Analyses on the Influence of Normal Nasal Morphological Variations on Odorant Transport to the Olfactory Cleft

Abstract

Objective:

Olfaction requires a combination of sensorineural components and conductive components, but conductive mechanisms haven’t typically received much attention. This study investigates the role of normal nasal vestibule morphological variations in ten healthy subjects on odorant flux in the olfactory cleft.

Materials and Methods:

Computed tomography images were used to create subject-specific nasal models. Each subject’s unilateral nasal cavity was classified according to its nasal vestibule shape as Standard or Notched. Inspiratory airflow simulations were performed at 15L/min, simulating passive breathing, using computational fluid dynamics. Odorant transport simulations for three odorants (limonene, 2,4-Dinitrotoluene, and acetaldehyde) were then performed at concentrations of 200ppm for limonene and acetaldehyde, and 0.2ppm for dinitrotoluene. Olfactory cleft odorant flux was computed for each simulation.

Results and Discussion and Conclusion:

Simulated results showed airflow in the olfactory cleft was greater in the Standard phenotype compared to the Notched phenotype. For Standard, median airflow was greatest in the anterior region (0.5006L/min) and lowest in the posterior region (0.1009L/min). Median airflow in Notched was greatest in the medial region (0.3267L/min) and lowest in the posterior region (0.0756L/min). Median olfactory odorant flux for acetaldehyde and limonene was greater in Standard (Acetaldehyde: Standard=140.45pg/cm²-s; Notched=122.20pg/cm²-s. Limonene: Standard=0.67pg/cm²-s; Notched=0.65pg/cm²-s). Median dinitrotoluene flux was greater in Notched (Standard=2.86×10⁻⁴pg/cm²-s; Notched=4.29×10⁻⁴pg/cm²-s). The impact of nasal vestibule morphological variations on odorant flux at the olfactory cleft may have implications on individual differences in olfaction, which should be investigated further.

Introduction

Olfaction is the sensation arising from the nasal cavity following stimulation of the olfactory receptors by odorant molecules. Olfactory dysfunction is characterized by reduced or absent sense of smell, ranging from hyposmia to total anosmia.(1) While olfactory perception is effective when the combination of sensorineural components and conductive factors function properly, the role of conductive factors (respiratory effort and nasal anatomical structure) in olfaction has been given less attention. The nasal cavity plays an essential role in odor perception, which consists of the transportation of volatile chemical molecules via airflow to the olfactory receptors.(2) In order to completely understand human olfaction, it is crucial to gain knowledge of the airflow patterns in the human nasal cavity and quantify the transport of odorant-laden air to the olfactory region.

According to Hahn et al.(3) and Keyhani, Scherer and Mozell (4), approximately 10% of inhaled air reaches the olfactory cleft during a normal resting breath, while Zhao and Jiang (5) reported about 6%±4% of flow volume passing through the olfactory cleft. Additionally, variations in nasal morphology have been traditionally understood to play a critical role in olfactory function,(6) as it influences airflow(7–13) and odorant transport in the nasal cavity.(13, 14) Morphological profiles that have been postulated to be associated with improved olfaction include the dorsal conduit, which delivers inhaled odorant-laden air to the olfactory cleft through enhanced olfactory airflow, and an enlarged olfactory cleft at the posterior end of the nasal cavity.(15–17) It has also been reported that anatomical changes in the olfactory cleft or the nasal valve region alter odorant transport to the olfactory epithelium.(18) Subsequently, as the nasal vestibule lies exactly posterior to the nostrils and is the primary point of contact between odorant-laden air and nasal mucosa, it is essential to explore the impacts of morphological variations in the nasal vestibule on airflow and odorant concentration in the olfactory regions.

One key feature of normal variations seen in both intra- and inter-individual nasal anatomy are the distinct nasal vestibule morphological phenotypes as described by Ramprasad and Frank-Ito (11) - Notched, Standard, and Elongated. Briefly, the Notched phenotype is characterized by a prominent notch at the junction of the sidewall and nasal ala, the Elongated phenotype is identified by an elongated nasal vestibule around the internal nasal valve and at the level of the nasal ala, and the Standard phenotype is considered the “typical” structure where some patients may have a small notch at the level of the nasal valve.(11) Three-dimensional representations of the phenotypes can be seen in Figure 1. These nasal vestibule morphological phenotypes were found across subjects with normal healthy nasal anatomy, and they do not influence global airflow dynamics and resistance in the nasal cavity;(8, 11, 19) nonetheless, local airflow resistance, air conditioning capability, and transport of particulate matters in the nasal cavity appear to be strongly influenced by these nasal vestibule morphological phenotypes.(8, 11, 20) Furthermore, recent findings by Sicard and Frank-Ito (21) suggest that on average olfactory airflow volume was highest in the Standard phenotype (3.0%–7.7% at 15 L/min flow simulations; 3.2%–7.8% at 30 L/min flow simulations), followed by Notched (2.0%–6.8% at 15 L/min flow simulations; 1.7%–6.1% at 30 L/min flow simulations) and Elongated (1.2%–5.4% at 15 L/min flow simulations; 2.1%–6.7% at 30 L/min flow simulations).

