# Fileset

[Advanced Sensor Research - 2026 - Shiba - Breathing‐Induced Airflow Measurements at Multiple Positions in a 3D‐Printed.pdf](https://mdr.nims.go.jp/filesets/2eb74754-b09a-4f92-846c-eec8817114dc/download)

## Creator

[Kota Shiba](https://orcid.org/0000-0001-7775-0318), Kayoko Saito, Masayoshi Tei, Nagomi Yonezawa, [Rumi Sekine](https://orcid.org/0000-0001-8686-195X), Monami Nagai, Hirotaka Tanaka, Yuji Kishimoto, Ryo Tamura, Genki Yoshikawa, Nobuyoshi Otori, Eri Mori

## Rights

[Creative Commons BY Attribution 4.0 International](https://creativecommons.org/licenses/by/4.0/)

## Other metadata

[Breathing‐Induced Airflow Measurements at Multiple Positions in a 3D‐Printed Nasal Cavity](https://mdr.nims.go.jp/datasets/a10842c0-6310-485d-91ad-db5e060b9552)

## Fulltext

Breathing‐Induced Airflow Measurements at Multiple Positions in a 3D‐Printed Nasal CavityAdvanced Sensor Research www.advsensorres.comRESEARCH ARTICLEBreathing-Induced Airflow Measurements at Multiple Positions in a 3D-Printed Nasal Cavity Kota Shiba1 Kayoko Saito1 Masayoshi Tei2 Nagomi Yonezawa2 Rumi Sekine2 Monami Nagai2 Hirotaka Tanaka2 Yuji Kishimoto2 Ryo Tamura3 , 4 Genki Yoshikawa1 , 5 Nobuyoshi Otori2 Eri Mori2 1 Research Center for Macromolecules and Biomaterials (RCMB), National Institute for Materials Science (NIMS), Tsukuba, Ibaraki, Japan 2 Department of Otorhinolaryngology, The Jikei University School of Medicine, Tokyo, Japan 3 Center for Basic Research on Materials, National Institute for Materials Science (NIMS), Tsukuba, Ibaraki, Japan 4 Graduate School of Frontier Sciences, The University of Tokyo, Kashiwa, Chiba, Japan 5 Materials Science and Engineering, Graduate School of Pure and Applied Science, University of Tsukuba, Tsukuba, Ibaraki, Japan Correspondence: Kota Shiba ( shiba.kota@nims.go.jp) Eri Mori ( morieri@jikei.ac.jp) Received: 13 February 2026 Revised: 11 May 2026 Accepted: 15 May 2026 Keywords: 3D printing | airflow | fluid dynamics | nasal cavity | pressure differential | rhinomanometry ABSTRACT Evaluating nasal cavity airflow is crucial in both fundamental science and clinical settings. While rhinomanometry (RM) effectively assesses global nasal resistance, developing alternative methods for detailed local structural evaluation remains challenging. Here, we report an approach to measure local pressure at multiple positions in the nasal cavity using 3D-printed models created based on computed tomography data of five healthy individuals. This approach records multiple pressure differential values at arbitrary positions in the nasal cavity during a breathing-like inhalation/exhalation cycle. The measured data are unique to the corresponding individuals and correlate well with the RM results from human-assisted breathing using the 3D-printed models. Our experimental approach will open up possibilities for a deeper understanding of the relationship between fluid dynamics in the nasal cavity and practical air flow in the nasal cavity. 1B  e  o  n  a  t  p  a  b  n  o  f  m  c               To©Ah Introduction reathing is essential for humans, enabling both pulmonary gasxchange and olfactory perception. During breathing, sniffing,r other inhalation/exhalation-processes, air flows through theasal cavity—a complex structure that significantly influencesirflow dynamics—and poses challenges for experimental inves-igation. Currently, rhinomanometry (RM) is widely used torovide an objective pressure-flow assessment of nasal patencynd to quantify global nasal airway resistance [ 1–10 ]. However,ecause the measurement is inherently lumped over the entireasal passage, it does not directly resolve where pressure lossesccur or how local anatomical features shape the intranasallow distribution. The inherent limitation of RM, which lumpseasurements over the entire nasal passage, often leads tolinical inconsistencies, such as the paradoxical sensation of  his is an open access article under the terms of the Creative Commons Attribution Licenriginal work is properly cited. 