Artificial olfaction system

To develop an artificial olfaction system and to build the pattern recognition and the odour regression model for this kind of application, we use a stepwise methodology. 

The most important element of an artificial olfaction system is the ensemble of sensors translating the primary stimuli into a measurable signal, usually electric. 

The previously mentioned quantities are completely general, and their importance holds for any kind of sensor. For chemical sensors an additional parameter of great importance is the selectivity. The selectivity defines the capability of a sensor to be sensitive only to one quantity rejecting all the others. In case of physical sensors, the number of quantities is limited to a dozen and the selectivity can be achieved in many practical applications. For chemical sensors, it is important to consider that the number of chemical compounds is of millions and that the structural differences among them may be extremely subtle. With these conditions the selectivity of chemical sensor can be obtained only in very limited conditions. Lack of selectivity means that the sensor responds with comparable intensity to different species and with such a sensor it is not possible to deduce any reliable information about the chemical composition of the measured sample. Selectivity is a straightforward requisite for analytical systems where sensors and its related measurement technique are addressed to the detection of individual compounds. As mentioned in the previous section, selectivity is not found in olfactory receptors. As a consequence, artificial olfaction systems are not based on individual selective sensors, but on sensors whose selectivity can be oriented towards molecular families, or better, towards interaction mechanisms. Figure 22.5 shows a typical selectivity map related to an array of quartz microbalances (see next section) coated with different metalloporphyrins based on the same macrocycle (tetraphenyl-porphyrin) but with different metal atoms. Figure 22.5 depicts well the concept of combinatorial selectivity, namely each compounds is identified by a unique sensitivity pattern that makes possible the identification. 

The oscillation of membrane current or membrane potential is well-known to occur in biomembranes of neurons and heart cells, and a great number of experimental and theoretical studies on oscillations in biomembranes as well as artificial membranes [1,2] have been carried out from the viewpoint of their biological importance. The oscillation in the membrane system is also related to the sensing and signal transmission of taste and olfaction. Artificial oscillation systems with high sensitivity and selectivity have been pursued in order to develop new sensors [3-8]. 

The artificial intelligence systems to which sensor arrays are coupled supply the closest likeness to the human olfactory system. Some of the recent theories on olfaction require that the human nose has only relatively few types of receptor, each with low specificity. The activation of differing patterns of these receptors supplies the brain with sufficient information for an odour to be described, if not recognized. As a consequence of this belief, the volatile chemical-sensing systems commercially available only contain from 6 to 32 sensors, each having relatively low specificity. Statistical methods such as principal component analysis, canonical discriminant analysis and Euclidian distances are used for mapping or linked to artificial neural nets as an aid to classification of the odour fingerprints . 

By considering the role of timing of odorant delivery in biological olfaction (Rubin Cleland 2006), we have recently built a novel machine olfaction technology, termed an artificial olfactory mucosa , which demonstrates clearly a third principle of odour discrimination in artificial olfactory systems ... 

Chemical sensors are an almost mature technology for many practical applications. For this scope it is necessary a strong co-operation between sensors developers and end-users in order to optimise practical solutions. At this level it is important a correct and careful analysis of user needs and expectations and an education effort towards the users in order to disseminate the intrinsic novelty carried by sensors systems such as those widely belonging to the class of artificial olfaction. 

K.C. Persaud, Olfactory System Cybernetics Artificial Noses, in Handbook of Olfaction and Gustation, 2nd edition, Richard L. Doty (ed.), Marcel Dekker (2001) ISBN 0-8247-0719-2. 

It has to be remarked that in spite of the widely accepted term electronic nose, current devices are still far from the structure and functions of natural olfaction sense. The unique common feature between artificial and natural system is that both are largely based on arrays of nonselective sensors. The concept underlying electronic nose systems has been demonstrated to be independent on the particular sensor mechanism indeed during the last two decades almost all the available sensor technologies have been utilized as electronic noses. Clearly, all these sensors are very different from the natural receptors. These dissimilarities make the perception of electronic nose very different from that of natural olfaction, so that the instrumental perception of the composition of air cannot be called odor measurement because odor is the sensation of smell as perceived by human olfaction. Nonetheless, the term odor analysis with electronic noses is now largely adopted, but it is important to keep in mind, especially in medical applications, that the electronic nose measurement may be very distant from the human perception. 

Furthermore, in this system it is also possible to follow the dynamic diffusion of airborne molecules through the sensing layer, giving rise to spatio-temporal response patterns resembling those observed in the olfaction of animal models. This feature can be adequately exploited to develop a sort of artificial olfactory mucosa [44]. 

Future olfactive (olfactive mucus) detection systems will probably be ultrasensitive chemical sensors [16] detecting nucleotides and hormones (e.g., pheromones). Biosensor research may also become orientated towards new bioreceptors [265], and perhaps even new biocatalysts, such as artificial enzymes [266]. 

---