Figure 1.

Sagittal view of the nasal vestibule airspace showing the soft tissue (transparent) and morphology of each phenotype. Reproduced with permission from Ramprasad and Frank-Ito.(11)

Therefore, the focus of the current study is to determine how these nasal vestibule morphological structures among adult healthy subjects influence the diffusion of odorant fluxes in the olfactory cleft using computational fluid dynamics (CFD) modeling. It is our expectation that these findings will provide new understanding of how nasal vestibule morphology contributes toward olfactory perception in humans. This contribution will be significant because detailed characterization of nasal vestibule morphology on olfactory function will be determined, which will demonstrate whether or not differences in nasal vestibule phenotypes may account for some degree of variability in human olfactory function.

Materials and Methods

Study Subject

This is a retrospective study approved by the Duke University Health System Institutional Review Board for Clinical Investigations involving high resolution cone beam computed tomography (CT) images of the head from eight female and two male healthy adult subjects. The resolution of CT images had pixel size between 0.293mm and 0.469mm, with thickness ranging from 0.313mm to 1.25mm, and number of slides between 127 and 498. In order to verify that each subject had a healthy normal nasal function, we accessed their self-reported survey of nasal patency, the validated Nasal Obstruction Symptom Evaluation (NOSE) questionnaire. The NOSE scale is a 5-question quality-of-life measure where subjects rate their symptoms of nasal obstruction and nasal patency sensation from 0 (no symptoms of Nasal Airway Obstruction) to 4 (severe symptoms of nasal airway obstruction) on each question, and the total value is then multiplied by 5 to get a score from 0 to 100. The exclusion criterion in the present study was based on subjects having a NOSE score > 32.(22, 23) Subject demographic information is presented in Table 1. Briefly, the subjects in this study ranged from 26 to 72 years old and consisted of 8 female and 2 male subjects.

Table 1:

Patient Demographics.

Subject	Left Airway	Right Airway	GENDER	RACE	AGE
1	Standard	Standard	Female	Caucasian	62
2	Notched	Notched	Female	Caucasian	64
3	Notched	Standard	Female	Caucasian	29
4	Notched	Standard	Female	Caucasian	68
5	Standard	Standard	Female	Asian	50
6	Standard	Standard	Female	African-American	54
7	Notched	Notched	Female	Latina	32
8	Standard	Standard	Female	Caucasian	26
9	Notched	Standard	Male	Caucasian	65
10	Standard	Notched	Male	Caucasian	72

Nasal Airway Reconstruction

Each subject’s CT images were imported into the analysis and segmentation software, Avizo Lite 2019.3 (Thermo Fischer Scientific, Waltham, Massachusetts), for creation of subject-specific three-dimensional (3D) reconstructions of the nasal cavity. Following the creation of 3D nasal cavity models, each subject’s unilateral nasal cavity was classified according to their observed nasal vestibule morphological phenotype – Notched or Standard, based on the definition of these phenotypes described in the Introduction Section. In brief, as depicted in Figure 1, the Notched phenotype has a prominent indentation at the junction of the ala and sidewall of the nasal vestibule, while the Standard phenotype resembles what a typical configuration of the nasal vestibule show look like. Each reconstructed model was exported in stereolithography file format from Avizo and into the computer aided design and mesh generating software package, ICEM-CFD 19.0 (ANSYS, Canonsburg, PA, USA). To specify the direction of airflow in the nasal passage, an outlet surface at the end of nasopharynx and inlet box covering the front of the entire front of the nose were created. Next, on each unilateral side of the nasal cavity model, the olfactory cleft region was defined. The horizontal delineation of the olfactory cleft airspace was from the anterior plane of the middle turbinate to the anterior portion of the sphenoid sinus. The vertical height was defined from the top of the cribriform plate down 10 mm inferiorly. (24–26)

