Pith. sign in

REVIEW 3 major objections 6 minor 28 references

Understanding the mismatch between in-vivo and in-silico rhinomanometry

T0 review · 3 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The position of the pressure tap that senses ambient pressure in a clinical rhinomanometer, not the numerical method, is what separates measured from computed nasal resistance.

desk verdict A plausible and testable mechanism for the CFD–rhinomanometry gap, but the 300–400% bias figures are likely inflated by the 2x phantom and need rescaling before the claim carries its weight. read the letter →

arxiv 2506.10865 v1 pith:EEBKB64C submitted 2025-06-12 physics.flu-dyn

Marco Atzori , Gabriele Dini Ciacci , Maurizio Quadrio This is my paper
classification physics.flu-dyn
keywords nasalresistancerhinomanometryanterioractivecomputationalfluiddynamicsdirectnumericalsimulationpressuretapbiasairwayobstructionsiliconephantom
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper argues that a large part of the longstanding disagreement between clinical rhinomanometry and numerical simulation of nasal resistance is not caused by the simulations, the anatomy reconstruction, or the turbulence model, but by where the clinical device places the pressure tap meant to measure ambient pressure. In anterior active rhinomanometry, that tap sits in the socket just before the air filter, so the mask and socket act as an artificial extension of the nasal cavity and the measured "ambient" pressure is actually reduced by flow development through a contraction. Experiments on a silicone phantom of a patient-specific nasal cavity, backed by dedicated sensors, show that using the mask tap instead of the filter tap lowers the measured pressure drop by more than 300% at low flow and more than 400% at high flow, bringing it into line with simulations. If this bias is general, it means the in-vivo/in-silico mismatch is largely a device-design problem; a simple change of probe location could improve diagnostic consistency.

What carries the argument

The load-bearing object is the pair of pressure taps that define the pressure drop in anterior active rhinomanometry: tap F, placed in the socket just before the air filter (the clinical default), and tap M, placed inside the mask in front of the nostrils. The mechanism that separates them is the sudden contraction in the duct connecting the mask to the rhinomanometer: as flow accelerates into the smaller socket, static pressure falls before the filter, so the F tap reports a pressure lower than true ambient. The paper isolates this mechanism with an idealized "mask junction" made of two cylinders and a spherical cap, simulated with DNS for three contraction radii, and quantifies the numerical background uncertainty with RANS (Reynolds-averaged Navier–Stokes, k–ω SST) and immersed-boundary DNS on grids ranging from $7\times10^5$ to $9\times10^7$ points.

What would settle it

Run the same two-tap comparison (mask tap versus pre-filter tap) on a 1:1 scale silicone phantom of the same anatomy, or on a consenting human subject, over flow rates from about 100 to 600 cubic centimeters per second; if the two taps agree within a few percent instead of the 300–400% reported here, the claimed systematic bias does not transfer to clinical size.

Watch

Extended reading notes

Core claim

Using the identical patient-specific anatomy for a silicone-phantom experiment and for simulations ranging from coarse RANS to fine direct numerical simulation, the paper establishes that the position of the reference-pressure tap is the dominant source of discrepancy between clinical and computational nasal resistance. The numerical side is internally consistent: excluding the two coarsest RANS runs, the computed pressure drops agree within about ±3% across fidelity levels, and the parallel-resistance formula used to combine left and right resistances is accurate to about 1.2%. In the experiment, the tap placed in the socket just before the filter reads a pressure that is not ambient: compared with a tap inside the mask in front of the nostrils, the filter tap inflates the pressure drop by roughly 1.3 Pa at the lowest flow rate and 17 Pa at the highest, which corresponds to more than 300% and 400%. At a common comparison point the mask-based experimental total resistance is 20.9% below the DNS reference while the filter-based estimate is 87.4% above it, and for individual nostrils the pre-filter errors reach +168% (left) and +334% (right). Simulations of an idealized mask junction show that a sudden contraction in the duct alone produces a hydraulic resistance larger than the total nasal resistance measured with the mask tap. The conclusion is that the mask and socket operate as an artificial extension of the nasal cavity, so clinical devices that assume the pre-filter probe senses ambient pressure carry a systematic bias large enough to explain the longstanding in-vivo/in-silico mismatch.

Load-bearing premise

The silicone phantom is twice life size, and the paper assumes that the pressure differences measured on it remain representative of real patient-sized anatomy even though Reynolds numbers and the relative weight of inertial and viscous losses do not scale simply with a twofold enlargement; the authors acknowledge this in the Discussion as a factor that could influence the quantitative outcome.

