as it reg- Particular to the food industry is the sample handling system
ulates the economy of food (e.g. pricing) and quality control (e.g. used for exposing the volatile compounds present in the headspace
detection of bacteria and spoilage). Specifically, this paper reviews (HS) to the sensor array in the e-nose. For some applications, spe-
the progress in the past decade for the following areas: milk, wine, cific techniques are used such as purge & trap (P&T), dynamic
coffee, tea, fish and meat. headspace (DHS), solid-phase microextraction (SPME) used by
 A. Loutfi et al. / Journal of Food Engineering 144 (2015) 103–111 105

Table 1
A list of some of the common commercially available electronic nose models.

Model No. of sensors Technology Manufacturer Country

i-Pen, PEN2, PEN3 6 MOS Sensor Airsense Analytics Germany
QCS 2 MOS Sensor Gerstel GmbH & Co. KG Germany
Artinose 38 MOS Sensor Sysca AG Germany
FF2 6 MOS Sensor RST Rostock System-Technik GmbH Germany
GFD1 6 MOS Sensor
FOX 2000, 3000 & 4000 6, 12 & 18 MOS Sensor Alpha MOS France
Promethus 18 MOS Sensor
Air quality module 2 MOS Sensor Applied Sensor Sweden
EOS 835, Ambiente 6 MOS Sensor Sacmi Italy
Bloodhound ST 214 14 Conducting polymers Scensive Technologies UK
Aromascan A32S 32 Conducting polymers Osmetech Plc USA
Cyranose 320 32 Conducting polymers Intelligent Optical Systems, Inc. USA

Berna et al., 2009 to preprocess wine samples. Comparisons quality, due to undesirable aroma, physical defects and metabolic
between methods is a further subject of study such as in, where toxicity. Hence identification of food spoilage, before the product
two types of purging, trapping methods and four types solid- formation became mandatory to avoid consumption of such spoiled
phase-micro extraction methods were compared (Lozano et al., food and this can be satisfied with the assistance of electronic
2008a,b). Aside from the sample handling, the sample itself may noses.
be preprocessed such as in the analysis of the quality of wine Likewise, spoilage of milk due to various other factors should
and beer, the preprocessing procedure of dehydration and dealco- also be identified at the earliest possible stage to avoid complica-
holization helped the e-nose to classify aromas to a better extent tions and complaints in the final product. Microbial milk spoilage
(Ragazzo-Sanchez et al., 2009). In the analysis of tea, a novel sam- can be identified from the presence of acetaldehyde, 3-methyl-1-
pling system has been devised based on illumination-controlled butanol, acetic acid and ethanol in headspace of the milk
heating together with physical raking (Bhattacharya et al., 2008c, (Ampuero and Bosset, 2003; Magan et al., 2001). In this work
2008d). authors determined the specified markers employing 14 conduct-
Today, several electronic noses were commercially available in ing polymer sensors and discriminant function analysis to separate
the market. These e-noses are often promoted as generic devices the unspoiled and bacterial containing milk. Whereas Marsili et al.
and suitable for a range of applications. Table 1 provides a list of studied the off-flavors in milk using solid-phase microextraction,
some of the common commercially available electronic nose mass spectrometry and multivariate analysis (SPME–MS–MVA)
models. and found the increase in hexanal and dimethyl disulfide concen-
tration in milk when it was exposed to the fluorescent light (typical
3. Application areas in food exposure in supermarkets). Also suggested that pentanal, hexanal,
heptanal and isopentanal can indicate the copper induced oxida-
The applications of electronic noses have been numerous and tion in milk (Marsili, 1999). Ageing of milk can be significantly
range from environmental monitoring (Kashwan and Bhuyan, identified through the presence of dimethyl sulfide, ethyl acetate,
2005) to medical applications (Längkvist and Loutfi, 2011). Since 2-heptanone, pentanal, etc. Also he observed (Marsili, 2000) that
1993, the amount of publications in the area of electronic olfaction the headspace of milk from the cow bearing genetic defect contains
is more than 12,000 articles. The main application areas related to trimethylamine. Authors (Eriksson et al., 2005) have observed
the food industry have been: fish, meat, milk, wine, coffee and tea, higher levels of sulfides, ketones, amines and acids in milk col-
and constitute approximately 5000 publications since 1993. This lected from the cow affected by mastic disease. For this experi-
demonstrates that food applications are central to electronic olfac- ment, a commercial hybrid sensor array system (Applied Sensor
tion and nearly half of the publications are in this area. Within each 3320, Linköping, Sweden) consisting of 10 MOSFET sensors, 12
application in the food industry, research contributions have MOS sensors, one sensor for humidity were used to collect the data
focused on the detection of a variety of aspects such as freshness, from milk and PCA was employed to analyze the milk data.
adulteration, off-flavors and bacteria detection. We provide in this Among these, the dominating components were ethanol, tri-
section a short description of the main application areas for a num- methylamine, 2, 3-butanedione, 2-pentanone, 2-methylbutanal,
ber of specific food stuff. methane-thiol, dimethyl sulfide, acetic acid and in general, the
infected milk had higher CO2 content than the healthy reference
milk. Therefore by detecting the higher concentration of CO2, mas-
3.1. Milk
tic milk can be identified.
Headspace of milk from a healthy cow typically contains com-
plex mixtures of VOCs (acetone, hexanal, 2-pentanone, 2-butanone, 3.2. Wine
toluene, limonene, heptanal, etc.) at various concentrations
(Ampuero and Bosset, 2003; Pardo and Sberveglieri, 2005). The con- The main application of electronic noses with respect of wine is
centration of these VOCs varies due to various factors such as bac- quality assurance. Wine is a commercial product, which can vary
terial metabolism, ageing, photo-oxidation and presence of pro- greatly in aroma and flavor according to the large possible varia-
oxidant metals such as copper, iron and nickel. Milk from a healthy tions in its production. The formation and transformation of
cow contains only few bacteria, which may multiply and the rate of organic acids at must fermentation and wine production are of
multiplication will increase, as time passes from the time of milking great importance in wine-making. Biochemical processes caused
to the time of processing; it also depends on the standard of milk- by yeast enzymes are significant to achieve better quality of wine.
ing, handling practices and storage. These bacterial growth leads Organic acids also protect wine against bacterial diseases. How-
to the spoilage of milk with the production of off-flavor. As the ever, high content of some acids influences wine flavor badly.
microbial spoilage of starting milk severely affects the industrial The qualitative and quantitative analyses of the most dominant
106 A. Loutfi et al. / Journal of Food Engineering 144 (2015) 103–111

aromas and flavors evolve from the headspace of wine namely Differential signal processing with SVM classifier (Brudzewski et al.,
Dimethylamine (DMA), Trimethylamine (TMA), Ethyl octanoate, 2012) schemes were used to classify the data for decision making.