Nasal Airflow Simulations

Following the creation of subject-specific 3D nasal cavity models, a hybrid tetrahedral-prismatic mesh was generated in the nasal cavity of each subject using ICEM-CFD, which is necessary for simulating airflow and odorant transport in the airways. Approximately 4 million unstructured tetrahedral elements and a three-layer prism elements were created in each nasal cavity. Mesh refinement analysis was deemed unnecessary for this study as mesh independent numerical results characterized by asymptotic behavior was seen starting at 4 million elements from our prior work.(27) Mesh quality analysis and subsequent smoothing were performed in order to ensure that the aspect ratio for the hybrid tetrahedral-prismatic mesh elements were smoothed and thus prevented distorted elements from impacting the accuracy of the numerical simulation.

Steady-state, laminar inspiratory airflow was simulated in the nasal cavities using the CFD software package, ANSYS Fluent 19.0 (ANSYS, Inc., Canonsburg, PA) to mimic physiological inhalation conditions at 15L/min (airflow rate when breathing at rest). Given that flow velocities through the nasal cavity during low to moderate breathing typically have a Mach number <<0.2,(28) the governing equations for simulating steady incompressible flow reduces to:

$$\nabla \cdot \overrightarrow{u}=0,$$

$$\rho (\overrightarrow{u}\cdot \nabla )\overrightarrow{u}=-\nabla p+\mu {\nabla }^2\overrightarrow{u},$$

where $\overrightarrow{u}=\overrightarrow{u}(x,y,z)$ is velocity vector field, ρ = 1.204 kg/m³ is fluid density, μ = 1.825 × 10⁻⁵ kg/m − s is dynamic viscosity, and p is pressure.

The “mass flow outlet” boundary condition was specified at the outlet to target 0.0003kg/s (15L/min) to simulate laminar inspiratory airflow dynamics. Atmospheric conditions were specified at the inlet with zero gauge pressure. No-slip boundary conditions were specified along the walls (nasal mucosa), and the walls were assumed to be stationary. The Coupled scheme was implemented in the pressure-velocity coupling and second order spatial discretization was specified for pressure and second order upwind for momentum.

Odorant Transport Simulations

In order to simulate transport of odorant molecules from the inspired airflow phase into the nasal and olfactory airspace, the method illustrated by Hahn et al. (29), Schroeter et al.(30), and Keyhani et al.(4) was adapted for use in this study. For the odorant concentrations simulated, the effects of odorant on airflow pattern is negligible; therefore, the convection-diffusion equation was specified in simulating odorant transport in the nasal cavity:

$$(\overrightarrow{u}\cdot \nabla )C_{air}={\mathcal{D}}_{oair}{\nabla }^2{C}_{air}$$

where C_air is concentration of inhaled odorant at the air-phase and ${𝒟}_{oair}$ is the diffusion coefficient of odorant in air. Three odorants used in olfactory psychophysical testing were simulated: [1] Acetaldehyde (C₂H₄O), which has a green apple-like smell; [2] 2,4-dinitrotoluene (C₇H₆N₂O₄); and [3] limonene (C₁₀H₁₆), which has an orange-like scent. Table 3 shows the chemical properties of these odorants. Acetaldehyde and limonene were simulated at an exposure concentration of 200ppm, and dinitrotoluene was simulated at a concentration 0.2ppm. Simulated exposure concentrations were determined for each odorant based on the United States health exposure permissible limits.

Table 3:

Air-water Henry’s Law constants (solubility) determined using HENRYWIN version 3.20 (US Environmental Protection Agency, Estimation Program Interface Suite, https://www.epa.gov/tsca-screening-tools/epi-suitetm-estimation-program-interface) were used to estimate air-mucus partition coefficients (K_p). The Environmental Protection Agency’s On-line Tools for Site Assessment Calculation (https://www3.epa.gov/ceampubl/learn2model/part-two/onsite/estdiffusion-ext.html) were used to determine the odorant diffusion coefficients in mucus and air, D_om and D_oa, respectively.

