Workshop 3: The Lung and the Respiratory (Structure, Oxygen Transport)

(November 6,2006 - November 10,2006 )

Organizers


Jason Bates
Medicine; Physiology; Biophysics, University of Vermont
Ken Lutchen
Department of Biomedical Engineering, Boston University
Bela Suki
Biomedical Engineering, Boston University

Computational modeling promises a new era in the fundamental understandings of how lung morphometry and biomechanical/biomaterial properties impact lung function. With continuous improvement in imaging modalities, it is becoming increasingly possible to establish precise physical locations and degrees of structural or functional defects in the lung during disease. Such data will beg the question of how explicit defects of biological components, processes, and structure at specific anatomic locations alter function. Computational power now permits one to develop models that are closer anatomic replicas of a real lung, while incorporating the fundamental biophysical properties and relations for each exquisite component of each airway. Rational and efficient disease management can be enhanced by understanding or predicting how alterations in the individual components of lung structure and properties impact the emergent lung function.

The workshop aims to:

  • indentify critical current questions in asthma, emphysema, and respiratory distress syndrome that can be addressed via mathematical modeling integrated across multiple scales;
  • identify modeling challenges associated with understanding fundamental biology and physiology of the constituent parts of the lung ranging from parencymal, alveolar and airway smooth muscle cells, to airways and alveoli and microvessels in the lung, to lung tissue;
  • identify role imaging can play to advance models of micro and macro structure and to link structure to function via computational models;
  • understand how emergent function and dysfunction in the lung for these diseases relates back to its constituent parts.

Accepted Speakers

Joe Anderson
Bioengineering, University of Washington
Ananth Annapragada
Health Information Sciences, University of Texas
Jason Bates
Medicine; Physiology; Biophysics, University of Vermont
James Butler
Physiology, Environmental Health, Harvard University
Chantal Darquenne
Dept. of Medicine, University of California, San Diego
Ben Fabry
Physical Medical Technology, Universität Erlangen-Nürnberg
Jeff Fredberg
Bioengineering and Physiology, Harvard University
Steve George
Biomedical Engineering, University of California, Irvine
Samir Ghadiala
Mechanical Engineering, Lehigh University
Susan Gunst
Physiology and Biophysics, Indiana University--Purdue University
Eric Hoffman
Physiologic Imaging, Radiology, University of Iowa
Hiroko Kitaoka
Graduate School of Medicine, Osaka University
Ching-Long Lin
Mechanical and Industrial Engineering, University of Iowa
Ken Lutchen
Department of Biomedical Engineering, Boston University
Geof Macksym
Biomedical Engineering, Dalhousie University
Brent McParland
Pharmacology, University of Sydney
Srboljub Mijailovich
Environmental Health, Harvard University
Wayne Mitzner
Physiology, Johns Hopkins University
Kim Prisk
Medicine, University of California, San Diego
Michael Sanderson
Physiology, University of Massachusetts
Brett Simon
Medicine, Anesthesiology, Johns Hopkins University
Dimitrije Stamenovic
Biomedical Engineering, Boston University
Akira Tsuda
Environmental Health, Physiology, Harvard Medical School
Jose Vengas
Biomedical Engineering, Massachusetts General Hospital
Sylvia Verbanck
Academic Hosptical (AZ VUB), Vrije Universiteit Brussel
Monday, November 6, 2006
Time Session
Tuesday, November 7, 2006
Time Session
09:00 AM
09:20 AM
Ching-Long Lin - Multi-scale Simulation of Pulmonary Air Flow

In this presentation I will talk about the progress in developing a comprehensive computational fluid dynamic (CFD) model for pulmonary air flow which utilizes subject-specific airway geometries and employs a Computed Tomography (CT) data-driven, multi-stage approach to provide accurate predictions of regional ventilation and gas transport through the entire moving airway tree. The model is based up an in-house parallel characteristic Galerkin fractional four-step finite element method. Both direct numerical simulation (DNS) and large-eddy simulation (LES) techniques have been employed to simulate turbulent flow in the proximal conducting airways. A three-stage approach has been adopted to simulate air flow in the acinar airways. Due to the large mesh size, some CFD simulations have been carried on TeraGrid clusters. The effects of boundary conditions and subject-specific human airways on air flow have been assessed. The transient air flow in a CT-based monopodial sheep (ovine) tracheobronchial tree of up to 13 generations has also been studied numerically. Currently, a geometric model of 11 generations of human airways has been constructed and the preliminary CFD simulation has been successfully performed. The progress in developing an in-house mesh generation software which aims to create CFD mesh from the trachea to the terminal bronchioles will be reported. Toward modeling a breathing lung, the host-mesh technique in conjunction with 4D dynamic imaging, the model for soft tissue mechanics, and the fluid-structure interaction (FSI) technique have been or are being developed, which will be briefly discussed as well.