Editorial extensions

If this is right

  • If the bias is real, clinical rhinomanometers that reference pressure before the filter systematically overestimate nasal resistance; some patients may be classified as obstructed when the true resistance is normal.
  • CFD studies that define pressure drop from a distant ambient to the nasopharynx are not measuring the same quantity as a clinical exam; comparisons should either model the mask and socket or use a mask-based reference pressure.
  • A controlled redesign of the probe position, moving the reference tap into the mask or calibrating the socket contribution, could remove most of the reported 200–300% in-vivo/in-silico gap without any change to the computational methods.
  • The parallel-resistance formula for combining left and right nasal resistances is accurate to about 1.2% and is not the source of the mismatch.
  • The resistance of the mask junction alone exceeds the nasal resistance measured with the mask tap for contraction radii of 7.0, 7.5 and 8.0 mm, so even minor duct geometry changes matter clinically.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Beyond the paper: if the bias is design-dependent, published comparisons of CFD against clinical rhinomanometry should be revisited with attention to each device's probe placement; part of the spread across studies may reflect different masks and sockets rather than different patients or codes.
  • A natural extension is a parametric sweep of mask and socket geometries on the same phantom, producing a correction factor for the pre-filter tap that could be applied retrospectively to existing clinical measurements.
  • The same principle, that a reference-pressure tap inside a measuring device can sit within the flow's influence, likely applies to other pressure-based respiratory or ventilatory measurements that route flow through masks, mouthpieces, or connectors before the sensor.
  • Because the phantom is twice life size, a patient-sized replica with simultaneous mask and pre-filter taps would directly test whether the 300–400% bias is a scale effect or a consistent feature of the contraction.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 6 minor

Summary. The paper investigates why clinical rhinomanometry and CFD simulations of nasal resistance disagree. Using a 2×-scale silicone phantom of a patient-specific nasal cavity, the authors run a clinical 4-phase rhinomanometer side by side with calibrated laboratory sensors, measuring pressure at the sealed nostril (N), inside the mask (M), and in the socket before the filter (F). They compare these measurements with RANS and DNS simulations on the same anatomy, spanning a wide range of fidelity levels. The numerical results are internally consistent, with simulation-to-simulation differences of a few percent. The experimental pressure drop relative to the mask tap (P_N−P_M) agrees reasonably with CFD, whereas the pressure drop relative to the filter tap (P_N−P_F) is much larger, by 300–400% at the flow rates studied. The authors conclude that the location of the ambient-pressure tap in clinical rhinomanometers can be a major source of systematic bias, potentially explaining a large part of the known in-vivo/in-silico mismatch.

Significance. The paper addresses a long-standing discrepancy between clinical and computational assessment of nasal resistance, and it does so with an unusually careful experimental design: a clinical rhinomanometer is used alongside independent transducers on the same physical model that feeds the CFD, and the numerical campaign spans from coarse RANS to fine DNS without invoking any fitted parameters. If the central quantitative claim survives scrutiny, the work would motivate a re-examination of rhinomanometer design and would substantially improve the interpretation of in-vivo versus in-silico comparisons. However, the headline percentage mismatch is based on a 2×-scaled phantom, and the paper does not account for the fact that this scaling systematically amplifies the measured filter-tap bias relative to a patient-sized nose.

