Wize Sniffer

A New Portable Device Designed for Selective Olfaction

Mario D’Acunto

1,2

, Antonio Benassi

, Federica Chiellini

, Danila Germanese

1,4

, Randa Ishak

Massimo Magrini

, Emanuele Pagliei

, Paolo Paradisi

, Marco Righi

and Ovidio Salvetti

1,2

Istituto di Scienza e Tecnologie dell’Informazione ISTI-CNR, via Moruzzi 1, 56124 Pisa, Italy

NanoICT Laboratory, Area della Ricerca CNR di Pisa, via Moruzzi 1, Pisa, Italy

Department of Chemistry and Industrial Chemistry, University of Pisa, Pisa, Italy

Department of Information Engineering, University of Pisa, Pisa, Italy

Department of Civil and Industrial Engineering, University of Pisa, Pisa, Italy

Cosmed Srl, via dei Piani di Monte Savello 37, 0041 Pavona di Albano, Rome, Italy

Keywords: Breath Analysis, Portable Device, Statistical Data Processing, Chemical Sensors.

Abstract: Digital semeiotics is one of the newest recent challenges for assessing a number of computational

descriptors to atherosclerotic cardiovascular diseases that are leading causes of mortality worldwide. These

descriptors can be expressed involving (i) morphometric, biometrics and colorimetric of the face; (ii)

spectroscopic analysis of skin and iris, of sub-cutaneous substances and the function of subcutaneous

tissues, and (iii) compositional analysis of breath and exhaled. In this paper, we describe the design and

functionality of the Wize Sniffer (WS), a new portable device for breath analysis limited to an effective

number of substances. Within the SEMEOTICON Project by the WS, we intend a hardware/software tool

for both the analysis of volatile organic compounds of breath and a platform for data mining and data

integration. The WS should be able to provide useful information about the “breathprint”, i.e., the analog of

fingerprint for the state of health of an individual.

1 INTRODUCTION

Atherosclerotic cardiovascular diseases (ACDs)

represent the leading cause of worldwide mortality

(World Health Organization. The Global Burden of

Disease: 2004 and update 2008; (Lison et al., 2010)).

Breath gases are recognized to be excellent

indicators of the presence of diseases and clinical

conditions. Such gases have been identified as

biomarkers using instrumentations such as gas

chromatography (GC) or electronic nose (e-nose)

(Guo et al., 2007). GC is very accurate but

expensive, time consuming and non portable. E-nose

has the advantages of low-cost and easy operation,

but is not particularly useful for analyzing breath

substances. As a consequence, in recent years, the

necessity to develop a portable device for breath

analysis, easy to use, and feasible for patients living

far from medical structures or physicians.

Essentially, a design of a portable device for

breath analysis is based on selected chemical sensors

that are sensitive to the biomarkers and compunds in

breath substances and make use of accurate

statistical tools for odor signal preprocessing,

classification methods and, in turn, identification of

possible diseases if any. In SEMEOTICONS

activity, we are developing a device following the

above requirement. The device, called Wize Sniffer

(WS), captures breath samples, the chemical

selective sensors sense the sample and accordingly

form a sort of odorprint of healthy people or patients

with known and specific diseases, in order to

evaluate the well-being state of a human subject

(Miekish et al., 2004). It should be noted that does

not exist a general definition of well-being, rather

some indices for well-being that can be correlated to

cardio-metabolic risk. In addition, such indices are

strictly connected to the recognition of behavioural

trends and the methods for computing the raw

numerical values of such well-being indices must be

designed for single individual. However, one critical

index of well-being is given by the breath

composition. Human breath is largely composed of

oxygen, carbon dioxide, water vapour, nitric oxide,

577

D’Acunto M., Benassi A., Chiellini F., Germanese D., Ishak R., Magrini M., Pagliei E., Paradisi P., Righi M. and Salvetti O..

Wize Sniffer - A New Portable Device Designed for Selective Olfaction.

DOI: 10.5220/0004938605770582

In Proceedings of the International Conference on Health Informatics (SUPERHEAL-2014), pages 577-582

ISBN: 978-989-758-010-9

 2014 SCITEPRESS (Science and Technology Publications, Lda.)

and numerous volatile organic compounds (VOCs)

(Di Francesco et al., 2005); (D’Amico et al., 2007).