2026 The Author(s). Advanced Sensor Research published by Wiley-VCH GmbH dvanced Sensor Research , 2026; 5:e70162 ttps://doi.org/10.1002/adsr.70162nasal obstruction, even with a seemingly patent airway [ 11, 12 ].While computational fluid dynamics (CFD) simulations offer apowerful alternative by providing spatially resolved predictionsof intranasal flow and transport [ 13–18 ], its quantitative outcomescan be sensitive to modeling choices and uncertainties in patient-specif ic geometry reconstruction from computed tomography(CT) images [ 19, 20 ]. Moreover, the development of CFD asa reliable, patient-specific decision-support tool is constrainedby the scarcity of spatially resolved validation data within thenasal cavity [ 21, 22 ]. In this context, there is a clear need foran experimental approach that enables direct, localized mea-surements of nasal airflow dynamics while preserving subject-specif ic anatomy. Unlike RM and CFD, position-specific pressuremeasurements within the nasal cavity have remained largelyunexplored. As only a few previous studies have demonstratedlocal pressure measurements in the nasal cavity [ 23, 24 ], devel-se, which permits use, distribution and reproduction in any medium, provided the 1 of 8http://www.advsensorres.comhttps://doi.org/10.1002/adsr.70162https://orcid.org/0000-0001-7775-0318https://orcid.org/0000-0001-8686-195Xmailto:shiba.kota@nims.go.jpmailto:morieri@jikei.ac.jphttp://creativecommons.org/licenses/by/4.0/https://doi.org/10.1002/adsr.70162http://crossmark.crossref.org/dialog/?doi=10.1002%2Fadsr.70162&domain=pdf&date_stamp=2026-06-01o  C  a  oI  p  e  m  i  t  a  W  m  d  o22CT  p  w  u  t  (  (  d  o  D  f  S  a  s  i  w  r  a  p  h  t  o  i  c  t  w  p  c  c  s  s  s  i  c  e  r  w  r                                           2 27511219, 2026, 6, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/adsr.70162 by Kota Shiba - National Institute For , Wiley Online Library on [02/06/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creping such an approach would provide crucial validation data forFD and deepen our understanding of the relationship betweenirflow, nasal structure, and physiological sensations such aslfaction. n this study, we described an approach for measuring localressure in the nasal cavity using 3D-printed nasal cavity mod-ls. We employed commercial pressure differential sensors toeasure pressure differential values during breathing-mimickednhalation/exhalation cycles at various flow rates. We insertedhese sensors into the 3D-printed models through holes createdt multiple positions to gather detailed site-specific information.e also performed RM measurements using the same 3D-printedodels and compared the results with the sensor-based values toemonstrate the validity and discussed the potential applicationf our experimental approach.  Methods and Materials .1 Reconstruction and 3D Printing of Nasal avity Models he first step in our approach involves the precise fabrication oferson-specific nasal cavity models. In this cross-sectional study,e recruited five healthy individuals (denoted as Ind. 1–5) andsed CT (3D Accuitomo, J. MORITA MFG. CORP.) to acquirehe nasal cavity data for 3D printing. The CT data’s voxel sizedissection to dissection distance) was 0.25 mm. The 3D printerAgilista, KEYENCE Corp.) offered high resolution (635 dpi × 400pi lateral, 0.015 mm vertical), ensuring accurate reproductionf the nasal cavity’s complex structure. Each CT dataset inigital Imaging and Communications in Medicine (DICOM)ormat was edited using commercial software (TomoShop Viewer,ystemcreate Co., Ltd.) to open 12 through-holes (both the rightnd left noses had six through-holes) for sensor insertion, ashown in Figure 1a . The holes of the right and the left nose weren the same position. The 12 through-holes for pressure sensorsere standardized across all subjects by targeting key anatomicalegions: the inferior meatus, the middle meatus, the nasal floor,nd the olfactory cleft. We chose these locations to represent therimary airflow pathways and clinically significant areas. Theoles were created with consistency for all five individuals. First,he holes for the middle meatus were created where the naturalstium of the maxillary sinus is located. Second, the holes for thenferior meatus, the nasal floor, and the middle part of olfactoryleft were created on the same coronal plane of the holes forhe middle meatus. Last, two more holes for the olfactory cleftere added on the anterior and posterior end of the cribriformlate. After creating the holes, the edited DICOM data wereonverted to STL format for 3D printing. The actual printed modelonsisted of three objects: a main body and two small parts, ashown in Figure 1b . These two parts were carved out to facilitateensor insertion and realize reproducible measurements. The twomall parts were detachable; they