1-Hexanol, Ethanethiol, Ethylacetate, and 2,4,6-Trichloroanisole The discrimination of aromas has also been fine-tuned by altering
play a significant role in determining the quality of wine (Arroyo the sensor materials (Wang et al., 2012) and data classification sys-
et al., 2009; Roy and Basu, 2004). Authors (Arroyo et al., 2009) have tems (Shi et al., 2012) which are employed in the electronic nose.
assessed the quality of wine using an electronic nose comprising of
16 tin oxide thin film based sensors. To enhance the sensitivity of
3.4. Fish and meat
these sensing elements, chromium and indium elements doped
thin films with varying thickness ranging from 200 to 800 nm were
In the recent past, inspection of food quality and safety especially
utilized. PCA, PNN and Leave One Out (LOO) algorithms were
to examine the sea foods because they are highly nutritious to pro-
applied for sensor data classification. Electronic nose techniques
vide better health care compounds. Fatty acids, amino acids, pro-
have been used to detect defects or spoilage e.g. caused by high
teins, fat content, flavor, color and texture of the fish depends
concentrations of 4-ethylphenol and 4-ethylguaiacol (Berna et al.,
upon the farming systems, control of growth rate, water chemistry,
2008; Cynkar et al., 2007). Other applications include the discrim-
water temperature, feeding behavior, nutrition intake and freshness
ination between aging techniques (Lozano et al., 2008a, 2008b;
(before and after the harvest). Among them freshness of the fish is
Prieto et al., 2012), discrimination between products (based on
the major issue in the fish market industry. Spectroscopy and imag-
geographic origin or grape) (Berna et al., 2009) and the prediction
ing techniques were used to test the freshness level of fish flesh. But
of sensorial descriptions (Buratti et al., 2007).
these techniques are having their own advantages and limitations
(Herrero, 2008). Sensory analysis is considered to be the best tool
3.3. Tea and coffee
to determine the fish freshness in the industry and consumer mar-
ket (Warm et al., 2001). Seafood export industry has taken several
The infusions and extract of the leaves Camellia sinensis (L.) is
steps to prevent the decomposition of fish flesh quality by reducing
the source for the beverage tea. The extract of leaves has the chem-
the microbial spoilage factor. Fatty acids profiles, aldehydes
ical components such as flavanol, caffeine, phenolic substances,
(hexanal, 2-methyl-1-butanal, nonanal, 2,4-heptadienal), ketones
fats, amino acids, theaflavin, thearubigin and volatile components
(2-octanone, 2-decanone, 2-propanone), trimethyl amine (TMA),
(Chaturvedula and Prakash, 2011). These chemical components
volatile organic compounds (1-butanol, 1-penten-3-ol) are the
are the source to determine the flavors and aromas of the beverage,
major aromatic compounds identified as the biomarkers of spoilage
which are depending on the leaves collected from plants and man-
levels. Sensing the concentration levels of these biomarkers
ufacturing processes like withering, pre-conditioning, cut-tear-curl
(Catarina Bastos and Magan, 2006; Macagnano et al., 2005;
operation, fermentation and drying (Bhattacharya et al., 2008c). All
Olafsdottir et al., 2004) using Electronic nose is found to be one of
these processes referred to focus on the enzymatic oxidation of tea
the best and promising solutions to analyze the quality of food
leaves. Thus degree of fermentation process dominantly deter-
(Gholamhosseini et al., 2007). The freshness levels of selected vari-
mined by the enzymatic oxidation of tea leaves which produces
eties of fish species namely Red Snapper, Gurnard, Tarakihi, and
specific aromas and flavors of tea. The pioneering work reported
Trevally using an array of thirty-two polymer carbon black compos-
by Dutta et al. employed Warwick metal oxide electronic nose
ite sensors in the portable e-nose (Cyranose 320, USA) with ANN
(WOLF – Enose, UK) with radial basis function (RBF) network pat-
classifiers have been reported (Gholam Hosseini et al., 2008). A
tern recognition system and successfully discriminated the flavors
fairly accurate assessment confirmed that this e-nose can be
of tea manufactured in different conditions (Dutta et al., 2003).
employed to estimate the fish freshness, rate of degradation and
Researchers (Bhattacharya et al., 2008c) have also calibrated an
quantitative time period for quality sustenance.
e-nose with an array of eight metal oxide sensors (TGS, Fiagaro
Similarly, the application of electronic nose technology to meats
Inc., Japan) toward volatile emission signatures which are evolving
is one of the main application areas in the food industry. A signif-
during the fermentation process. This work has resulted in the
icant number of works have been presented over the past two dec-
e-nose which can be used to predict the optimum fermentation
ades and (Ghasemi-Varnamkhasti et al., 2009) gives an additional
time in real time applications.
review of the field. The main applications of electronic noses with
Similarly, Coffee also provides various flavors depending on spe-
respect to meat are in assessing quality, spoilage identification,
cies like Arabica, Robusta, Bengal, Congo and etc., and variety of
detection of off-flavors, taints and classification of bacterial strains.
plant, other agricultural factors, harvesting, picking, sorting, pro-
In addition, processing techniques and discrimination of different
cessing, hulling, polishing, roasting, and packaging. In addition, cof-
types of meats have been examined (Blixt and Borch, 1999;
fee is changing in all its forms, from green to roast to brew.
Hansen et al., 2002; Olsen et al., 2005). The recognition of meat
Traditionally, human olfaction technique was used to discriminate
freshness as well as decomposition of meat food under various sit-
the quality of coffee beans aroma and flavors. In general, presence
uations like temperature and volume were successfully tested
of species like Arabica has been considered as one of the markers
using an e-nose (KAMINA e-nose, Germany) with an array of 38
indicating the high quality coffee bean due to its pleasing flavors,
sensors in which LDA algorithm was used to classify the observed
aromatic properties, low in caffeine and low in acidity. Since coffee
data (Musatov et al., 2010).
has gained significant commercial and economic importance, peo-
ple adulterate it by mixing certain low quality or even other species
of coffee beans to yield more profit. But, it is very difficult to dis- 4. Common methodologies
criminate the quality of adulterated coffee using human sensory
panels and gas measurement system (Kottawa-Arachchi et al., 4.1. Electronic nose comparison with sensor panels
2012). In this scenario, Electronic noses have been designed to
quantify the concentration levels of the identified aromas in coffee Comparative studies between sensory panels and the response
and its quality levels (Bhattacharya et al., 2008a; Dutta et al., 2003). from an electronic nose dates back to the early 1990s. In the liter-
An e-nose comprising of 12 preheated metal oxide gas sensors (Fig- ature there has been a twofold motivation for developing correla-
aro Inc., Japan) were used to recognize the flavors and aromas by tions with human panels. On one hand, the pattern recognition
detecting the variations of volatile organic components in the mix- process, whose output is highly dependent on the given labels, is
tures of low quality Robusta beans and high quality Arabica beans. synchronized with current descriptors used in the food industry.