Odorant	Molecular Formula	Kp	Doa(m2/s)	Dom(m2/s)	Solubility
Limonene	C10H16	1.05	6.30×10−6	7.00×10−10	Low
Acetaldehyde	C2H4O	190	1.24×10−5	1.23×10−9	Moderate
2,4-Dinitrotoluene (DNT)	C7H6N2O4	2.21×10−6	6.50×10−6	7.30×10−10	High

For boundary conditions to simulate odorant transport, odorant concentration at the inlet was specified with a user-defined scaler. Along the airway walls (nasal mucosa), a user-defined function (UDF) was compiled to calculate odorant-specific mass transfer coefficient for odorant flux using appropriate air-mucus partition coefficients (K_p) and diffusion coefficient in air (D_oa) for each odorant as indicated in Table 3. Tissue depth in the human nose of 1×10⁻⁴ m was used.

Results

The surface area and volume of both the entire nasal airway and nasal vestibule are displayed in Figure 2. The nasal airway surface area was slightly greater in the Standard phenotypes (Standard median=84.357mm²(IQR=10.195mm²) vs Notched median=83.620mm²(IQR=12.328mm²)) while the nasal vestibule surface area was slightly greater in the Notched phenotypes (Standard median=9.805mm²(IQR=5.531mm²) vs Notched median=10.665mm²(IQR=4.460mm²)). However, the volume of both the nasal airway and the nasal vestibule were greater in the Standard phenotypes (Standard airway median=6.772mm³ (IQR=2.345mm³); Notched airway median=5.657mm³ (IQR=1.919mm³); Standard vestibule median=0.977mm³ (IQR=0.708mm³); Notched vestibule median=0.730mm³ (IQR=0.401mm³)). The resulting surface area to volume ratios were thus greater in the Notched phenotypes for both nasal airway and nasal vestibule (Standard airway median=13.511mm⁻¹ (IQR=4.717mm⁻¹); Notched airway median=15.566mm⁻¹ (IQR=4.139mm⁻¹); Standard vestibule median=10.030mm⁻¹ (IQR=1.886mm⁻¹); Notched vestibule median=14.128mm⁻¹ (IQR=3.948mm⁻¹)).

Figure 2.

Box plot comparing anatomical size of Standard and Notched airspace and nasal vestibules. (A) Nasal Surface Area; (B) Nasal Volume; and (C) Nasal Surface Area to Volume ratio (SA:V)

To verify laminar flow in the nasal cavity, Reynolds numbers were calculated at the choana for each airspace for each subject and are presented in Table 2. Briefly, all Reynolds numbers were below the 2,000 maximum for laminar airflow and ranged from 732 to 1396. Figure 3 shows the nasal airspace of five representative subjects and their respective unilateral nasal vestibule morphological classifications - Notched or Standard. This figure depicts nasal airflow streamline in the airspace colored by velocity magnitude including regions of fastest (in red) and slowest (in blue) airflow in the nasal and olfactory airspace (Figure 3). Additionally, contours showing local air velocity and pressure from four cross-sections across the airspace are displayed next to the airflow streamlines. It is important to note that the apparent holes in the airspace of subjects 1–4 are caused by mucous plugs, and are not symptomatic.

Table 2:

Reynolds numbers calculated at the choana for each subject’s airspace.

Subject	Left Side	Right Side
1	821	679
2	732	837
3	1401	1363
4	971	901
5	799	1084
6	1226	1114
7	1396	1394
8	857	959
9	1077	1086
10	1082	1005

Figure 3.

Streamlines of airflow through the nasal airspace of five representative subjects (MIDDLE PANEL) with arrow points at Standard (S) and Notched (N) nasal vestibule morphologies. Velocity contour plots (LEFT PANEL) and pressure contour plots (RIGHT PANEL) from four cross-sections (X1 – X4) across the left (L) and right (R) nasal airspace.

To adequately describe airflow volume in the olfactory cleft across the different nasal vestibule phenotypes, volumetric flow passing through four cross-sectional slices extending across the entire olfactory airspace was computed (Figure 4A). Across the olfactory cleft, the median (IQR=interquartile range) airflow volume was greatest at the anterior region of the Standard nasal vestibule at OZ1=0.5006 L/min (IQR = 0.3367 L/min) and at the medial region of the Notched nasal vestibule at OZ3=0.3267 L/min (IQR = 0.3771 L/min). Furthermore, the most posterior end (OZ4) of the olfactory cleft had the lowest median airflow volume at 0.0756 L/min (IQR = 0.4182 L/min) for Notched and 0.1009 L/min (IQR = 0.1751 L/min) for Standard. The median olfactory airflow volume in Standard decreased anterior to posteriorly. However, for Notched, the median olfactory airflow volume in the medial regions (OZ2 and OZ3) were greater than the anterior and posterior regions. At all cross-sections, median airflow volume in Standard was greater than Notched.