were removed during sensornsertion into the main body, then reattached and sealed withommercial plumbing putty to ensure an airtight measurementnvironment. The step to wash out unnecessary 3D printingemaining in the model took several days with extra care andas completed with endoscopic observation to ensure thoroughemoval. of 8a2.2 Measurement System and Procedure The experimental setup was constructed to simulate physiologicalairflow under strictly controlled and reproducible conditions.The system consisted of several specialized units, including apump, solenoid valves, flow meters, and a high-precision pressuredifferential sensor. We employed a double-head type oil-less piston pump (100RND-ED, G&M Tech Inc.) because its flow rate range sufficientlycovered typical human breathing levels. As shown in Figure 2a ,the pump featured both suction and discharge ports, which wereintegrated with four direct-acting 2-port solenoid valves (FFBseries, CKD Corporation). This configuration allowed us to mimicbreathing cycles by alternating the flow direction; specifically, weprogrammed the system for a standard cycle of 5.0 s for inhalationand 5.0 s for exhalation. Furthermore, the system demonstratedthe flexibility to replicate faster breathing patterns, such as 1.0 sinhalation and 1.5 s exhalation, or even rapid cycles as short as 0.2s. To capture the subtle aerodynamic changes within the model, weused a low-differential-pressure sensor (PSE550, SMC Corpora-tion). This sensor was selected for its high precision (0.001 kPaaccuracy), which was essential for the objectives of this study.Airflow rates were monitored using a 2-color display digital flowswitch (PF2M7 series, SMC Corporation), which provided reliablemeasurements across the required flow range. All data wererecorded via a data logger (USB-6002, Emerson Electric Co.) andcontrolled using LabVIEW software. The pressure differential sensor was connected to a pipe viatubing, and the tip of this pipe was inserted into the modelthrough the standardized holes (Figure 2b ). A custom-built setupheld the pipe securely to maintain a constant measurement depth.To ensure the integrity of the measurements, we conducted apre-test by flowing air at each target rate (25, 30, 35, 40, and 45L min− 1 ) and verifying that the flow meter readings remainedstable and accurate, confirming the absence of any leakage. Allunoccupied through-holes and nostrils were carefully pluggedand sealed during the process. 2.3 RM To obtain a clinical baseline for each subject, active anteriorRM was performed using a standard clinical rhinomanometer(MPR3100, NIHON KOHDEN). The measurements were con-ducted in accordance with established clinical guidelines andstandard operating procedures [ 25 ]. The total nasal resistance wascalculated at a reference pressure of 75 and 150 Pa. These globalresistance values served as the reference standard for validatingour 3D model-based local pressure measurements. 2.4 Outcome Parameters The primary outcome of this study was the local pressure differ-ential ( ΔP , Pa) measured at 12 anatomical locations (L1–L6 andR1–R6) to characterize regional airflow dynamics. The secondaryoutcome was the local nasal resistance ( R , Pa/cm3 /s), calculatedAdvanced Sensor Research, 2026tive Commons LicenseFIGURE 1 Detailed structure of the 3D-printed nasal cavity model. (a) CT images of five individuals recruited for the present study. For each individual, sagittal and coronal CT images are shown side by side. The blue lines indicate the plane of the coronal section, while the green lines indicate the plane of the sagittal section. The exact positions where through-holes were created for the pressure measurements are indicated by orange dots and arrows in each image. (b) Photographs of the 3D-printed nasal cavity model taken from different directions. The model was fabricated as a three-part assembly: one main body (shown in the “Bottom” and “Top” views) and two detachable small parts (shown in the two smaller photographs between the “Bottom” and “Top” views). The small images of the two parts were included to show the internal contact surfaces that interface with the main body. The labels “Bottom” and “Top” denote the inferior and superior views of the main body, respectively. The positions for inserting a sensor for the pressure differential measurements are marked with red circles and labeled as L1–L6 for the left nasal cavity and R1–R6 