The type and number of VOCs in the breath changes

among different individuals, but there is a common

core of breath VOCs which are present in all

individuals. The molecules in an individual’s breath

may be exogenous or endogenous. Exogenous

molecules are those that have inhaled or ingested

from the environment or other sources such as air or

food, thus giving no diagnostic value. Endogenous

molecules are produced by metabolic processes and

partition from blood via the alveolar pulmonary

membrane into the alveolar air. Such endogenous

molecules are present in breath relatively to their

types, concentrations, volatilities, lipid solubility and

rates of diffusion as they circulate in the blood and

cross the alveolar membrane. Changes in the

concentration of the molecules in VOCs could

suggest various diseases or at least changes in the

metabolism.

In this paper, for the first time we present the

design of a portable device able to operate with a

limited number of breath VOCs, normally connected

to oxidative stress, thus giving information to

physicians on the possible state of wellness of an

individual.

2 BREATH ANALYSIS AND

WELLNESS STATE

In this section, we describe the correlation between

exhaled VOCs and oxidative stress. Oxidative stress

is the cause of over 100 diseases, including

atherosclerosis or hypertension. In addition,

hypertension is a risk factor that can lead to

myocardial injury, cardiac failure, aortic aneurysm.

The design of a portable device for breath analysis

must necessarily work on a limited space of breath

VOCs. The most relevant VOCs and their

connection to oxidative stress are labelled as

follows.

Carbon Monoxide (CO): it is naturally produced

by the action of heme oxygenase on the heme for

haemoglobin breakdown. This produces

carboxyhemoglobin, which is a more stable

molecule than oxyhemoglobin. For example, an

increase of CO leads haemoglobin to carry less

oxygen through the vessels. CO is present in

cigarette smoke, and air pollution too. In addition, it

is a blood vessel relaxant (vasodilator), and a

promoter of neurovascular growth, therefore it could

be considered as an indicator of brain tumor growth.

Carbon Dioxide (CO

): the variation rate of CO

can be associated to different metabolic response,

for example, CO

increases during physical activity

or in general with the increment of blood-heat as in a

fever. Analogously, there is a decrease in the case of

hypothermia. In healthy individuals, partial pressure

in arterial blood is very close to the partial pressure

in expired gases. In most forms of lung diseases and

some of congenital heart disease (cyanotic lesions-

bluish-grey discoloration of the skin, lack of O

the body), a decrease of CO

exhaled rate is

commonly observed. Capnography is a technique

commonly used for monitoring the concentration of

, or, similarly, its partial pressure using such

technique, the max value of a capnogram

corresponds to the end of tidal volume of exhaled

breath and the steady-state concentrations of each

breath. It must be noted that the breathing rate

influences the level of CO

in the blood: slow

breathing rates cause Respiratory Acidosis (i.e.,

increase of blood CO

partial pressure, that may

stimulate hypertension, or heart rate acceleration).

On the contrary, too rapid breathing rate leads to

hyperventilation, which may provoke Respiratory

Alkalosis (i.e., decrease of blood CO

partial

pressure, no longer fits its role of vasodilator,

leading to possible arrhythmia or heart trouble).

Ethanol: the ethanol compound in a breath

composition can be classified as endogenous or

exogenous. Ethanol endogenous may increase in

exhaled gas mixtures because of alcoholic

fermentation of an excessive over-load of

carbohydrates. Indeed it is originated from microbial

fermentation of carbohydrates in the gastro-intestinal

tract. Then, it enters in the blood. On the contrary,

Ethanol exogenous comes from alcoholic drinks. It

is important to note that it is recognized that ethanol

breakdown leads to an accumulation of free radicals

into the cells, a clear example of oxidative stress.

Ethanol may cause arrhythmias and depresses the

contractility of cardiac muscle.

Nitrogen Oxide (NO): it is a vasodilator and it

modulates inflammatory response (operating in

combination with CO and Hydrogen Sulfide).

Hydrogen Sulfide (H

S): it is a vascular relaxant

agent, and has a therapeutic effect in various

cardiovascular diseases (myocardial injury,

hypertension). In general, H

S could have

therapeutic effect against oxidative stress due to its

capability to neutralize the action of free radicals. In

patients with coronary heart disease, H

S level in

blood is normally reduced, as in the case of

hypertension. On the contrary, in haemorrhagic

shock, H

S plasma levels are increased.