major comments (3)
  1. [§2.2, §3.3, Table 2] The central quantitative evidence for the tap-position bias is the large difference between ΔP_F and ΔP_M, reported as 300–400% in §3.3. This ratio is computed on a silicone phantom that is twice the linear size of the anatomy (§2.2). For a fixed volumetric flow rate Q, the factor-2 enlargement reduces velocities by a factor 4 and the nasal pressure drop P_N−P_M by roughly 8–16, depending on whether losses are predominantly viscous or inertial. In contrast, P_F−P_M is governed by the contraction into the unscaled clinical socket and therefore remains close to its clinical value at the same Q. The ratio (P_F−P_M)/(P_N−P_M) that carries the headline result is consequently amplified by approximately the scale factor. The Discussion acknowledges only that the 2× model 'could influence the quantitative outcome'; the text does not identify that this influence systematically inflates the main measured mismatch. Please provide a quantitative scaling analysis, or a life-size validation, showing how P_F−P_M compares with a patient-scale P_N−P_M. Without this, the claim that tap location causes the clinically observed 200–300% mismatch is not established.
  2. [§3.5, Fig. 9] The idealized mask-junction resistance is brought into comparison with the experimental total nasal resistance obtained from ΔP_M. The same scale mismatch applies: the junction geometry is described as representative of an actual mask, whereas the nasal resistance is measured on the 2× phantom and is therefore 8–16 times too small at the same Q. The conclusion that the junction resistance exceeds the nasal resistance may therefore be a consequence of the phantom scale rather than a property of clinical geometry. The authors should report the junction resistance against a correctly scaled estimate of nasal resistance, or simulate a life-size nasal airway with the same socket, before using Fig. 9 to support the proposed mechanism.
  3. [§3.3, abstract] The statement that the tap position 'can cause a mismatch comparable to that generally observed' relies on relative errors taken with respect to the numerically computed ΔP on the enlarged geometry. Because the phantom's nasal pressure drop is artificially small, a fixed absolute bias P_F−P_M translates into an exaggerated percentage error. At Q≈576 cm3/s, for example, ΔP_F−ΔP_M≈17 Pa; whether this is comparable to the clinical in-vivo/in-silico discrepancy depends on the patient-scale nasal pressure drop, which is not measured here. Please report the absolute bias against a clinical-scale reference, or state clearly that the quantitative transfer to patient geometry is not yet demonstrated.
minor comments (6)
  1. [Abstract, §1] The phrase 'The aim of present paper' should read 'The aim of the present paper' or 'The aim of this paper'.
  2. [§2.2, §2.3] The description of the custom mask is ambiguous: it is said to be 'built to precisely follow the contour of patient's face', but the phantom is twice the linear size of the anatomy; please clarify whether the mask was scaled to the phantom or the original face.
  3. [§3.1] There is a typo, 'laringeal striction', which should be 'laryngeal striction'.
  4. [§3.3] The sentence 'simulations gives an estimate which is around 5 Pa' should be 'simulations give an estimate'.
  5. [Table 2] Please clarify the meaning of the 'Left' and 'Right' columns: state explicitly which nostril is sealed and which is measured, and explain why the flow rate Q differs between the two sides.
  6. [§3.3] The text says the errors related to the choice of the outer-pressure proxy 'are found to be above 300%', but Table 2 lists a left-side value of +168.0% at the intermediate flow rate; please reconcile the wording with the tabulated values.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the tap-position bias is a direct measured quantity, and the CFD and mask-junction results are independent forward simulations.

full rationale

The paper's central claim is not derived from a fitted parameter or from the authors' prior results. The experimental part directly measures pressures at three locations (sealed nostril N, filter socket F, mask M) on a physical phantom, and the reported 300-400% discrepancies are ratios of these measured pressure drops (Section 3.3, Fig. 5, Table 2). The CFD contribution is an independent simulation of the same anatomy; DNS is not calibrated to the experiment, and the few operating points are matched by prior RANS estimates only to align flow-rate ranges, not to force the resulting resistance values. The idealized mask-junction study (Sections 2.4.3 and 3.5) is a forward parametric sweep over socket radius, not a fit to the measured bias. Self-citations (e.g., Ref. [17] on numerical-scheme accuracy) are used as motivation; the paper independently reproduces that conclusion by running first- and second-order RANS, so the citation is not load-bearing. The acknowledged 2x phantom scale is a limitation that could affect quantitative transfer to patient size, but this is a correctness/validity concern, not circular reasoning. No equation in the paper is equivalent to its input by construction.

Assumptions & free parameters 1 free parameters · 5 assumptions · 0 invented entities

The central claim is not derived from fitted parameters. The geometric parameter in the idealized mask junction is varied to show sensitivity, not tuned to reproduce the measured mismatch. The main burden is carried by domain simplifications such as the rigid model, steady flow, 2x scale, and custom mask, and by the standard AAR assumption that the sealed-nostril pressure represents the nasopharynx. No new physical entities are introduced.