for the right nasal cavity. Specifically, points 1 correspond to the nasal floor, 2–4 to the olfactory cleft, 5 to the inferior meatus, and 6 to the middle meatus. u  r  e  t  T  o  g     A 27511219, 2026, 6, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/adsr.70162 by Kota Shiba - National Institute For , Wiley Online Library on [02/06/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicablesing the standard formula: R = ΔP / V , where V is the flowate (cm3 s− 1 ). Although nasal airflow is considered non-laminarven under quiet breathing conditions, this formula was used ashe conventional physiological expression of nasal patency [ 26 ].hese parameters were used to evaluate the diagnostic resolutionf our local measurement approach compared to conventionallobal rhinomanometry. dvanced Sensor Research, 20262.5 Statistical Analysis All pressure-differential data were obtained from five repeatedmeasurement cycles at each flow rate to ensure reproducibility,and the results were presented as mean ± standard deviation.To assess the uniqueness of the aerodynamic profiles for eachsubject, principal component analysis (PCA) was performed3 of 8 Creative Commons LicenseFIGURE 2 Experimental setup for measuring flow-induced pressure differential in a 3D-printed nasal cavity model. (a) Schematic and (b) photograph of the experimental setup. w  a  s  v  C33SU  s  n  T  v  i  r  a  n  s  r  d  s  p  d  a  p  l  e  n  f  o  r  u  o                                     4 27511219, 2026, 6, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/adsr.70162 by Kota Shiba - National Institute For , Wiley Online Library on [02/06/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creaith local pressure values as input variables. This statisticalpproach allowed for the objective classification of individual-pecif ic nasal structures. All statistical calculations and dataisualizations were conducted using OriginPro 2026 (OriginLaborporation).  Results and Discussion .1 Dynamic Response of Local Pressure During imulated Breathing Cycles nder the established flow conditions, the pressure differentialensor captured a regularly modulated response that synchro-ized with the simulated breathing cycles, as shown in Figure 3a .he observed alternation between positive and negative pressurealues directly corresponded to the periodic switching betweennhalation and exhalation phases. We found that the sensoresponse’s increment and decrement aligned perfectly with thepplied flow rate, although pump pulsation became more pro-ounced as the flow rate decreased due to lower pump rotationalpeeds. Regarding the temporal resolution, the sensor respondedapidly, reaching stability within several hundred millisecondsuring the initial inhalation at 45 L min− 1 and even faster inubsequent cycles (Figure 3b ). To further validate the system’serformance under more rapid breathing conditions, we con-ucted additional measurements with shorter durations of 1.0nd 1.5 s. Despite a slight delay in the flow meter’s response, theressure sensor consistently reached a plateau without significantag across all flow rates, as shown in Figure S1a . Remarkably,ven with an extremely short duration of 0.2 s, the setup yieldedearly equivalent pressure values, with the exception of theirst inhalation cycle (Figure S1b ). These results confirm thatur experimental setup is capable of capturing accurate sensoresponses under simulated breathing conditions, justifying these of a 5.0 s duration for subsequent high-precision evaluationsf the nasal cavity models. of 8With the experimental setup optimized and measurement con-ditions established, we proceeded to collect comprehensivepressure differential data from the five 3D-printed nasal cavitymodels. Our initial findings revealed position-dependent varia-tions, as shown in Figure 4a,b . To accurately evaluate the effectof sensor insertion/removal, we repeated all measurements fivetimes, disconnecting and reconnecting the sensor before eachsubsequent measurement. A specific pressure differential valuewas extracted for each site by averaging the sensor responsesat each flow rate for 5 s. Consequently, each model exhibitedunique sensor responses, reflecting the structure of the nasalcavity, as shown in Figure 1a . Compared to the right noses,sensor responses from the left noses differed significantly, indi-cating clearer structural differences among the five individuals.Considering that a higher