Figure 4.

(A) Air flow rate at five slices anteriorly (OZ1) to posteriorly (OZ4) across the olfactory cleft; and (B) air flow partition at five slices across the olfactory cleft.

Figure 4B illustrates the fraction of airflow volume going into the olfactory cleft (relative to the rest of the nasal cavity) at the four cross-sectional planes. For Standard, the median fraction of airflow volume was greatest at OZ1 (Standard: median = 6.57%, IQR = 5.31%) and lowest at OZ4 (Standard: median = 1.31%, IQR = 1.82%). However, for Notched, median factional airflow volume was greatest at OZ2 (Notched: median = 4.84%, IQR = 2.43%) and lowest at OZ4 (Notched: median = 1.25%, IQR = 5.96%). The fraction of airflow volume reaching the olfactory cleft was greater in the Standard phenotype across all cross-sections.

Figure 5 shows box plots quantifying the amount of diffusive flux in the olfactory cleft for each inhaled odorant (acetaldehyde, dinitrotoluene, and limonene). In general, the Standard phenotype had a greater distribution of flux in the olfactory cleft for acetaldehyde and limonene. Median flux of acetaldehyde in the Notched olfactory cleft was 122.20 pg/cm²-s (IQR=56.73 pg/cm²-s) at 200ppm, while the median flux of acetaldehyde in the olfactory cleft for Standard was 140.45 pg/cm²-s (IQR=74.32 pg/cm²-s) at the same concentration. Additionally, median fluxes of limonene in the olfactory cleft at 200ppm were 0.65 pg/cm²-s (IQR=0.15 pg/cm²-s) and 0.67 pg/cm²-s (IQR=0.20 pg/cm²-s) for Notched and Standard phenotypes, respectively. In the case of dinitrotoluene, median odorant flux at 0.2ppm was greater in the olfactory cleft for the Notched phenotype; 4.29×10⁻⁴ pg/cm²-s (IQR=7.48×10⁻⁴ pg/cm²-s) for Notched and 2.86×10⁻⁴ pg/cm²-s (IQR=4.57×10⁻⁴ pg/cm²-s) for Standard. Nonetheless, as evidenced by IQR values, there was more variability of dinitrotoluene flux in the olfactory cleft in Notched than in Standard.

Figure 5.

Odorant flux at the olfactory cleft for (A) acetaldehyde; (B) dinitrotoluene; and (C) limonene.

Figure 6 shows surface contour plots of odorant flux in the olfactory cleft for a representative subject with Notched and Standard phenotypes on either side. The surface flux for acetaldehyde in the olfactory cleft was generally lower for Notched compared to Standard (Figure 6- TOP). In addition, the intensity of acetaldehyde surface flux in the Notched airspace was more elevated in the posterior region of the olfactory cleft, which is in contrast to the Standard airspace where elevated levels of acetaldehyde flux dominated the entire middle-to-superior and posterior regions of the olfactory cleft. The MIDDLE panel of Figure 6 depicts contrasting dinitrotoluene surface fluxes in the olfactory cleft between Notched and Standard. Unlike in the Notched airspace where increased intensity of dinitrotoluene fluxes were limited to a couple isolated regions in the olfactory cleft, the olfactory region for Standard is mostly covered by increased dinitrotoluene intensity levels. With regards to limonene (Figure 6- BOTTOM), differences in olfactory fluxes between the Notched and Standard sides were not as prominent as in acetaldehyde and dinitrotoluene.

Figure 6.

Contour plot of odorant flux at the olfactory cleft of a representative subject for acetaldehyde (TOP), dinitrotolune (MIDDLE), and limonene (BOTTOM).

Discussion

The ordinary human nasal cavity is distinguishable by several intra- and inter-individual morphological variabilities that can influence a typical nasal airflow profile.(10, 11, 19, 31, 32) The nasal vestibule allows for inhaled ambient air to flow into the respiratory tract. According to Inthavong et al.(8), inhaled air flowing into the notched nasal vestibule twists and swirls through to the nasal valve. However, depending on nasal anatomical variation, the vortex created can stretch posteriorly to the anterior end of the middle turbinate and beyond.(11, 19) Furthermore, Ramprasad and Frank-Ito(11) imply that local airflow patterns at the superior end of the nasal cavity were greatly influenced by nasal vestibule phenotype. This suggests that nasal morphology influences the transport of odorant-laden molecules into the olfactory cleft, due to its effect on nasal airflow patterns.(7, 12, 13)