Ammonia (NH

): an increase of NH

in blood

HEALTHINF2014-InternationalConferenceonHealthInformatics

578

may be caused by cigarette smoke, renal failure,

cardiac failure, changes in cardio-circulatory system.

Hydrogen and Methane: they are produced by

the breakdown of the carbohydrates in the colon by

anaerobic bacteria. They can be correlated to

overweight problems, and probably to an improper

life-style. Nevertheless, their quantity is very low

presenting some problems for their detection.

Cigarette smoke: it contains principally Nicotine,

CO, NH

, oxidant substances leading to oxidative

stress. After 1 minute smoking, an increase of heart

rate can be revealed. Tobacco combustion causes a

constriction of blood vessels.

Other factors leading to oxidative stress are

cholesterol and fibrinogen. An increase of

cholesterol leads to atherosclerotic issues and,

similarly, an increase of fibrinogen leads to blood

viscosity. In addition, a strong correlation exists

between exhaled hydrocarbons and cardiovascular

diseases, principally pentane and ethane. In healthy

subjects, there is a balance between oxidant

molecules (free radicals) and anti-oxidant molecules.

When free radicals are balanced, they have a

positive effect on the general person wellness and

health, because they are involved in many

inflammatory processes, thus having a therapeutic

potential. On the contrary, when free radicals

number exceeds the anti-oxidant ones, this leads to

oxidative stress.

3 THE WS DESIGN

The design of the WS platform device operates in

three phases: gas collection, sampling and data

analysis. During the first phase the subject breaths

into a gas sampling box. Our design follows the

recent proposal by Guo et al., (2007). The gas is

injected into the chamber containing a sensor array.

Then, a measurement circuit measures the

interaction between the breath and the array sensors.

Then signals are filtered and amplified and

afterwards sent to computer for further analysis. The

WS components are array sensors contained in a

customized box, the microcontroller board and the

statistical analysis tool.

The array sensors are both commercial and

customized with conductive polymers electrospun

nanofibers as sensing material. When using a

conductive polymer, it is reasonable to have an

increased sensor sensitivity and, moreover, the

possibility to measure nicotine concentration. Note

that a correspondent commercial sensor for nicotine

is not available.

Table 1: Specific VOC and corresponding sensitive gas

capability of Figaro sensors.

Breath substance

Baseline levels

(ppm)

Sensors

Sensitivities (ppm)

(in percentage

4%)

40000 30-300

3.5 (14-30 with

moke)

50-1000

Ethanol

0.62

153-180

1-10

50-1000

Hydrogen sulfide 0.33 0.1-1

Ammonia 0.42-0.46

1-10

30-300

Hydrogen 9.1

1-10

10-100

The used commercial chemical sensors are

provided by the Figaro Engineering Inc. These

sensors are very sensitive, robust and resistant to

humidity and ageing. In table 1, the sensitivities of

such sensors for any VOC to be detected are

reported.

Any chemical sensor should have has unique

odour-print corresponding to the capacity to sense a

single VOC substance. The framework of the WS

consists of three modules: signal measurement,

signal conditioning and signal acquisition. The array

of sensors is contained in the signal measurement

module. The measurement circuit is responsible for

transforming gas signals into electronic signals. The

array sensors are mounted directly or connected with

a griddle to a microcontroller board that can operate

on the data collected filtering and amplifying them.

The microcontroller used in our system is a widely

employed open source controller: Arduino 2560

system that can be powered very simply via a USB

connection (see Figure 1).

Figure 1: The microcontroller used for the electronic data

collection: Arduino Mega 2560 system (based on the AT

mega 2560), that can be powered very simply via a USB

connection. The VOCs data collected can be loaded on

Arduino or sent to computer with the USB connection.

The length of the Mega 2560 system is 10 cm.

