Odor sensory tests vs. In-silico prediction for the high-definition quantification of olfaction dynamics

The intricate dynamics of volatile organic compounds (VOCs) in the human respiratory system remain poorly understood. In the present study, we integrate odor sensory tests (OSTs) coupled with computational fluid dynamics and a physiologically based pharmacokinetic (CFD-PBPK) model to elucidate various aspects of the olfaction process. Safe yogurt-derived substances were incorporated in OSTs to prevent harmful exposure. Acetaldehyde was identified as a key active component in determining odor intensity, prompting further analysis of acetone and other four constituents of yogurt. Logarithmic correlations were established between the perceived odor intensity from the OSTs and both time-averaged absorption flux and equilibrium concentration within the olfactory mucus layer. These parameters were numerically captured, enabling the logarithmic approximation of odor intensity for different breathing profiles and the development of reliable prediction models for odor sensation based on quantifiable physiological parameters. The CFD-PBPK model captured detailed spatial and temporal variations of these parameters, which offers potential for future integrated/in-silico applications. Minor peaks of odor concentration were observed in the posterior olfactory regions during exhalation, revealing a retro-nasal phenomenon. Location-specific analysis revealed the nostrils and olfactory regions as the most accurate indicators of perceived odor intensity, proving the limitations of rough sensory assessments in the indoor/breathing zone scales. Acetone exhibited distinct absorption and desorption trends during the transitional phase between inhalation and exhalation, owing to the physical properties (diffusion and partition coefficients) that strongly characterize the olfaction dynamics. Author SummaryWe developed an integrated method using odor sensory tests (OSTs) coupled with computational fluid dynamics and physiologically based pharmacokinetic models (CFD-PBPK) to assess the temporal and spatial transport of yogurt odorants to the human olfactory region. Multiple phenomena were observed, including ortho-nasal and retro-nasal olfaction, temporal changes in the perceived odor intensity associated with breathing/sniffing profiles, absorption and desorption curves of acetone in the mucus epithelium, and regional-based olfaction distribution. The perceived odor intensity from OSTs can be predicted logarithmically in correlation with both the time-averaged absorption flux and the equilibrium concentration in the olfactory mucus layer, offering a reliable in-silico prediction model for odor sensation based on numerically quantifiable parameters. This model offers potential implications for multiple computational, biomedical, and industrial applications, such as the electric noses, smart odor sensors, food assessments, and fragrance development, particularly for long-term exposure in the industries that emit odorous compounds. It can open the door for more accurate predictions of the complex micro-fluid dynamics in the microbial ciliated tissues in the olfactory receptors.