free parameters (1)
  • radius r of the idealized mask junction small cylinder = 7.0, 7.5, 8.0 mm
    Chosen by hand as representative values and varied parametrically in Section 2.4.3 to show sensitivity; not fitted to reproduce the experimental mismatch.
assumptions (5)
  • domain assumption Pressure at the sealed nostril equals nasopharyngeal pressure
    Standard AAR assumption invoked in Section 2.1 to define PN; CFD in Section 3.1 shows a fairly uniform nasopharynx pressure, providing partial support.
  • domain assumption Steady inspiration is representative of the clinical measurement
    Adopted throughout Sections 1.2 and 2.4 to remove unsteady modelling errors; the authors note it is a simplification.
  • domain assumption The 2x enlarged silicone model preserves the qualitative pressure-loss behaviour of real anatomy
    The model is twice life size in Section 2.2; the Discussion concedes the quantitative outcome could be influenced by the scale factor.
  • domain assumption The custom semi-rigid mask is a valid substitute for the clinical compliant mask
    The compliant mask was replaced with a 3D-printed mask to avoid leaks in Section 2.3, so the exact mask geometry differs from the clinical device.
  • standard math Incompressible Newtonian fluid model is valid for nasal airflow
    Standard assumption in both RANS and DNS of nasal flow, used throughout Section 2.4.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Understanding the mismatch between in-vivo and in-silico rhinomanometry." pith.science (2026). https://pith.science/paper/EEBKB64C

@misc{pith2026250610865,
  author       = {Pith},
  title        = {Pith review of: Understanding the mismatch between in-vivo and in-silico rhinomanometry},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/EEBKB64C}},
  note         = {Machine review of arXiv:2506.10865}
}
read the original abstract

Numerical simulations and clinical measurements of nasal resistance are in quantitative disagreement. Bias introduced by the design of medical devices has not been considered until now as a possible explanation. The aim of present paper is to study the effect of the location of the probe on the rhinomanometer that is meant to measure the ambient pressure. Rhinomanometry is carried out on a 3D silicone model of a patient-specific anatomy; a clinical device and dedicated sensors are employed side-by-side for mutual validation. The same anatomy is also employed for numerical simulations, with approaches spanning a wide range of fidelity levels. We find that the intrinsic uncertainty of the numerical simulations is of minor importance. To the contrary, the position of the pressure tap intended to acquire the external pressure in the clinical device is crucial, and can cause a mismatch comparable to that generally observed between in-silico and in-vivo rhinomanometry data. A source of systematic bias may therefore exist in rhinomanometers, designed under the assumption that measurements of the nasal resistance are unaffected by the flow development within the instruments.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

28 extracted references · 28 canonical work pages

  1. [1]

    & Smith, T

    Rhee, J., Book, D., Burzynski, M. & Smith, T. Quality of Life Assessment in Nasal Airway Obstruction.The Laryngoscope113, 1118–1122 (2003)

    work page 2003

  2. [2]

    & Raviv, J

    Chandra, R., Patadia, M. & Raviv, J. Diagnosis of Nasal Airway Obstruction. Otolaryngologic Clinics of North America42, 207–225 (2009)

    work page 2009

  3. [3]

    Septoplasty and compensatory inferior turbinate hypertrophy: Long- term results after randomized turbinoplasty.Eur Arch Otorhinolaryngol254, S89–S92 (1997)

    Illum, P. Septoplasty and compensatory inferior turbinate hypertrophy: Long- term results after randomized turbinoplasty.Eur Arch Otorhinolaryngol254, S89–S92 (1997)

    work page 1997

  4. [4]

    Dinis, P. B. & Haider, H. Septoplasty: Long-term evaluation of results.American Journal of Otolaryngology23, 85–90 (2002)

    work page 2002

  5. [5]

    & Kaya, E

    Demirbas, D., Cingi, C., Chakli, H. & Kaya, E. Use of rhinomanometry in common rhinologic disorders.Expert Rev. Med. Devices8, 769–777 (2011)

    work page 2011

  6. [6]

    Quadrio, M.et al.Review of computational fluid dynamics in the assessment of nasal air flow and analysis of its limitations.Eur Arch Otorhinolaryngol271, 2349–2354 (2014)

    work page 2014

  7. [7]

    Otolaryngol.44, 801–809 (2019)

    Radulesco, T.et al.Functional relevance of computational fluid dynamics in the field of nasal obstruction: A literature review.Clin. Otolaryngol.44, 801–809 (2019)

    work page 2019

  8. [8]

    Waldmann, M.et al.An effective simulation- and measurement-based work- flow for enhanced diagnostics in rhinology.Medical & Biological Engineering & Computing(2021). 17

    work page 2021

Show all 28 references
  1. [9]

    Radulesco, T.et al.Computational fluid dynamics and septal deviations— Virtual surgery can predict post-surgery results: A preliminary study including two patients.Clinical Otolaryngology45, 286–291 (2020)

    2020

  2. [10]

    Calmet, H.et al.Computational modelling of nasal respiratory flow.Computer Methods in Biomechanics and Biomedical Engineering24, 440–458 (2021)

    2021

  3. [11]