 A. Loutfi et al. / Journal of Food Engineering 144 (2015) 103–111 107

The emerging challenge is to correlate the rich descriptors used by insensitive to androsterone (Bonneau and Squires, 2004), making
humans to the response of the electronic nose e.g., by using PLS. the selection of human panels particularly challenging.
Other approaches have attempted to provide human like descrip- In sum, sensor panels are an essential aspect in the industrial
tors to an electronic nose response using natural language symbols food processes, but the difficulty and cost to maintain such panels
which have been manipulated to provide new descriptions to is high. Consequently, for day-to-day production having a tool such
unseen samples via relation to known concepts e.g. ’’smells a lot as an electronic nose which would in effect be synchronized with
like lavender’’ (Loutfi et al., 2001). A secondary motivation for cor- an eventual panel would be valuable for industrial use. The use
relation with human panels is to encourage the adaptation of new of an electronic nose to complement human panels particularly if
standards and eventually exploit the e-nose’s ability to provide ‘‘non-odorant’’ gases are to be detected is promising. Nevertheless
finer grading of attributes e.g. freshness. In determination of fish the general practice is to use the sensor panel to define the targets
freshness, the most promising works has been done in the project for quality detection and train the nose on these descriptors. In
FAIR CT98-4076 (MUSTEC) by Olafsdottir et al. (2004) by consider- order for a symbiosis between electronic noses and human panels
ing various combinations of physico-chemical techniques and its to emerge in the industry, a high reproducibility remains to be
integration to attain output that relates fish freshness. The project demonstrated in the electronic counterpart.
aims at producing a Artificial Quality Index (AQI) combining the
outputs of the instrumental techniques and calibrating them with 4.2. Hybrid electronic nose technologies
sensory scores of Quality Index Method (QIM) used by test panel
for attributes like appearance, smell and texture. While e-noses present a number of advantages over traditional
The works which are most prevalent in sensor comparison with analysis, the sensors also present a number of shortcomings which
panels are related to wine quality testing. This is rather expected have yet to be solved. These include issues such as sensor poison-
due to the rich and varied qualifiers often used in the wine indus- ing, sensor drift and sensitivity. Selectivity and sensor drift have
try. Arroyo et al. investigated a homemade electronic nose with been the focus of investigation. Specific to the food industry is
semiconducting thin film based sensing elements and twenty-five profile masking (e.g. ethanol) which affects the response. Profile
human tasters were trained to classify seventeen different aromas masking for example can be found in wines and often requires a
which are used to discriminate the quality of wine. It was found pre-processing step as described in Section 3.2. Recent trends to
that human tasters performed better in identifying certain aromas overcome sensor shortcomings include combining semiconductor
(Arroyo et al., 2009). However, an electronic nose subsequently chemical sensors with other types of gas sensors. While this
developed by Santos et al. was found to be better in detecting complicates the sampling system (requiring more bulk and elec-
the specific thresholds of typical red wine compounds such as tronics), this hybrid technology is able to compensate for the short-
ethyl acetate and eugenol and white wine compounds such as hex- comings in current chemical sensor technology. In particular, a
anol and ethyl octanoate (up to 8 times lower than a human panel) new generation of electronic noses referred to as mass spectrome-
(Santos et al., 2010). In another work, the success rate of sixteen tin try-based electronic noses (MS-E-nose) are increasingly used in the
oxide sensors based e-nose system was compared to a human sen- literature where a MS-E-nose consists of a mass spectrometer
sory panel and to a HP-6890 gas chromatograph with HP Mass instrument without prior chromatographic separation.
Selective 5973 Spectrometer. Twenty-eight wine samples were Comparisons of MS-based e-nose and other commercial e-nose
analyzed. From this comparison it was found that the results pro- systems have been presented in several works. The work by Berna
vided by e-nose employing PLS regression algorithm corresponded et al. has compared the performance of MOS-e-nose (FOX 3000
better to the sensory panel results than to the predictions of clas- E-Nose from Alpha MOS, France) and MS-E-nose in estimating
sification form the GC–MS. the defects in red wine (Berna et al., 2008). These two techniques
Tea, like wine is also characterized by rich descriptors and elec- were used to identify the presence of the two major components
tronic noses can be trained to achieve a high correlation with tea of taint in red wine namely 4-ethylphenol and 4-ethylguaiacol
tasters. For example, researchers (Tudu et al., 2009a) used various (Cynkar et al., 2007) used MS-e-nose for identifying spoilage factor
methods to correlate with the tester’s scores for assessing the qual- due to the presence of 4-ethylphenol and a success rate of 67% was
ity of black tea samples. Typically, testers assign scores to tea on a achieved with Stepwise Linear Discriminant Analysis (SLDA) classi-
scale of 1 to 10 based on flavor, aroma and taste of the sample. In fier. A combination of twelve metal oxide sensors (MOS-e-nose)
this work, a set of Figaro gas sensors were first selected for black and a mass spectrometry-based e-nose (MS-E-nose) was used
tea classification by measuring the sensitivity to various com- (Ragazzo-Sanchez et al., 2009). The preprocessing procedure of
pounds that are typical in black tea aroma (Geraniol, Linalool, dehydration and dealcoholization helped the e-nose to do better
2-phenyl-ethanol). In total, an array of five sensors were then used classification of aromas (Cozzolino et al., 2008). This approach
in an instrument whose responses were correlated with two has coupled the mass spectrometry and e-nose for better estima-
groups of tea testers – thereby deriving a computational model tion of aroma properties in Australian Riesling wines. After collect-
to predict the score for unknown samples. Testing with unknown ing the data using MS-e-nose, it was classified and analyzed using
samples resulted in a classification rate of more than 90%. PCA and partial least squares (PLS1) regression adopting leave one
In certain applications such as meat quality testing, the detec- out method. Even though this technique did not help for quantita-
tion of specific bacteria or chemical compounds is the primary tive discrimination of various aroma elements of selected com-
aim, and thus correlation with human panels can be limited. Nev- pounds, it helped for better screening of wines prior to sensory
ertheless, certain off-odors such as boar taint, a unpleasant cooking analysis.