pressure value indicates a narrowernasal cavity structure, it seemed reasonable that Ind. 3 provideda relatively high pressure compared with Inds. 1, 2, and 4. Thistrend was consistent with the CT data shown in Figure 1a , right.A comparison of the CT images of the right and left nosesalso indicated that the patency of the nasal cavity was inverselycorrelated with the sensor responses. The most interesting findinghere was that the sensor response from Ind. 5 (except for theL5 hole that was not used for the measurements owing to ahole that was too shallow and a printed wall that was toothin, which hindered stable and reproducible measurements)exceeded the upper limit even at 20 L min− 1 and barely stayedin the measurable range at 17.5 L min− 1 , although its CT imagewas not very different from the others. The anomalously highpressure observed in Ind. 5-L, despite its CT image not appearingdrastically different from the others, highlighted the sensitivityof our approach to subtle structural nuances. This sensitivityunderscores the potential application of our method to identifyfunctionally significant stenosis that might be overlooked byvisual inspection of CT scans or less localized measurementsalone. To further contextualize these individual variations andvalidate our local pressure measurements, we proceeded to con-duct comparative RM measurements using identical 3D-printedAdvanced Sensor Research, 2026tive Commons LicenseFIGURE 3 Measured pressure differential under regularly switched inhalation/exhalation cycles. (a) Change in the pressure differential as a function of the flow rate. The flow rate profile was also recorded during the pressure differential measurement and is shown above the pressure differential data. (b) Zoom-in of the graph shown in (a). The transition between inhalation and exhalation at the minimum and maximum flow rates (25 and 45 L min− 1 , respectively) was focused on to show the sensor response time. FIGURE 4 Measured pressure differential data using five 3D-printed nasal cavity models. Position-dependent change in pressure differential as a function of flow rate: (a) left nose and (b) right nose. Note that the data for Ind. 5-L were recorded only at lower flow rates (17.5 and 20 L min− 1 , measured repeatedly) than at other flow rates (25 to 45 L min− 1 , measured sequentially). The dotted line represents the upper/lower limit of the pressure differential sensor used in this study. All error bars represent standard deviation. Advanced Sensor Research, 2026 5 of 8 27511219, 2026, 6, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/adsr.70162 by Kota Shiba - National Institute For , Wiley Online Library on [02/06/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons LicenseTABLE 1 Summary of subject profiles and clinical rhinomanomet- ric data. Subject ID Age /gender RM resistance [Pa/cm3 /s] Left Right 75 Pa 150 Pa 75 Pa 150 Pa Ind. 1 40s / Male 0.376 0.521 0.638 0.816 Ind. 2 60s / Male 0.358 0.417 0.455 0.546 Ind. 3 30s / Female 0.544 0.641 0.338 0.452 Ind. 4 40s / Male 0.278 0.407 0.487 0.588 Ind. 5 40s / Female 1.985 2.675 0.620 0.795 m  e3RT  c  v  c  i  c  s  1  n  o  [F  c  m  a  i  n  i  sT  w  (  i w  (  i  a  t  R  p  c  e  l                                                  6 27511219, 2026, 6, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/adsr.70162 by Kota Shiba - National Institute For , Wiley Online Library on [02/06/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creaodels. This allowed us to correlate our localized data withstablished global nasal resistance metrics. .2 Comparison Between Model-Derived esistance and Clinical RM o validate our methodology against the clinical standard, weonducted a comparative analysis using the global resistancealues obtained from RM. The clinical characteristics and theorresponding RM baseline for the five subjects are summarizedn Table 1 . All RM measurements exhibited typical S-shapedurves, as shown in Figure 5a . Following the internationaltandards for RM [ 27 ], the pressure differential values at 75 and50 Pa were used as representative reference points to evaluateasal patency. These specific values correspond to flow conditionsften observed prior to the onset of significant turbulence effects 28 ]. ollowing the outcome parameters defined in Section 2.4 , wealculated R for both the RM data and our localized sensoreasurements. Interestingly, the R values for both showed