The three odorants examined in this study – limonene, acetaldehyde, and 2,4-dinitrotoluene, – vary in terms of their physical properties. Limonene is a relatively insoluble odorant with a lemon-like odor and is considered to have a low level of toxicity.(33) Acetaldehyde carries a sharp and fruity odor and is miscible in proportions with water and most organic solvents.(34) 2,4- Dinitrotoluene is a highly toxic and soluble odorant with a slight odor.(35) Given that limonene has a very low toxicity level according to the National Institute for Occupational Safety and Health and that the revised Immediately Dangerous to Life or Health (IDLH) concentration for acetaldehyde was deemed to be 2,000ppm, these two odorants (limonene and acetaldehyde) were simulated at safe exposure concentrationof 200ppm.(36) However, because the IDLH concentration for dinitrotoluene is 6.71ppm, a much lower exposure concentration of 0.20ppm was simulated.(35)

Results from this study indicate that airflow rate and flow partition were greatest at the anterior region of the olfactory cleft for Standard nasal vestibule phenotype and middle region for Notched nasal vestibule phenotype. However, the Standard phenotype always had a greater amount of olfactory flow volume compared to Notched. This study also suggests potential implications for the impact of nasal vestibule morphology on the ability of different odorants to reach the olfactory receptors in the olfactory cleft, which has implications on olfaction. For example, while median olfactory fluxes for limonene and acetaldehyde were greater in the Standard phenotype, median flux for dinitrotoluene was greater in Notched.

In addition, our simulations results suggest that the diffusion of limonene and acetaldehyde in the olfactory cleft behaved similarly, and are both different in comparison to dinitrotoluene. The relative variability of odorant flux for limonene and acetaldehyde in the olfactory cleft was much smaller than dinitrotoluene.. One plausible explanation for the similar behavior noticed in limonene and acetaldehyde may be due to each odorant’s chemical composition. Limonene is a completely non-polar, insoluble compound, solely containing hydrogens and carbons.(37) Acetaldehyde, although slightly polar and miscible in all proportions of water, is generally composed of hydrogens and carbons with the exception of one oxygen.(34) Nonetheless, it is important to note that the amount of odorant flux in the olfactory cleft was much greater for acetaldehyde than limonene. On the other hand, dinitrotoluene is a large, highly polar and soluble compound, containing two nitrogen and four oxygens.(38) Another possible explanation relates to the simulated exposure concentrations levels, dinitrotoluene was simulated at much smaller concentrations than limonene and acetaldehyde, and given its high density(14), it could have influenced its diffusive flux in the olfactory cleft quite differently than the other two odorants (limonene and acetaldehyde).

Potential limitations are related to this being a retrospective study using existing CT scans of patients in Duke Health’s database. As a result, olfactory tests were unable to be given to patients to corroborate the CFD findings, so we are unable make any definite comment on how these results will affect clinical practice. However, this paper has clinical relevance in the hopes that it will trigger increased collection of olfactory assessment data that can be used to identify the impact of nasal vestibule phenotype on sense of smell. Importantly, it is our opinion that because all subjects in this study were normal by their NOSE score, their sense of smell should be normal. This would also imply that a correlation between NOSE score and CFD findings may not be informative due to the small variability in the subjective NOSE score combined with the objective, precise CFD measurements. However, observed differences between odorant flux in Notched and Standard phenotypes may help to illustrate why certain patient groups are predisposed to diminished sense of smell under presence of underlying nasal disease. Further investigation with a prospective study involving acoustic rhinometry and olfactory assessments would be valuable in attaching clinical significance to the results of CFD simulations.

In conclusion, this study uses computational modeling to provide important insights into the effects of nasal vestibule morphological variabilities on olfactory airflow profile, and the corresponding direct impact on the transport of molecules in odorant-laden air to the olfactory cleft. Our results indicate that median olfactory airflow volume and percent of flow partition were highest among nasal cavities with Standard nasal vestibule phenotype. In the odorant transport simulations of acetaldehyde, dinitrotoluene, and limonene, the nasal vestibule phenotype with largest median diffusive flux in the olfactory region was not the same across all odorants (Standard for acetaldehyde and limonene, while Notched for dinitrotoluene), suggesting the possible impact of a molecule’s physical properties on olfactory transport to the human olfactory cleft.