odor from non-castrated male pigs, are often characterized by Similarly, Berna (Berna et al., 2008) have compared the perfor-
human panels. Five different meat parts (loin, neck, shoulder, outer mance of MOS-Enose (FOX 3000 E-Nose from Alpha MOS, France)
and inner thigh) were tested with an electronic nose and validated and MS-Enose in estimating the defects in red wine. These two
by a trained human sensory panel (Kirsching, 2012). A high deter- techniques were used to identify the presence of the two major
mination coefficient (R2 = 0.915) was obtained between reference components of taint in red wine namely 4-ethylphenol and
values of boar taint (obtained by sensory panel) and predicted val- 4-ethylguaiacol. MS-E-nose was used prior to ethanol removal
ues calculated from e-nose data. Two compounds deposited in the from the sample and MOS-Enose later. As expected the selectivity
fat tissue of pig are held responsible for boar taint: androstenone of MOS-Enose was not helpful for rapid discrimination of quality of
and skatole. Typically in any sample population about 23% are red wine. In an electronic nose, ion mobility spectrometry was
108 A. Loutfi et al. / Journal of Food Engineering 144 (2015) 103–111

used for boar taint measurements where samples varying in nanostructured metal oxide thin films based sensors are playing
androstenone and skatole levels were tested (Vestergaard et al., a key role in the field of perishable food quality assessment partic-
2006). Their findings indicated that sensory perceptible boar taint ularly meat and fish (Galdikas et al., 2000; Jung and Lee, 2011;
was found to be more related to androstenone than to skatole. Spe- Mani and Rayappan, 2013, 2014a, 2014b; Muniyandi et al., 2014;
cifically the e-nose was indentified as a useful tool in for ordering Panigrahi et al., 2006; Perera et al., 2010). Researcher Roy and Basu
samples with respect to low and high levels of androsterone and modified the surface of nanostructured zinc oxide thin film by dis-
skatole. The e-nose based on ion mobility spectrometry was partic- persing palladium and doping aluminum (Roy and Basu, 2004). The
ularly useful for fast analysis and identified as suitable approach authors reported that the surface modified ZnO film with palla-
for application in production e.g. at the slaughter line. dium exhibited a better sensitivity to DMA, whereas Al doped
In sum, electronic noses based on chemical gas sensors still ZnO showed better sensitivity toward TMA over DMA. Since TMA
present the advantage of ease of use and low costs. Nevertheless, and DMA are the very good indicators of fish freshness level, the
the more complex MS-nose could be used as a complement to pro- developed ZnO films can be used as one of the sensors in the sensor
vide better verification of results based on an initial screening from array. SnO2 and CuO mixed oxide nanowires exhibited a better sen-
a more portable and lightweight device. sitivity than that of pure tin oxide nanowires toward acetone, eth-
anol and ethyl acetate which are all the good markers of meat, fish
4.3. New pattern recognition methods applied to food analysis freshness level (Li et al., 2011). The works by Kumar et al. have
reported that the nanostructured cerium oxide thin film is highly
The pattern recognition component of the electronic nose is selective for one of the markers in perishable food items TMA com-
non-trivial due to: the non-linearity of the sensor response; the pared with the ethanol (Kumar et al., 2013) .
need to compensate for sensor drift; and the selectivity which Clearly for electronic nose applications, the selection of the
requires that an electronic nose is trained to recognize specific pat- materials used in the array is important to achieve good discrimi-
terns representing odors. The machine learning community has nation for the intended application. However, solving the problems
used electronic nose data as a way to validate new and generic related to the selectivity in gas sensing applications still remains a
algorithms that cope with the aforementioned challenges. In par- major challenge. In this context, the different sensing behavior of F
ticular with respect of machine learning applied to food related and Mn doped ZnO thin films toward ethanol and TMA is encour-
applications, a number of aspects have been considered that aging and will help to solve the problem of selectivity (Sivalingam
include, feature extraction, sensor selection and incremental learn- et al., 2011, 2012). Though F/Mn doped films are n-type semicon-
ing. In (Bhattacharyya et al., 2007) a selection of specific sensors for ductors and ethanol/TMA are reducing gases, the presence of grain
a customized e-nose for tea is presented. In this, (Bhattacharya boundary scattering during sensing in the first case and the
et al., 2008b) an incremental probabilistic neural network is used absence of the same in the second case helped to develop sensing
for black tea grade discrimination. The incremental classifiers elements with inherent selective nature. Adding such sensing ele-
show the ability to accommodate new classes and new knowledge ments to the array of sensors, electronic nose will provide new
within an already trained model, thus promoting the possibility to possibilities for detection of specific markers.
dynamically augment the training data. This is similar to the works
done by Tudu et al., in which a radial basis function (RBF) is com- 5. Discussion and future perspectives
bined with incremental learning techniques that preserve the
knowledge already learnt by the network on previous training sets In this paper, we have outlined the major contributions of elec-
while allowing the acquisition of new knowledge as it becomes tronic nose technologies relevant to the most published fields
available (Tudu et al., 2009a). Other combinations of incremental within the food industry (Table 2). As each paper addresses just
learning have been shown by the same authors using fuzzy tech- one application area, the field gives the impression of being frag-
niques (Tudu et al., 2009b). Other learning methods that have been mented. This is in general due to the need to tune either the soft-
applied also include non-linear support vector machines in the ware and/or hardware to the specific application. However, much
area of spoilage detection in meat. For example, using non-linear overlap exists in hardware and software algorithms used. It is also
support vector machines, a high classification rate of 98.81% and clear that the utility of using e-noses in an industrial context is
96.43% were obtained in identifying spoiled/unspoiled samples of high, and that most works have in fact shown cooperation with
beef and sheep meat respectively (El Barbri et al., 2008). A trend industrial partners that have made available the samples and con-
in machine learning that can also have relevant application in elec- ditions of use. The open question is therefore why there is an
tronic olfaction is the use of deep learning. A first attempt has been absence of electronic noses in industrial processes. It is likely that
applied to the classification of meat spoilage markers where the the underlying reason for the reluctance of the uptake of e-noses in
combination of novel sensing materials and auto encoders were an industrial context is multifold. On the materials side, major
shown to provide an electronic nose system with high selectivity focus must be given to the design and development of drift free
to TMA as well as fast response when using the transient response sensors that can be used reliably over long temporal horizons. It
for detection (Längkvist et al., 2013). For real industrial deploy- is likely that only experiments showing long term use will be a
ments, it will be necessary to have systems that can be easily main- convincing factor for industry when considering the uptake of such
tained, and do not require an expert in machine learning to a device. Consequently, the internal drift influencing factors like
manually tune the learning parameters for optimal performance. crystalline structure variation, grain size variations, grain bound-
This is the main challenge for the machine learning community ary effects, uniformity in the dopant concentrations, perfect con-
in developing algorithms for food related applications. tact materials, thickness of the sensing elements should be
addressed. In addition external factors like humidity, ambient tem-
4.4. Toward new sensor materials for electronic nose instruments perature and atmosphere, presence of other gases/aromas during
standardization procedure, accuracy in analog to digital data con-
An important new trend is the developing of nanostructured version process, etc. must be given special attention for enhancing
sensors for electronic nose instruments. This new kind of sensors the performance of the sensor system as a whole. In this context,
have promising features as they offer controllable grain size nanostructured surfaces may be used for better selectivity and
(Yamazoe et al., 2003). This can open an entirely new era in sensitivity and Metal organic framework may be incorporated for
development of innovative metal oxide gas sensors. Recently, filtering the atmospheric effects and cross interferences and hence
 A. Loutfi et al. / Journal of Food Engineering 144 (2015) 103–111 109

Table 2
Summary of the applications of e-nose in different food matrices.

Application Sensing E-Nose Model Number Data Processing Ref.