goodgreement, especially for the right noses, regardless of thenconsistent flow rate range, as shown in Figure 5b,c . The leftoses also showed similar trends; however, as the nasal resistancencreased, a clear discrepancy emerged between the RM andensor responses. his divergence likely stemmed from differences in flow regimesithin the complex nasal geometry, requiring Reynolds numberRe) estimation to understand the underlying fluid dynamics. Res defined as Re =𝜌vL 𝜇(1)here ρ is the density of air (kg m− 3 ), v is the velocity of airflowm s− 1 ), L is the characteristic length (m; the hydraulic diameters usually used for the nasal cavity), and μ is the viscosity ofir (Pa s). Since the nasal cavity structure is very complex, theransition from laminar to turbulent flow occurs with a lowere number, even less than 1,000, compared to the transition in aipe (approximately 2300) [ 29, 30 ]. Considering that typical nasalavities have a hydraulic diameter of 1 cm or less [ 30 ], Re canasily exceed 1,000, indicating that the flow becomes more oress turbulent. To investigate this aerodynamic behavior at eachof 8measurement site, we applied the power-law fitting (see FiguresS2 and S3 for detailed plots) [ 25, 31–33 ]. The resulting fittingparameters—specifically the exponents ( b )—provided criticalinsights into the local fluid dynamics. At most measurement sites,the b values consistently approached 2.0 (ranging from 1.8 to 2.1).This near-quadratic scaling aligned with the Rohrer model, Δ𝑃 = 𝑘1 𝑣 + 𝑘2 𝑣2 (2)where the quadratic term ( k2 v2 ) represents the dominance ofinertial or turbulent losses. The high b values observed in ourstudy confirmed that the k2 term was dominant under the presentexperimental conditions. Importantly, nasal resistance coulddramatically increase in a remarkably narrow nasal structure,such as Ind. 5-L, leading to a large discrepancy in nasal resistance.These insights into flow dynamics, particularly in challengingcases like Ind. 5-L, reinforced the necessity for localized measure-ments and their improvements for a better understanding of thesecomplicated phenomena. 3.3 Subject Discrimination Using PCA Another practical merit of localized measurements was that wecould collect detailed person-specific information. Consideringthat RM provided a global nasal resistance that almost linearlyincreased in the representative measurement range of 75 and150 Pa, the RM results included limited information about themeasured persons regardless of flow rate (Figure 5d ). In contrast,the position-wise measurements reported here offered richerinformation than the RM. Even if the averaged nasal resistanceamong the six positions was similar, such as in the case of the rightnose (Figure 5c ), the independent nasal resistance values fromeach position included unique information. Taking advantageof these non-linear characteristics, similar nasal structures werediscriminated using PCA, as shown in Figure 5e ). Thus, bycollecting a large amount of data, including data from personswith a specific nasal disease, the present approach could lead tofacile classification based on nasal conditions. 4 Conclusion We developed a facile, pressure-based approach to evaluatelocalized airflow dynamics within the nasal cavity using person-specif ic 3D-printed models. This method directly addressed thecritical limitation of conventional RM, its inability to provide localstructural details, by offering position-dependent pressure values.We demonstrated that nasal resistance estimated using ourapproach closely correlated with global resistance obtained usingrhinomanometric measurements, thereby validating its utilitywhile significantly enhancing the diagnostic resolution. Giventhat RM is used as a worldwide standard technique to evaluatenasal flow that reflects nasal conditions, such as the presenceof obstructions or physical abnormalities, the present approachoffered equivalent information with an easy and reproducibleoperation. Compared with an RM result that simply connectedwith the global resistance of the nasal cavity, the present approachprovided an additional aspect of acquiring position-dependentpressure values, which allowed us to discuss the correlationbetween nasal flow and our olfactory ability in more detail. TheAdvanced Sensor Research, 2026tive Commons LicenseFIGURE 5 Advantage of the present approach validated using conventional RM. (a) RM results measured using five 3D-printed nasal cavity