Ge-NOSE: Electronic Nose for Sniffing Food-Borne Bacteria

Gastronomy practice has become major attraction in tourism also promote food importation globally. So, controlling bacterial contamination to comply biosecurity regulations is one of imperative task for quarantine services. However detection method of bacteria causing food poisoning is laborious. Electronic nose technology has ability to recognise volatile compounds (VOCs) emitted by biological materials. Recently, the Elecetronic Nose is one of the best choice since it does not need reagen, cheap and fast. To proof-of concept, an investigation was carried out employing Ge-Nose (Universitas Gadjah Mada, Indonesia) to captured volatile emission of four food-borne bacteria: E.coli (ATCC 25922), S.thypimurium (ATCC 14028), L.monocytogenes 4b (ATCC 13932) and B.cereus (ATCC 10876). All of sample were then incubated at 37°C for 2, 8, 16, 24, 32, 40, and 48 hours then analysed using different methods such as Linear Discriminan Analysis (LDA), Quadratic Discriminant Analysis (QDA), and SupportVectorMachine(SVM). The result showed, using LDA methods, accuracy value of E.coli was 97.80±2.20; S.thypimurium: 94.60±5.40:; L.monocytogenes 98.00±2.00: and B.cereus 95.00±5.00. Using QDA methods, the accuarcy value of E.coli was 94.80±5.20; S.thypimurium: 95.60±4.40:; L.monocytogenes 92.00±8.00: and B.cereus 95.00±5.00; whereas SVM methods, it has been showed: E.coli was 97.00±3.00; S.thypimurium: 92.40±7.60:; L.monocytogenes 89.00±11.00: and B.cereus 89.00±11.00. Highest accuracy classification average (98) was achieved. Therefore, Ge-NOSE's discriminate power is able to deliver faster, accurate yet simple and inexpensive diagnostic result. © 2019 Published by ISRES.