Speaker's Highlight

• Don Sin University of British Columbia, St. Paul Hospital (Canada)
  Kenneth R. Chapman Toronto General Hospital Research Institute (Canada)
• Parameswaran Nair McMaster University (Canada)
  Carolyn Calfee UCSF (U.S.A.)
• Gregory P. Downey University of Colorado School of Medicine (U.S.A.)
  David A. Schwartz University of Colorado School of Medicine (U.S.A.)
• Neil Schluger Tuberculosis Control Branch, California Department of Public Health (U.S.A.)
  Nick Kim Critical Care & Sleep Medicine, University of California San Diego (U.S.A.)
• Nicola Hananiah Baylor College of Medicine (U.S.A.)
  Jae-Joon Yim Seoul National University College of Medicine (Republic of Korea)
• Koichiro Asano Tokai University School of Medicine (Japan)
  Diahn-Warng Perng Taipei Veterans General Hospital (Taiwan)
• Konstantinos Kostikas University of Ioannina (Greece)
  Karin Klooster University Medical Center Groningen (Kingdom of the Netherlands)

• Computational Modeling of Airflow in Asthmatic Airways

  Asthma remains one of the most common chronic respiratory conditions in Australia, where more than two million adults and children live with the diagnosis. Behind every wheeze, cough, or tight chest lies a physical reality: airways that narrow, swell, and remodel in patterns unique to each patient. Computational fluid dynamics now offers clinicians a way to look inside those airways without a scalpel, turning medical imaging into a numerical laboratory where airflow can be measured, visualised, and predicted.

  The field has moved well beyond textbook diagrams. Patient-specific models can reconstruct a single trachea down to the eighth bronchial generation and simulate how a puff from a salbutamol puffer might reopen a constricted segment. For respiratory physicians preparing for the APSR 2022 congress, the topic sits at a meeting point between engineering, physiology, and personalised medicine, and it is rapidly becoming one of the most discussed themes in pulmonary research.

  The Mechanics Behind Airflow Simulation

  Computational fluid dynamics, often shortened to CFD, treats inhaled air as a moving fluid governed by the Navier-Stokes equations. In an asthmatic lung, these equations become harder to solve because the geometry is irregular and the walls are not rigid. Bronchoconstriction, mucosal oedema, and dynamic airway collapse all change local resistance in ways a uniform tube model cannot capture.

  Modern solvers tackle this with high-resolution meshes of millions of polyhedral elements that follow the curvature of each branch. Turbulence models such as k-omega SST or large eddy simulation are applied when velocities exceed the laminar threshold during a forced expiratory manoeuvre. The result is a colour map of pressure, velocity, and wall shear stress that a clinician can read almost like a weather chart of the lung.

  Why Airflow Patterns Matter in Asthma

  Asthma is rarely a uniform disease. Some patients have predominantly large-airway obstruction, while others show patchy small-airway involvement that only appears on sophisticated lung function tests. Computational models help distinguish these phenotypes by quantifying where the resistance actually lives, branch by branch.

  Wall shear stress has emerged as a biomarker of interest. Low shear at the airway surface may encourage remodelling, while elevated shear can trigger the release of inflammatory mediators. Simulating an inhalation lets researchers pinpoint hot spots of abnormal shear and correlate them with bronchoscopic findings, imaging features, or sputum eosinophil counts. For clinicians in Australian tertiary centres such as the Royal Melbourne Hospital or the Alfred, this mechanistic insight is starting to influence decisions about biologic therapy.

  Building a Patient-Specific Model

  The workflow typically starts with a high-resolution CT scan acquired during a breath-hold. Segmentation software, often built on deep learning architectures, isolates the airway tree from the surrounding parenchyma. Clean segmentation remains the foundation of any reliable simulation and one of its most time-consuming steps.

  Once the geometry is ready, it is meshed and imported into a solver such as ANSYS Fluent, OpenFOAM, or SimVascular. Boundary conditions mimic either quiet tidal breathing or a forced expiratory effort, depending on the clinical question. Outputs include airway resistance, particle deposition patterns for inhaled corticosteroids, and pressure gradients across individual segments. Researchers at the Woolcock Institute of Medical Research in Sydney have been refining these pipelines for several years, often linking them to severe asthma registries across New South Wales.