models. (b, c) Nasal resistance-based comparison: (b) RM results obtained at 75 and 150 Pa, and (c) experimental data collected from the six positions. (d) Relationship between nasal resistance obtained at 75 and 150 Pa. The linear fitting results were also shown as dotted lines. (e) Principal component analysis based on the data collected from the right nostrils of the five individuals. u  e  s  fAT  R  e  gFT  J     A 27511219, 2026, 6, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/adsr.70162 by Kota Shiba - National Institute For , Wiley Online Library on [02/06/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Crease of CFD and/or other relevant techniques will help furtherxtend the possibility of applying the present approach in clinicalituations, olfactory sensing, and understanding complex fluidlow. cknowledgments his work was financially supported by a Grant-in-Aid for Scientificesearch (B), JSPS, MEXT, Japan (No. 24K01520). The authors acknowl-dge the use of Gemini 3 (Google AI) for the conceptual design and partialeneration of the Table of Contents figure. unding his work was supported by a Grant-in-Aid for Scientific Research (B),SPS, MEXT, Japan (No. 24K01520)., dvanced Sensor Research, 2026Ethics Statement This study was approved by the Ethics Committee of the Jikei UniversitySchool of Medicine (Approval No. 34–250[11402]). Informed consent wasobtained from all participants prior to their inclusion in the study. Conflicts of Interest The authors declare no conflicts of interest. Data Availability Statement The data that support the findings of this study are available from thecorresponding authors upon reasonable request. References 1 . E. H. C. Wong and R. Eccles, “Comparison of the Classic and BromsMethods of Rhinomanometry Using Model Noses,” European Archives ofOto-Rhino-Laryngology 272 (2015): 105–110. 7 of 8tive Commons License2  R  53  R4  N5  b  E6  B  D  A  S7  R  t  J8  n  D  S9  A  e1  “  (1  A  P1  r  a1  p  S  T1  S  s  M1  i  t1  R  J1  t  R1  i1  “  FP2  D  S  N                      8 27511219, 2026, 6, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/adsr.70162 by Kota Shiba - National Institute For , Wiley Online Library on [02/06/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable  . E. H. C. Wong and R. Eccles, “Comparison of Classic and 4-Phasehinomanometry Methods, Is There Any Difference?,”Rhinology Journal2 (2014): 360–365.  . H. L. Thulesius, A. Cervin, and M. Jessen, “Can We Always Trusthinomanometry?,” Rhinology Journal 49 (2011): 46–52.  . R. Eccles, “A Guide to Practical Aspects of Measurement of humanasal Airflow by Rhinomanometry,” Rhinology Journal 49 (2011): 2–10.  . M. Atzori, G. D. Ciacci, and M. Quadrio, “Understanding the Mismatchetween In-Vivo and In-Silico Rhinomanometry,” Medical & Biologicalngineering & Computing 64 (2025): 219–229.  . N. Schmidt, H. Behrbohm, L. Goubergrits, T. Hildebrandt, and J.rüning, “Comparison of Rhinomanometric and Computational Fluidynamic Assessment of Nasal Resistance with Respect to Measurementccuracy,” International Journal of Computer Assisted Radiology andurgery 17 (2022): 1519–1529.  . G. B. Cherobin, R. L. Voegels, F. R. Pinna, E. M. M. S. Gebrim,. S. Bailey, and G. J. M. Garcia, “Rhinomanometry versus Computa-ional Fluid Dynamics: Correlated, but Different Techniques,” Americanournal of Rhinology & Allergy 35 (2021): 245–255.  . M. Berger, A. I. Giotakis, M. Pillei, et al., “Agreement between Rhi-omanometry and Computed Tomography-Based Computational Fluidynamics,” International Journal of Computer Assisted Radiology andurgery 16 (2021): 629–638.  . N. Patil and S. Jain, “Rhinomanometry: a Comprehensive Review of Itspplications and Advancements in Rhinology Practice,”Cureus 16 (2024):61370. 0 . K. Naito, S. Horibe, Y. Tanabe, H. Kato, S. Yoshioka, and I. Tateya,Objective Assessment of Nasal Obstruction,” Fujita Medical Journal 92023): 53–64. 1 . C. Li, A. A. Farag, G. Maza, et al., “Investigation of the Abnormal Nasalerodynamics and Trigeminal Functions among Empty Nose Syndromeatients,” International Forum of Allergy & Rhinology 8 (2018): 444–452. 2 . A. Dayal, J. S. Rhee, and G. J. M. Garcia, “Impact of Middle versus Infe-ior Total Turbinectomy on Nasal Aerodynamics,” Otolaryngology–Headnd Neck Surgery 155 (2016): 518–525. 3 . G. L. Liu, W. J. Martin, Y. Mirmozaffari, R. Ni, and Z. Li, “Com-utational Modeling of Nasal Cavity Aerodynamics: Implications forurgical Outcomes and Targeted Drug Administration,” Ear, Nose, &hroat Journal 9 (2025): 1455613251335109. 