Element of Algorithm
sensors
Rancidity of two different kinds of MOS (SnO2) Lab Made 5 PCA Capone et al. (2001)
milk (ultrahigh temperature
processed and pasteurized)
To detect unspoiled and spoiled Conducting Bloodhound Sensors Ltd. UK (BH-114) 14 DFA Magan et al. (2001)
milk polymer
To discriminate different milk MOS Alpha MOS, France (Fox 4000) 18 PCA Wang et al. (2010)
flavorings
Identification of adulterated milk MOS Airsense Analytics, Germany (PEN 2) 10 PCA, LDA Yu et al. (2007)
Identification of odours (milk, MOS Figaro Engineering Inc., Japan (TGS 2620, 2610 5 Neuro Solutions 5.0 Botre et al. (2009)
rancid milk and yoghurt) and 2600) from Neuro
Dimension Inc., USA
Differentiate the milk collected MOSFET + MOS Applied Sensor, Sweden (3320) 22 PCA Eriksson et al. (2005)
from healthy and mastitic
disease affected cows
Seasonal changes in milk powder Conducting OSMETECH PLC, England 28 LDA Biolatto et al. (2007)
polymer
Bacterial contamination in milk QCM SES Piezo Ltd., England 6 PCA Ali et al. (2003)
Detect different off-flavors in wines MOS Alpha MOS, France (Fox 4000) 18 PCA, DFA Ragazzo-Sanchez et al.
of different origin (2009)
Prediction of Italian red wine MOS Airsense Analytics, Germany (PEN 2) 10 Genetic Algorithms Buratti et al. (2007)
Classification of four types of red MOS (SnO2) Lab Made 16 PCA Lozano et al. (2008a,b)
wines
Detection of aroma from green tea MOS Airsense Analytics, Germany (PEN 2) 10 PCA, LDA and BPNN Yu et al. (2009)
at different storage times
Identification of optimum MOS Figaro Engineering Inc., Japan (TGS-832, 823, 8 m-TDNN Bhattacharya et al.
fermentation time for black tea 831, 816, 2600, 2610, 2611 and 2620) (2008c)
Quality evaluation of black tea MOS Figaro Engineering Inc., Japan (TGS 832, 823, 5 Incremental fuzzy Tudu et al. (2009b)
2600, 2610, and 2611), classification
Quality evaluation of black tea MOS Figaro Engineering Inc., Japan (TGS 832, 823, 5 Incremental RBF Tudu et al. (2009a)
2600, 2610 and 2611) neural network
Classification of black tea MOS Figaro Engineering Inc., Japan (TGS 816, 823, 8 PCA, PNN, BP-MLP, Bhattacharya et al.
831, 832, 2600, 2610, 2611 and 2620) RBF (2008a)
Discriminating different grades of MOS Airsense Analytics, Germany (PEN 2) 10 PCA, LDA Yu et al. (2008)
green teas
Distinguish the Japanese green teas MOS Fragrance & Flavor Analyzer, Shimadzu, Japan 10 PCA, CA Yang et al. (2009)
with different content of (FF-2A)
coumarin
Discrimination of tea with different MOS Figaro Engineering Inc., Japan (TGS 880, 826, 4 MLP, LVQ, PNN, RBF Dutta et al. (2003)
qualities 825 and 822)
Classification of different brands of MOS (SnO2) Lab Made 4 PCA, MLP Pardo et al. (2000)
Espresso coffee
Classification of different coffee MOS Figaro Engineering Inc., Japan (TGS 800, 815, 14 ANN, FNN Singh et al. (1996)
816, 821, 823, 824, 825, 830, 831, 842, 880, 881,
882, 883)
Analysis of Colombian coffee MOS Figaro Engineering Inc., Japan and FIS Inc., Japan 8 PCA, MLP Rodríguez et al. (2009)
samples (SP-12A, SP 31, TGS 813, TGS 842, SP-AQ3, TGS
823, ST-31 and TGS 800)
Shelf-life determination of red meat MOS Figaro Engineering Inc., Japan (TGS 815, 821, 6 PCA Amari et al. (2009)
822, 824, 825 and 842)
Discrimination of boar tainted MOS Alpha MOS France (Fox 4000) 18 MGLH, DA Kirsching (2012)
samples of different meat parts
Quality detection of porcine meat MOS Danish Odour Sensor System (DOSS) 6 APLSR Hansen et al. (2002)
loaf
Detection of boar taint MOS Figaro Engineering Inc., Japan (TGS 825, 882, 5 – Bourrounet et al.
824, 822, 800) (1995)
Spoilage classification of red meat MOS Figaro Engineering Inc., Japan (TGS 823, 825, 6 PCA, PLS, SVM El Barbri et al. (2008)
826, 831, 832, 882)
Meat spoilage markers detection MOS (ZnO) Lab Made 3 PCA, SVM, DBN Längkvist et al. (2013)
Discrimination of microbial MOS Figaro Engineering Inc., Japan (TGS 812, 822, 6 LDA, QDA Panigrahi et al. (2006)
population 880, 2602, 2611 and 2611)
Quality identification of rapid MOS Alpha MOS, France (Fox 4000) 18 PCA, SQC Yang et al. (2009)
fermented fish

improving selectivity and sensitivity. On the software side, of such markers has been done by building relevant models which
researchers apply many of the available linear and nonlinear correspond to GC–MS studies, till today, there is no significant and
algorithms (either separately or combined). While complex data accurate mathematical modeling proposed. It is also possible to
processing can achieve good results on a specific dataset it fails address the problem of selectivity via changing properties in the
to compensate for the core problems of stability and reliability of sensing array, for example, synchronization or modulation of the
the sensors. Clearly, selectivity is also a key issue and it is impor- operating temperatures are viable approaches to improve selectiv-
tant that the relevant markers for the given application are in fact ity. Ultimately, however, such methods still must demonstrate a
detected and in certain cases quantifiable. So far, the examination robustness over time to be considered for industrial use. On the
110 A. Loutfi et al. / Journal of Food Engineering 144 (2015) 103–111

market a selection of generic instruments are available as shown in Capone, S., Epifani, M., Quaranta, F., Siciliano, P., Taurino, A., Vasanelli, L., 2001.
Monitoring of rancidity of milk by means of an electronic nose and a dynamic
Table 1, it is quite possible that an indication of a maturing of the
PCA analysis. Sensors Actuators B Chem. 78, 174–179.
field will be demonstrated when specific instruments are available Catarina Bastos, A., Magan, N., 2006. Potential of an electronic nose for the early
for the specific applications and when such instruments are being detection and differentiation between Streptomyces in potable water. Sensors
used over a long time period in real settings. In addition, the top Actuators B Chem. 116, 151–155.