WizeSniffer-ANewPortableDeviceDesignedforSelectiveOlfaction

579

Figaro Engineering provides the chemical

sensors and a general purpose electric scheme to

adopt according to the specific use. The ability to

sense a selective VOC substance depends on the

employed sensing materials. For example,

considering the sensor named TGS2600, able to

sense CO, Ethanol and iso-butane, the

manufacturing procedure is as follows: using thick

film techniques, the sensor material is printed on

electrodes (noble metal) that have been printed onto

an alumina substrate. The main sensing material of

the sensor element is tin dioxide (SnO2). One

electrode of the sensing material is connected to a

pin 2 (see Figure 2) and the other is connected to

another pin 3. An RuO2 heater printed onto the

reverse side of the substrate and connected to pins 1

and 4 heats the sensing material. Lead wires are Pt-

W and connected to sensor pins which are made of

Ni-plated Ni-Fe 50%. The sensor base is made of

Ni-plated steel. The sensor cap is made of stainless

steel and contains 6 pin holes on the top of the

sensor. The basic measuring circuit is represented in

Figure 2.

Figure 2: Basic measuring circuit of Figaro chemical

sensors.

Circuit voltage (VC) is applied to the electrodes of

the sensor element. The behaviour of the sensor

element is that of a variable resistance (RS). In order

to accord the circuit to the acquiring device it is

necessary to use the load resistor (RL). Vc is always

required for the circuit power supply, and the

polarity shown in Figure 2 must be maintained. The

sensor value is measured indirectly as a change in

voltage across RL and RS. The

signal measurement module measures these voltages

and converts them into digital signals. The analog

signals are then conditioned by signal filtering and

amplifying. The sampling procedure is based on four

steps described as follows:

1) before the sampling, the chamber is purged and

the sensor returns to a steady state. The baseline

value is measured and recorded for data

manipulation and normalization;

2) 0~5s (injection stage): sampled gas is injected

into the chamber at an invariable rate. The

particles of sampled gas that are injected and

accrue inside the chamber produce a resistance

change and cause the amplitude of the signal to

rise;

3) 6 ~10s (reaction stage): particles in the chamber

continue to accumulate on the sensors but the

accumulation rate is decreasing. The resistance

of the sensor monotonically increases at a

decreasing rate, as the amplitude of the signals;

4) 15~90s (purge stage): the chamber is purged

again. The pump quickly draws out the

remaining analyte, thereby shortening the

sampling time as well as refreshing the air for the

next use.

3.1 Electrospun Nanofibers as Sensing

Materials

The sensitivity of commercial sensors, as well as

selectivity, can be improved reducing the sensing

area. This achievement can be obtained using as

sensing material conductive polymers elestrospun

nanofibers, (see Figure 3).

Figure 3: A schematic sketch of electrospinning process.

Electrospinning is a manufacturing process able

to produce ultra-fine fibers with diameters ranging

from 50-500nm (Ding et al., 2009); (Chen et al.,

2011). In the electrospinning process an electric

field is generated between a polymer solution and a

metal collection screen (collector). The polymer

solution is contained in a syringe with a steel

capillary tip and a voltage is applied between the tip

and the conductive collector. As the electrical

potential is increased, the charged polymer solution

is attracted to the collector. When the voltage

reaches a critical value, the charge overcomes the

surface tension of the polymer solution formed on

HEALTHINF2014-InternationalConferenceonHealthInformatics

580

the capillary tip and an electrified jet is produced. As

the charged jet flies through the air, it loses the

solvent through a quick evaporation and impinges on

the collection screen as a thin fiber mat, as shown in

Figure 4.

Figure 4: Example of electrospun nanofibers as observed

at a Scanning Electron Microscope.

In the development of the WS, we will use

electrospun nanofibers as sensing material and

conductive polymers, such as polyaniline (PANi),

polypirrole (PPY), poly(p-phenylene vinylene)

(PPV) and others to be defined yet (Slater et al.,

1992); (Slater et al., 1993); (Savage et al., 2001).

The advantage of sensors using nanofibers as

sensing material would be a general increment of

sensitivity and a significant reduction of the

response time. Nevertheless, such properties must be

evaluated during the measurement stages, and the

possible response of such sensors is beyond the

scope of the present paper.

3.2 Statistical data processing

Since the operability of WS works on a reduced

space of breath substances, data analysis requires a

particular care because of the various sources of

errors. The sensors measure changes in voltage

across each sensor and convert the raw signal into a

digital value that can be managed by physicians. Our

analysis of sensor data requires three steps: signal

preprocessing, feature extraction and classification.