4 . H. Fatahi, A. Dastan, S. Sadrizadeh, and O. Abouali, “Numericaltudy of Nasal Hair Effects on Breathing Comfort and Particle Depo-ition in a Simplified Vestibule Region,” Biomechanics and Modeling inechanobiology 24 (2025): 1513–1533. 5 . K. Bradshaw, P. Warfield-McAlpine, S. Vahaji, et al., “New Insightsnto the Breathing Physiology from Transient respiratory Nasal Simula-ion,” Physics of Fluids 34 (2022): 14. 6 . A. A. T. Borojeni, G. J. M. Garcia, M. G. Moghaddam, et al., “Normativeanges of Nasal Airflow Variables in Healthy Adults,” Internationalournal of Computer Assisted Radiology and Surgery 15 (2020): 87–98. 7 . K. Zhao and J. B. Jiang, “What Is Normal Nasal Airflow? A Compu-ational Study of 22 Healthy Adults,” International Forum of Allergy &hinology 4 (2014): 435–446. 8 . P. Niegodajew, “Flow Patterns and Vortex Formation Mechanismsnside a Human Nasal Cavity,” Physics of Fluids 37 (2025): 19. 9 . G. B. Cherobin, R. L. Voegels, E. M. M. S. Gebrim, and G. J. M. Garcia,Sensitivity of Nasal Airflow Variables Computed via Computationalluid Dynamics to the Computed Tomography Segmentation Threshold,”LoS ONE 13 (2018): e0207178. 0 . S. Basu, D. O. Frank-Ito, and J. S. Kimbell, “On Computational Fluidynamics Models for Sinonasal Drug Transport: Relevance of Nozzleubtraction and Nasal Vestibular Dilation,” International Journal forumerical Methods in Biomedical Engineering 34 (2018): e2946. of 821 . S. G. Johnsen, “Computational Rhinology: Unraveling Discrepanciesbetween In Silico and In Vivo Nasal Airflow Assessments for EnhancedClinical Decision Support,” Bioengineering-Basel 11 (2024): 82. 22 . C. Li, J. Jiang, H. Dong, and K. Zhao, “Computational Modelingand Validation of Human Nasal Airflow under Various BreathingConditions,” Journal of Biomechanics 64 (2017): 59–68. 23 . R. H. J. Hebbink, B. J. Wessels, R. Hagmeijer, and K. Jain, “Com-putational Analysis of human Upper Airway Aerodynamics,” Medical &Biological Engineering & Computing 61 (2023): 541–553. 24 . J. Van Strien, K. Shrestha, S. Gabriel, et al., “Pressure Distribution andFlow Dynamics in a Nasal Airway Using a Scale Resolving Simulation,”Physics of Fluids 33 (2021): 14. 25 . P. A. Clement and F. Gordts, “Consensus Report on AcousticRhinometry and Rhinomanometry,” Rhinology 43 (2005): 169–179. 26 . K. Naito, S. Horibe, Y. Tanabe, H. Kato, S. Yoshioka, and I. Tateya,“Objective Assessment of Nasal Obstruction,” Fujita Medical Journal 9(2023): 53–64. 27 . P. A. Clement, “Committee Report on Standardization of Rhino-manometry,” Rhinology 22 (1984): 151–155. 28 . J. R. Wheatley, T. C. Amis, and L. A. Engel, “Nasal and Oral AirwayPressure-Flow Relationships,” Journal of Applied Physiology 71 (1991):2317–2324. 29 . D. J. Doorly, D. J. Taylor, and R. C. Schroter, “Mechanics of Airflowin the human Nasal Airways,” Respiratory Physiology & Neurobiology 163(2008): 100–110. 30 . K. Keyhani, P. W. Scherer, and M. M. Mozell, “Numerical Simulationof Airflow in the Human Nasal Cavity,” Journal of BiomechanicalEngineering 117 (1995): 429–441. 31 . P. Broms, B. Jonson, and C. J. Lamm, “Rhinomanometry. Rhino-manometry. II. A System for Numerical Description of Nasal AirwayResistance,” Acta Oto-Laryngologica 94 (1982): 157–168. 32 . D. W. Cockcroft, D. W. MacCormack, S. M. Tarlo, F. E. Hargreave, andL. D. Pengelly, “Nasal Airway Inspiratory Resistance,” American Reviewof Respiratory Disease 119 (1979): 921–926. 33 . J. Pfitzner, “Poiseuille and His Law,” Anaesthesia 31 (1976): 273–275. Supporting Information Additional supporting information can be found online in the SupportingInformation section. Supporting File : adsr70162-sup-0001-SuppMat.docx. Advanced Sensor Research, 2026Creative Commons License Breathing-Induced Airflow Measurements at Multiple Positions in a 3D-Printed Nasal Cavity 1 | Introduction 2 | Methods and Materials 2.1 | Reconstruction and 3D Printing of Nasal Cavity Models 2.2 | Measurement System and Procedure 2.3 | RM 2.4 | Outcome Parameters 2.5 | Statistical Analysis 3 | Results and Discussion 3.1 | Dynamic Response of Local Pressure During Simulated Breathing Cycles 3.2 | Comparison Between Model-Derived Resistance and Clinical RM 3.3 | Subject Discrimination Using PCA 4 | Conclusion Acknowledgments Funding Ethics Statement Conflicts of Interest Data Availability Statement References Supporting Information