Chaturvedula, V., Prakash, I., 2011. The aroma, taste, color and bioactive
five needs in this area of research are (1) drift free sensors that constituents of tea. J. Med. Plants Res. 5, 2110–2124.
can be used reliably (2) investigation of new material for achieving Cozzolino, D., Smyth, H.E., Cynkar, W., Janik, L., Dambergs, R.G., Gishen, M., 2008.
better selectivity (3) better modeling and correlation between Use of direct headspace-mass spectrometry coupled with chemometrics to
predict aroma properties in Australian Riesling wine. Anal. Chim. Acta 621,
presence of chemical markers and the sensor response in the e- 2–7.
nose array (4) application of specific instruments with carefully Cynkar, W., Cozzolino, D., Dambergs, B., Janik, L., Gishen, M., 2007. Feasibility study
selected sensor arrays and sampling system (5) Better understand- on the use of a head space mass spectrometry electronic nose to monitor red
wine spoilage induced by Brettanomyces yeast. Sensors Actuators B Chem. 124,
ing of the industrial need related to quality control and monitoring 167–171.
of food processing. Dutta, R., Hines, E.L., Gardner, J.W., Kashwan, K.R., Bhuyan, M., 2003. Tea quality
prediction using a tin oxide-based electronic nose: an artificial intelligence
approach. Sensors Actuators B Chem. 94, 228–237.
Acknowledgements El Barbri, N., Llobet, E., El Bari, N., Correig, X., Bouchikhi, B., 2008. Electronic nose
based on metal oxide semiconductor sensors as an alternative technique for the
The authors wish to express their sincere thanks to the Depart- spoilage classification of red meat. Sensors 8, 142–156.
Eriksson, Å., Waller, K.P., Svennersten-Sjaunja, K., Haugen, J.-E., Lundby, F., Lind, O.,
ment of Science & Technology, New Delhi, India and VINNOVA, Eriksson, A., 2005. Detection of mastitic milk using a gas-sensor array system
Sweden for their financial support (Project ID: INT/SWD/VINN/ (electronic nose). Int. Dairy J. 15, 1193–1201.
P-04/2011). We also thank SASTRA University, Thanjavur, India Falasconi, M., Concina, I., Gobbi, E., Sberveglieri, V., Pulvirenti, A., Sberveglieri, G.,
2012. Electronic nose for microbiological quality control of food products. Int. J.
and Orebro University, Orebro, Sweden for extending the infra- Electrochem. 2012, 1–12.
structural support to carry out this work. Frost, J.A., 2001. Current epidemiological issues in human campylobacteriosis. J.
Appl. Microbiol. 90, 85S–95S.
Galdikas, A., Mironas, A., Senulien, D., Strazdien, V., Šetkus, A., Zelenin, D., 2000.
References Response time based output of metal oxide gas sensors applied to evaluation of
meat freshness with neural signal analysis. Sensors Actuators B Chem. 69, 258–
Alam, H., Saeed, S.H., Engg, C., 2012. Electronic nose in food and healh applications: 265.
a review. Int. J. Comput. Corp. Res. 2, 1–17. Gardner, J., Bartlett, P., 1999. Electronic Noses, Principles and Applications. Oxford
Ali, Z., O’Hare, W., Theaker, B., 2003. Detection of bacterial contaminated milk by University Press, New York, NY, USA.
means of a quartz crystal microbalance based electronic nose. J. Therm. Anal. Ghasemi-Varnamkhasti, M., Mohtasebi, S.S., Siadat, M., Balasubramanian, S., 2009.
Calorim. 71, 155–161. Meat quality assessment by electronic nose (machine olfaction technology).
Amari, A., El Barbri, N., El Bari, N., Llobet, E., Correig, X., Bouchikhi, B., Pardo, M., Sensors 9, 6058–6083.
Sberveglieri, G., 2009. Potential application of the electronic nose for shelf-life Gholam Hosseini, H., Luo, D., Xu, G., Liu, H., Benjamin, D., 2008. Intelligent fish
determination of raw milk and red meat. AIP Conf. Proc. 1137, 457–460. freshness assessment. J. Sensors 628585, 8 pages.
Ampuero, S., Bosset, J.O., 2003. The electronic nose applied to dairy products: a Gholamhosseini, H., Luo, D., Liu, H., Xu, G., 2007. Intelligent Processing of E-nose
review. Sensors Actuators B Chem. 94, 1–12. Information for Fish Freshness Assessment. In: Intelligent Sensors, Sensor
Arroyo, T., Lozano, J., Cabellos, J.M., Gil-Diaz, M., Santos, J.P., Horrillo, C., 2009. Networks and Information, 2007. ISSNIP 2007. 3rd International Conference on.
Evaluation of wine aromatic compounds by a sensory human panel and an pp. 173–177.
electronic sose. J. Agric. Food Chem. 57, 11543–11549. Hansen, T., Boholt, K., Gammelgaard, E., Byrne, D.V., 2002. Sensory and electronic
Berna, A.Z., Trowell, S., Clifford, D., Cynkar, W., Cozzolino, D., 2009. Geographical nose analysis of porcine meat loaf in relation to ingredient quality. In: Sensors,
origin of Sauvignon Blanc wines predicted by mass spectrometry and metal 2002. Proceedings of IEEE, vol. 1. pp. 736–740.
oxide based electronic nose. Anal. Chim. Acta 648, 146–152. Haugen, J.E., Chanie, E., Westad, F., Jonsdottir, R., Bazzo, S., Labreche, S., Marcq, P.,
Berna, A.Z., Trowell, S., Cynkar, W., Cozzolino, D., 2008. Comparison of metal oxide Lundby, F., Olafsdottir, G., 2006. Rapid control of smoked Atlantic salmon
based electronic nose and mass spectrometry based electronic nose for the (Salmo salar) quality by electronic nose: correlation with classical evaluation
prediction of red wine spoilage. J. Agric. Food Chem. 56, 3238–3244. methods. Sensors Actuators B Chem. 116, 72–77.
Bhattacharya, N., Bandyopadhyay, R., Bhuyan, M., Tudu, B., Ghosh, D., Jana, A., Hedberg, C., 1999. Food-related illness and death in the United States. Emerg. Infect.
2008a. Electronic nose for black tea classification and correlation of measure- Dis. 5, 607–625.
ments with tea taster marks. IEEE Trans. Instrum. Meas. 57, 1313–1321. Herrero, A.M., 2008. Raman spectroscopy a promising technique for quality
Bhattacharya, N., Metla, A., Bandyopadhyay, R., Tudu, B., Jana, A., 2008b. assessment of meat and fish: a review. Food Chem. 107, 1642–1651.
Incremental PNN classifier for a versatile electronic nose. In: 2008 3rd Int. James, D., Scott, S.M., Ali, Z., O’Hare, W.T., 2005. Chemical sensors for electronic nose
Conf. Sens. Technol. 242–247. systems. Microchim. Acta 149, 1–17.
Bhattacharya, N., Tudu, B., Jana, A., Ghosh, D., Bandhopadhyaya, R., Bhuyan, M., Jung, J.-Y., Lee, C.-S., 2011. Characteristics of the TiO2/SnO2 thick film
2008c. Preemptive identification of optimum fermentation time for black tea semiconductor gas sensor to determine fish freshness. J. Ind. Eng. Chem. 17,
using electronic nose. Sensors Actuators B Chem. 131, 110–116. 237–242.