The aim of signal preprocessing is to compensate

drifts and to eliminate irrelevant information so to

improve the performance of the subsequent pattern

recognition and classification. It is assumed that the

dynamic response of any sensors is sampled

introducing reasonable time intervals in a particular

way for the transient phase. In addition,

normalization is used to compensate for sample-to-

sample variations caused by analyte concentrations

and pressure of oxygen. At this point, we introduce

feature extraction from any sample using the

standard procedure of principal component analysis

(PCA) and, finally, k-nearest neighbor voting rule

(KNN) will be used as a classifier for the features

that extracted by PCA, (Guo et al., 2007).

4 CONCLUSIONS

In this paper, we have presented the basic

requirements for the development of the portable

device WS, operating on a number of effective

VOCs in the breath. Within the SEMEOTICON

Project, by the WS we intend a hardware/software

tool for both the analysis of VOCs of breath and a

platform for data mining and data integration.

Similar to a fingerprint, every individual has a

“breathprint” that can provide useful information

about his or her state of health. Thousands of unique

substances in exhaled breath have been identified.

These substances include elemental gases, such as

hydrogen, nitric oxide (NO), and carbon monoxide,

and a vast array of VOCs. The WS design has been

conceived to identify target VOCs with selective

sensors such as: nitric oxide, alcohols, carbon

monoxide and, possibly, other VOC substances

based on general requirement by physicians. In turn,

the breath gas sensors will be included in a box and

interacting with a main processing unit. In its

ultimate shape and functionality the WS will be a

compact and feasible device to be used in many

open environments and the most disparate situations

without losing efficiency.

ACKNOWLEDGEMENTS

Sara Colantonio, Paolo, Marraccini, Giuseppe

Coppini and Serena Detti are warmly acknowledged

for useful suggestions.

REFERENCES

World Health Organization. The Global Burden of

Disease: 2004 and update 2008.

Lison M. W.; Carl H.; Rafael P.; Dan L.; David N.; Mant

D., Glasziou P., 2010. What are the basic self-

monitoring components for cardiovascular risk

management?, MC Medical Research Methodology,

10:105.

Miekisch W.; Schubert J.; Nöeldge-Schomburg, G., 2004.

Breath analysis in critically ill patients: potential and

WizeSniffer-ANewPortableDeviceDesignedforSelectiveOlfaction

581

limitations, Expert Review Molecular Diagnosys , 4,

619-629.

Di Francesco F.; Fuoco R.; Trivella M.; Ceccarini A.

2005. Breath Analysis: trends in techniques and

clinical applications, Micromechanical Journal, 79,

405-410.

D’Amico, A.; Di Natale C.; Paolesse R.; Macagnano A.;

Martinelli E.; Pennazza G.; Santonico M.; Bernabei

M.; Roscioni C.; Galluccio G. et al., 2007. Olfactory

systems for medical applications, Sensors &

Actuators: B-Chemical, 130, 458-465.

Guo D.; Zhang D.; Li N.; Zhang L.; Yang J., 2007. A

Novel Breath Analysis System Based on Electronic

Olfaction, IEEE Transaction on Biomedical

Engineering.

Ding B.; Wang M.; Yu J.; Sun G., 2009. Gas Sensors

based on Electrospun Nanofibers, Sensors, 9, 1609-

1624.

Chen D.; Lei S.; Chen Y., 2011. A single Polyaniline

Nanofiber Field Effect Transistor and Its Gas Sensing

Mechanisms, Sensors, 11, 6509-6516.

Slater, J.; Watt, E.; Freeman, N.; May, I.; Weir, D., 1992.

Gas and vapor detection with poly(pyrrole) gas

sensors. Analyst, 117, 1265-1270.

Slater, J.; Paynter, J.; Watt, E., 1993. Multilayer

conducting polymer gas sensor arrays for olfactory

sensing. Analyst, 118, 379-384.

Savage, N.; Chwierogh, B.; Ginwalla, A.; Patton, B.;

Akbar, S.; Dutta, P., 2001. Composite n-p

semiconducting titanium oxides as gas sensors. Sens.

Actuat. B-Chem, 79, 17-27.

HEALTHINF2014-InternationalConferenceonHealthInformatics

582