    Cherobin, G.et al.Rhinomanometry Versus Computational Fluid Dynamics: Correlated, but Different Techniques.Am J Rhinol Allergy35, 245–255 (2020)

    2020

  4. [12]

    Kimbell, J.et al.Computed nasal resistance compared with patient-reported symptoms in surgically treated nasal airway passages: A preliminary report.Am. J. Rhinology\& Allergy26, e94–e98 (2012)

    2012

  5. [13]

    Osman, J.et al.Assessment of nasal resistance using computational fluid dynamics.Current Directions in Biomedical Engineering2, 617–621 (2016)

    2016

  6. [14]

    Berger, M.et al.Agreement between rhinomanometry and computed tomography-based computational fluid dynamics.Int J CARS16, 629–638 (2021)

    2021

  7. [15]

    & Br¨ uning, J

    Schmidt, N., Behrbohm, H., Goubergrits, L., Hildebrandt, T. & Br¨ uning, J. Com- parison of rhinomanometric and computational fluid dynamic assessment of nasal resistance with respect to measurement accuracy.Int J CARS(2022)

    2022

  8. [16]

    & Szaleniec, J

    Karbowski, K., Kopiczak, B., Chrzan, R., Gawlik, J. & Szaleniec, J. Accuracy of virtual rhinomanometry.Polish Journal of Medical Physics and Engineering29, 59–72 (2023)

    2023

  9. [17]

    & Quadrio, M

    Schillaci, A. & Quadrio, M. Importance of the numerical schemes in the CFD of the human nose.Journal of Biomechanics138, 111100 (2022)

    2022

  10. [18]

    Computational Rhinology: Unraveling Discrepancies between In Silico and In Vivo Nasal Airflow Assessments for Enhanced Clinical Decision Support

    Johnsen, S. Computational Rhinology: Unraveling Discrepancies between In Silico and In Vivo Nasal Airflow Assessments for Enhanced Clinical Decision Support. Bioengineering11, 239 (2024)

    2024

  11. [19]

    & Argale, M

    Vogt, K., Wernecke, K.-D., Behrbohm, H., Gubisch, W. & Argale, M. Four- phase rhinomanometry: A multicentric retrospective analysis of 36,563 clinical measurements.European Archives of Oto-Rhino-Laryngology273, 1185–1198 (2016)

    2016

  12. [20]

    & Meinke, M

    H¨ orschler, I., Schr¨ oder, W. & Meinke, M. On the assumption of steadiness of nasal cavity flow.Biomechanics43, 1081–1085 (2010)

    2010

  13. [21]

    & Sznitman, J

    Inthavong, K., Das, P., Singh, N. & Sznitman, J. In silico approaches to respiratory nasal flows: A review.J. Biomech97, 109434 (2019). 18

    2019

  14. [22]

    Quadrio, M.et al.Effect of CT resolution and radiodensity threshold on the CFD evaluation of nasal airflow.Med Biol Eng Comput54, 411–419 (2016)

    2016

  15. [23]

    & Skorek, A

    Tretiakow, D., Tesch, K., Meyer-Szary, J., Markiet, K. & Skorek, A. Three- dimensional modeling and automatic analysis of the human nasal cavity and paranasal sinuses using the computational fluid dynamics method.Eur Arch Otorhinolaryngol278, 1443–1453 (2020)

    2020

  16. [24]

    Tauwald, S.et al.Tomo-PIV in a patient-specific model of human nasal cavities: A methodological approach.Meas. Sci. Technol.35, 055203 (2024)

    2024

  17. [25]

    & Langtry, R

    Menter, F., Kuntz, M. & Langtry, R. Ten Years of Industrial Experience with the SST Turbulence Model.Turbulence, Heat and Mass Tansfer4, 8 (2003)

    2003

  18. [26]

    Immersed-boundary simulations of turbulent flow past a sinusoidally undulated river bottom.Eur

    Luchini, P. Immersed-boundary simulations of turbulent flow past a sinusoidally undulated river bottom.Eur. J. Mech. B/Fluids55, 340–347 (2016)

    2016

  19. [27]

    Luchini, P.et al.A simple and efficient immersed-boundary method for the incompressible Navier–Stokes equations.J. Comp. Phys.(Submitted)(2025)

    2025

  20. [28]

    & Luchini, P

    Russo, S. & Luchini, P. A fast algorithm for the estimation of statistical error in DNS (or experimental) time averages.J. Comput. Phys.347, 328–340 (2017). Short Bio of the Authors Marco Atzoriobtained a master’s degree in physics at the University of Genoa and a doctorate in...

    2017

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.