Bhattacharya, N., Tudu, B., Jana, A., Ghosh, D., Bandhopadhyaya, R., Saha, A.B., 2008d. Kashwan, K.R., Bhuyan, M., 2005. Robust electronic-nose system with
Illumination heating and physical raking for increasing sensitivity of electronic temperature and humidity drift compensation for tea and spice flavour
nose measurements with black tea. Sensors Actuators B Chem. 131, 37–42. discrimination. In: Sensors Int. Conf. new Tech. Pharm. Biomed. Res. 2005
Bhattacharyya, N., Seth, S., Tudu, B., Tamuly, P., Jana, A., Ghosh, D., Bandyopadhyay, Asian Conf. 154–158.
R., Bhuyan, M., Sabhapandit, S., 2007. Detection of optimum fermentation time Kirsching, A., 2012. Classification of meat with boar taint using an electronic nose.
for black tea manufacturing using electronic nose. Sensors Actuators B Chem. Acta Agric. Slov. 3, 99–103.
122, 627–634. Kottawa-Arachchi, J.D., Gunasekare, M.T.K., Ranatunga, M.A.B., Jayasinghe, L.,
Biolatto, A., Grigioni, G., Irurueta, M., Sancho, A.M., Taverna, M., Pensel, N., 2007. Karunagoda, R.P., 2012. Analysis of selected biochemical constituents in black
Seasonal variation in the odour characteristics of whole milk powder. Food tea (Camellia sinensis) for predicting the quality of tea germplasm in sri lanka.
Chem. 103, 960–967. Trop. Agric. Res. 23.
Blixt, Y., Borch, E., 1999. Using an electronic nose for determining the spoilage of Kumar, K.L.A., Durgajanani, S., Jeyaprakash, B.G., Rayappan, J.B.B., 2013.
vacuum-packaged beef. Int. J. Food Microbiol. 46, 123–134. Nanostructured ceria thin film for ethanol and trimethylamine sensing.
Bonneau, M., Squires, J., 2004. Boar taint: causes and measurement. Encycl. Meat Sensors Actuators B Chem. 177, 19–26.
Sci., 91–96. Längkvist, M., Coradeschi, S., Loutfi, A., Rayappan, J.B.B., 2013. Fast classification of
Botre, B., Gharpure, D., Shaligram, A., Sadistap, S., Pardo, M., Sberveglieri, G., 2009. meat spoilage markers using nanostructured ZnO thin films and unsupervised
Semiconductor sensor array based electronic nose for milk, rancid milk and feature learning. Sensors (Basel) 13, 1578–1592.
yoghurt odors identification. AIP Conf. Proc. 1137, 587–590. Längkvist, M., Loutfi, A., 2011. Unsupervised feature learning for electronic nose
Bourrounet, B., Talou, T., Gaset, A., 1995. Application of a multi-gas-sensor device in the data applied to bacteria identification in blood. In: NIPS 2011 Workshop on
meat industry for boar-taint detection. Sensors Actuators B Chem. 27, 250–254. Deep Learning and Unsupervised Feature Learning.
Brudzewski, K., Osowski, S., Dwulit, A., 2012. Recognition of coffee using differential Li, X., Gu, Z., Cho, J., Sun, H., Kurup, P., 2011. Tin–copper mixed metal oxide
electronic nose. IEEE Trans. Instrum. Meas. 61, 1803–1810. nanowires: synthesis and sensor response to chemical vapors. Sensors
Buratti, S., Ballabio, D., Benedetti, S., Cosio, M.S., 2007. Prediction of Italian red wine Actuators B Chem. 158, 199–207.
sensorial descriptors from electronic nose, electronic tongue and Loutfi, A., Coradeschi, S., Duckett, T., Wide, P., 2001. Odor source identification by
spectrophotometric measurements by means of Genetic Algorithm regression grounding linguistic descriptions in an artificial nose. Proc. SPIE Conf. Sens.
models. Food Chem. 100, 211–218. Fusion Archit. Algorithms Appl. V 4385, 273–282.
 A. Loutfi et al. / Journal of Food Engineering 144 (2015) 103–111 111

Lozano, J., Arroyo, T., Santos, J.P., Cabellos, J.M., Horrillo, M.C., 2008a. Electronic nose Rodríguez, J., Durán, C., Reyes, A., 2009. Electronic nose for quality control of
for wine ageing detection. Sensors Actuators B Chem. 133, 180–186. colombian coffee through the detection of defects in cup tests. Sensors 10, 36–
Lozano, J., Santos, J.P., Horrillo, M.C., 2008b. Enrichment sampling methods for wine 46.
discrimination with gas sensors. J. Food Compos. Anal. 21, 716–723. Roy, S., Basu, S., 2004. ZnO thin film sensors for detecting dimethyl- and trimethyl-
Macagnano, A., Careche, M., Herrero, A., Paolesse, R., Martinelli, E., Pennazza, G., amine vapors. J. Mater. Sci.: Mater. Electron. 15, 321–326.
Carmona, P., D’Amico, A., Natale, C.D., 2005. A model to predict fish quality from Santos, J.P., Lozano, J., Aleixandre, M., Arroyo, T., Cabellos, J.M., Gil, M., del Carmen
instrumental features. Sensors Actuators B Chem. 111–112, 293–298. Horrillo, M., 2010. Threshold detection of aromatic compounds in wine with an
Magan, N., Pavlou, A., Chrysanthakis, I., 2001. Milk-sense: a volatile sensing system electronic nose and a human sensory panel. Talanta 80, 1899–1906.
recognises spoilage bacteria and yeasts in milk. Sensors Actuators B Chem. 72, Scott, S.M., James, D., Ali, Z., 2006. Data analysis for electronic nose systems.
28–34. Microchim. Acta 156, 183–207.
Mani, G.K., Rayappan, J.B.B., 2013. A highly selective room temperature ammonia Shi, B., Zhao, L., Zhi, R., Xi, X., 2012. Optimization of electronic nose sensor array
sensor using spray deposited zinc oxide thin film. Sensors Actuators B Chem. by genetic algorithms in Xihu-Longjing Tea quality analysis. Math. Comput.
183, 459–466. Model.
Mani, G.K., Rayappan, J.B.B., 2014a. Novel and facile synthesis of randomly Singh, S., Hines, E.L., Gardner, J.W., 1996. Fuzzy neural computing of coffee and
interconnected ZnO nanoplatelets using spray pyrolysis and their room tainted-water data from an electronic nose. Sensors Actuators B Chem. 30, 185–
temperature sensing characteristics. Sensors Actuators B Chem. 198, 125–133. 190.
