Olfactory-taste synesthesia model: An integrated method for flavor responses of electronic nose and electronic tongue

Flavor perception involves a mixing phenomenon between smell and taste. This phenomenon is known as the olfactory-taste synesthesia effect (including halo, horn, nasal, olfactory-taste coexistence, and posterior nasal effects). Research on the synesthesia effect of electronic nose (e-nose) and electronic tongue (e-tongue) for flavor substance analysis is limited, which affects the analysis performance. Therefore, in this study, an olfactory-taste synesthesia model (OTSM) is proposed for human nervous systems to integrate e-nose and e-tongue data for flavor substance identification. Next, the synesthesia effect is simulated to demonstrate the effectiveness of the OTSM for flavor substance analysis. Finally, four flavor substances are identified using the OTSM. The results suggest that, first, excellent recognition results in the simulation of nasal, olfactory-taste coexistence, and posterior nasal effects are acquired with the OTSM. Second, in the simulation of halo and horn effects, a high correlation between the OTSM output and sensory scores of the halo and horn effects is obtained. The simulation results indicate that the OTSM is effective for flavor substance analysis. Finally, compared with the identification results for signal preprocessing and multipattern recognition models, the best identification results for four flavor substances, including an accuracy of 95.56%, Kappa coefficient of 94.17%, and F1-score of 95.58%, are acquired by the OTSM. In conclusion, effective identification of flavor substances with the e-nose and e-tongue is achieved using the OTSM.

Automated summary

This study proposes an olfactory-taste synesthesia model (OTSM) to integrate data from electronic nose and electronic tongue for flavor substance identification. The effectiveness of OTSM for flavor substance analysis is demonstrated through the simulation of synesthesia effect. Four flavor substances are successfully identified using OTSM. Compared with other models, OTSM achieves the best identification results for flavor substances.