  Clinical Applications in Australian Practice

  In practice, computational modelling is shifting from a research curiosity to a decision-support tool. Severe asthma clinics are beginning to use airway flow simulations alongside spirometry and FeNO testing to decide whether a patient is more likely to respond to a T2-biologic or to a small-airway-targeted intervention. The Pharmaceutical Benefits Scheme subsidises several biologic agents, and getting the right one to the right patient has measurable cost implications, particularly when scripts are repeated monthly at the local GP.

  Bushfire smoke events along the east coast, including the 2019-2020 Black Summer fires, have sharpened interest in airway physiology. Modelling how fine particulates deposit in already-inflamed airways helps explain why some patients deteriorate weeks after the air has cleared. Public health messaging in Sydney and Brisbane now routinely references these mechanisms in plain language that families can act on.

  Emerging Methods and AI Integration

  Machine learning is reshaping the field from two directions. Surrogate models trained on thousands of CFD simulations can predict pressure drops in milliseconds, allowing near-real-time feedback during a clinic visit. Convolutional networks can read a CT scan and infer flow characteristics without ever solving the Navier-Stokes equations directly. Hybrid approaches, where physics-informed neural networks blend both worlds, are gaining traction at conferences and in journals.

  For clinicians and engineers thinking about presenting their own work, the speaker instructions for the upcoming congress offer a useful starting point, especially for those moving into hybrid poster-presentation formats on tight timelines.

  Choosing the Right Modelling Approach

  There is no single correct method for every question. Lumped-parameter models give quick answers about global resistance but miss branch-level detail. Three-dimensional CFD delivers rich spatial information but demands hours of compute time and skilled operators. One-dimensional tree models sit in between and are popular for studying ventilation heterogeneity across many generations.

  Choosing among them usually comes down to the clinical question, available imaging, and local expertise. A team with strong radiology but limited HPC access will get further with one-dimensional models than with full turbulent 3D simulation.

  Approach	Spatial Detail	Compute Time	Best Use Case
  Lumped-parameter	Whole lung	Seconds	Screening and global resistance
  1D tree model	Branch generation order	Minutes	Heterogeneity and deposition
  3D CFD (steady)	Local segments	Hours	Branch-level pressure and shear
  3D CFD (transient)	Local segments	Days	Wheeze mechanics and dynamic collapse

  When the goal is to compare two biologics in a single severe-asthma patient, a targeted 3D steady-state simulation is often enough. When the goal is to model wheeze sounds across a full respiratory cycle, transient CFD or coupled fluid-structure interaction becomes necessary. Knowing the question first prevents over-engineering and keeps the project within an ordinary research budget.

  Practical Recommendations for Clinicians and Researchers

  • Start with a clear clinical question before selecting any modelling method.
  • Pair every computational output with conventional lung function tests for validation.
  • Use reproducible segmentation pipelines built on open-source tools such as 3D Slicer.
  • Engage biomedical engineers early; Australian universities including the University of Melbourne and UNSW run embedded clinical-engineering collaborations.
  • Keep patient data governance central, particularly under the Australian Privacy Principles.
  • Plan validation studies alongside model development rather than as an afterthought.
  • Communicate uncertainty honestly, especially when simulation outputs may guide medication changes.

  What stays with the reader after all of this is the sense that airflow inside an asthmatic lung is no longer a black box. From the imaging suites of the Woolcock to severe asthma clinics in Melbourne, the tools now exist to map individual airways, test ideas without harming patients, and bring physics-informed reasoning into a field that has long relied on indirect measurements.

  Richard Russell Nuffield Department of Clinical Medicine, University of Oxford (United Kingdom)
• Mona Bafadhel King’s College London (United Kingdom)
  David Jackson Guy’s and St Thomas’ Hospital, King’s College London (United Kingdom)
• James Chalmers University of Dundee (United Kingdom)
  David Price University of Aberdeen (United Kingdom)