Mani, G.K., Rayappan, J.B.B., 2014b. Influence of copper doping on structural, optical Sivalingam, D., Gopalakrishnan, J.B., Rayappan, J.B.B., 2012. Structural,
and sensing properties of spray deposited zinc oxide thin films.pdf. J. Alloys morphological, electrical and vapour sensing properties of Mn doped
Compd. 582, 414–419. nanostructured ZnO thin films. Sensors Actuators B Chem. 166–167, 624–631.
Marsili, R.T., 1999. SPME-MS-MVA as an electronic nose for the study of Off-flavors Sivalingam, D., Rayappan, J.B.B., Gandhi, S., Madanagurusamy, S., Sekar, R.K.,
in milk. J. Agric. Food Chem. 47, 648–654. Krishnan, U., 2011. Ethanol and TMA sensing by ZnO based nanostructured thin
Marsili, R.T., 2000. Shelf-life prediction of processed milk by solid-phase films. Int. J. Nanosci. 10, 1161–1165.
microextraction, mass spectrometry, and multivariate analysis. J. Agric. Food Tudu, B., Jana, A., Metla, A., Ghosh, D., Bhattacharyya, N., Bandyopadhyay, R., 2009a.
Chem. 48, 3470–3475. Electronic nose for black tea quality evaluation by an incremental RBF network.
Muniyandi, I., Mani, G.K., Shankar, P., Rayappan, J.B.B., 2014. Effect of nickel doping Sensors Actuators B Chem. 138, 90–95.
on structural, optical, electrical and ethanol sensing properties of spray Tudu, B., Metla, A., Das, B., Bhattacharyya, N., Jana, A., Ghosh, D., Bandyopadhyay, R.,
deposited nanostructured ZnO thin films. Ceram. Int. 40, 7993–8001. 2009b. Towards versatile electronic nose pattern classifier for black tea quality
Musatov, V.Y., Sysoev, V.V., Sommer, M., Kiselev, I., 2010. Assessment of meat evaluation: an incremental fuzzy approach. IEEE Trans. Instrum. Meas. 58,
freshness with metal oxide sensor microarray electronic nose: a practical 3069–3078.
approach. Sensors Actuators B Chem. 144, 99–103. Vestergaard, J.S., Haugen, J.-E., Byrne, D.V., 2006. Application of an electronic nose
Olafsdottir, G., Nesvadba, P., Di Natale, C., Careche, M., Oehlenschlager, J., for measurements of boar taint in entire male pigs. Meat Sci. 74, 564–577.
Tryggvadottir, S.V., Schubring, R., Kroeger, M., Heia, K., Esaiassen, M., Wang, B., Xu, S., Sun, D.-W., 2010. Application of the electronic nose to the
Macagnano, A., Jorgensen, B.M., 2004. Multisensor for fish quality identification of different milk flavorings. Food Res. Int. 43, 255–262.
determination. Trends Food Sci. Technol. 15, 86–93. Wang, G., He, X., Zhou, F., Li, Z., Fang, B., Zhang, X., Wang, L., 2012. Application of
Olsen, E., Vogt, G., Ekeberg, D., Sandbakk, M., Pettersen, J., Nilsson, A., 2005. Analysis gold nanoparticles/TiO2 modified electrode for the electrooxidative
of the early stages of lipid oxidation in freeze-stored pork back fat and determination of catechol in tea samples. Food Chem. 135, 446–451.
mechanically recovered poultry meat. J. Agric. Food Chem. 53, 338–348. Warm, K., Martens, M., Nielsen, J., 2001. Sensory quality criteria for five fish species
Panigrahi, S., Balasubramanian, S., Gu, H., Logue, C.M., Marchello, M., 2006. Design predicted from near-infrared (NIR) reflectance measurement. J. Food Qual. 24,
and development of a metal oxide based electronic nose for spoilage 389–403.
classification of beef. Sensors Actuators B Chem. 119, 2–14. Wilson, A.D., 2013. Diverse applications of electronic-nose technologies in
Pardo, M., Niederjaufner, G., Benussi, G., Comini, E., Faglia, G., Sberveglieri, G., agriculture and forestry. Sensors (Basel) 13, 2295–2348.
Holmberg, M., Lundstrom, I., 2000. Data preprocessing enhances the Wilson, A.D., Baietto, M., 2011. Advances in electronic-nose technologies developed
classification of different brands of Espresso coffee with an electronic nose. for biomedical applications. Sensors (Basel) 11, 1105–1176.
Sensors Actuators B Chem. 69, 397–403. Yamazoe, N., Sakai, G., Shimanoe, K., 2003. Oxide semiconductor gas sensors. Catal.
Pardo, M., Sberveglieri, G., 2005. Classification of electronic nose data with support Surv. Asia 7, 63–75.
vector machines. Sensors Actuators B Chem. 107, 730–737. Yang, X., Xie, W., Zhang, C., Wu, H., Li, S., Yang, L., 2009. Identification of sensory
Perera, A., Pardo, A., Barrettino, D., Hierlermann, A., Marco, S., 2010. Evaluation of quality of rapid fermented fish using electronic nose. Int. Conf. Inf. Eng. Comput.
fish spoilage by means of a single metal oxide sensor under temperature Sci. 1–3.
modulation. Sensors Actuators B Chem. 146, 477–482. Yang, Z., Dong, F., Shimizu, K., Kinoshita, T., Kanamori, M., Morita, A., Watanabe, N.,
Peris, M., Escuder-Gilabert, L., 2009. A 21st century technique for food control: 2009. Identification of coumarin-enriched Japanese green teas and their
electronic noses. Anal. Chim. Acta 638, 1–15. particular flavor using electronic nose. J. Food Eng. 92, 312–316.
Persaud, K., Dodd, G.H., 1982. Analysis of discrimination mechanisms in the Yu, H., Wang, J., Xu, Y., 2007. Identification of adulterated milk using electronic
mammalian olfactory system using a model nose. Nature 299, 352–355. nose. Sensors Mater. 19, 275–285.
Prieto, N., Rodriguez-MÃéndez, M.L., Leardi, R., Oliveri, P., Hernando-Esquisabel, D., Yu, H., Wang, J., Zhang, H., Yu, Y., Yao, C., 2008. Identification of green tea grade
Iñiguez-Crespo, M., de Saja, J.A., 2012. Application of multi-way analysis to UV– using different feature of response signal from E-nose sensors. Sensors
visible spectroscopy, gas chromatography and electronic nose data for wine Actuators B Chem. 128, 455–461.
ageing evaluation. Anal. Chim. Acta 719, 43–51. Yu, H., Wang, Y., Wang, J., 2009. Identification of tea storage times by linear
Ragazzo-Sanchez, J.A., Chalier, P., Chevalier-Lucia, D., Calderon-Santoyo, M., discrimination analysis and back-propagation neural network techniques based
Ghommidh, C., 2009. Off-flavours detection in alcoholic beverages by on the eigenvalues of principal components analysis of e-nose sensor signals.
electronic nose coupled to GC. Sensors Actuators B Chem. 140, 29–34. Sensors 9, 8073–8082.