Differentiation smelling footprints of the Chagas disease vector using an electronic nose based on artificial intelligence algorithms.

The present study aims to disclose the design of an electronic nose capable of learning and differentiating semiochemical signals from insects usable to identify species that transmit Chagas disease. The proposed device used different non-specific resistor gas sensors integrated into a system of artificial intelligence models. To validate the nose, we used eight insect species of the Triatominae subfamily and one population that was a natural carrier of the parasite Trypanosoma cruzi. Also, the discriminatory capacity of distant species was tested with other insects like Aedes aegypti (arbovirus vector) and Sitophilus oryzae (stored grains plague). As a result, the electronic nose was able to differentiate up to gender level with an accuracy of 89.64% and to differentiate Rhodnius pallensces naturally infected with T. cruzi with less than 1% error in classification. These results show that our designed device can detect particular smelling footprints, and one electronic nose like that could be a tool to discriminate against insects in the future.