 
09:20 AM
09:40 AM
Jason Bates - Modeling emergent behavior

The lung consists of myriad disparate components that interact in a rich variety of ways. Accordingly, a quantitative understanding of how the lung functions can be pursued at many levels of scale, from that of the molecule up to the whole organ. Importantly, however, a complete understanding of lung function at the organ level does not follow automatically from knowledge of the individual characteristics of its components. Indeed, understanding the components in isolation may not even constitute the major requirement for an understanding of the lung as a whole. This is because the lung is a complex system - something with emergent properties that arise from the way its components operate as an ensemble. Emergent properties in a complex system typically bear little qualitative relationship to the individual properties of the system components, yet somehow arise from these components and their interactions in frequently subtle ways. Prediction of emergence in complex systems is well known to physicists, as it constitutes the central paradigm of statistical mechanics that began with the kinetic theory of gases more than a century ago. Emergence has been formally embraced by the biomedical community only more recently, but it impacts the life of every scientist, whether they know it or not, through the Central Limit Theorem which provides a theoretical basis for the widespread appearance of the Gaussian distribution throughout the natural world. Even the ubiquity of power law processes in nature, which is now exciting a great deal of interest in biology and physics and makes its appearance in the lung, likely represents an emergent phenomenon. Understanding the genesis of emergent behavior, however, is frequently not straightforward. Highly nonlinear interactions between system components are typical of complex biological systems, which severely limits the use of analytical methods for predicting the emergent properties arising there from. The exploration of emergent behavior in the lung is thus going to be based on computational models based constructs such as cellular automata, artificial network networks, percolation networks, and distributed recruitment mechanisms. As specific examples, we will consider how percolation and recruitment mechanisms might be invoked to account for the bulk mechanical properties of lung tissue.



 
10:40 AM
11:00 AM
Brett Simon - Clinical Perspectives: Modeling in Acute Lung Injury

Acute lung injury (ALI) and the adult respiratory distress syndrome (ARDS) are syndromes of respiratory failure with mortality in excess of 30% affecting an estimated 25-50 patients per 100,000 per year worldwide. While there are a variety of initiating events, the final common pathway is characterized by hypoxemic respiratory failure with a heterogeneous loss of aerated lung volume and reduced compliance due to flooding and collapse, surfactant inactivation, increased dead space, and inflammation. Clinical management is currently limited to supportive measures, primarily low tidal volume mechanical ventilation with positive end-expiratory pressure (PEEP), in an attempt to recruit the poorly aerated lung while reducing injurious cyclic end-expiratory airspace opening and closing and end-inspiratory regional overdistension. Many other therapeutic approaches, including pharmacologic therapy to reduce edema formation or inflammation or improve perfusion distribution, exogenous surfactant treatment, and novel mechanical ventilation strategies, have been tried or are in development but none have yet been shown to improve outcome.


Modeling of ALI/ARDS pathophysiology to date has been focused on the quasi-static pressure-volume properties of the injured lungs, primarily in an attempt to predict recruitment and optimal mechanical ventilation strategies, but these approaches have had limited success, possibly because they do not adequately reflect the very different dynamic operating conditions of the injured lung. Opportunities for modeling to contribute to improved understanding of ALI pathophysiology and management abound. Better characterization of the dynamic mechanical properties of the lung, particularly the distributed regional contributions, non-linear effects over the tidal breathing cycle, and a focus on heterogeneity may help optimize mechanical ventilation management strategies. Next, the consequences of these interventions on ventilation and perfusions distributions and matching, dead space, and vascular regulation are important in translating mechanics to improved physiology and effective gas exchange. Novel therapies such as high-frequency oscillatory ventilation, airway pressure-release ventilation (APRV), and "noisy" ventilation need to be understood in terms of their mechanisms of gas exchange and effects on lung mechanics and V/Q. ALI/ARDS represents an area in which there is an opportunity for model-based advances to very rapidly achieve translation to the clinical arena and thus have great potential to improve patient care and outcomes.

11:00 AM
11:20 AM
Steve George - An Integrative Approach Towards Understanding Nitrogen Oxide Biology in the Lungs

Asthma is a chronic disease that affects approximately 10% of the population in the United States. Although our understanding of underlying mechanisms (e.g., Th2 type inflammation) has continued to increase over the past several decades, the prevalence continues to rise. Furthermore, the mainstay of monitoring (spirometry) and therapy (anti-inflammation and -agonist bronchodilation) are essentially unchanged over the same time period. Over the past ten years, the interest in using exhaled nitric oxide as a non-invasive marker of inflammation has steadily increased despite a limited understanding of the cell and molecular mechanisms. More recently, the loss of naturally occurring NO-based bronchodilators (S-nitrosothiols or SNOs) in an animal model of asthma has sparked renewed interest in the key role of NO biochemistry in modulating asthma. Our group combines advanced primary cell culture techniques, direct measurement of gas phase NO release, quantitation of enzymatic activity, exhaled nitric oxide from human subjects, and mathematical models in an integrative fashion to understand how multiscale systems interact to influence NO storage and release in the lungs. We hope this approach will significantly improve our mechanistic understanding of NO metabolism and gas phase release, and thus non-invasive inflammatory monitoring of asthma.



 
01:30 PM
01:50 PM
Samir Ghadiala - Fluid-Structure Modeling of Cellular Deformation and Injury in Pulmonary Airways

Several respiratory disorders including infant RDS, adult RDS and acute lung injury, are characterized by the fluid-occlusion of small pulmonary airways and alveoli. The reopening of these fluid-filled structures involves the movement of air-liquid interfaces and the application of complex hydrodynamic and surface tension forces on the epithelial cells which line airway and alveolar walls. Although these fluid mechanical forces may contribute to ventilator-induced lung injury, the mechanism of cellular injury during these free-surface flows is not well established. Under these conditions, the amount of cellular deformation is not known a priori and may depend on several multi-scale factors including a) reopening dynamics, b) surfactant transport to the air-liquid interface, c) morphological and/or micro-mechanical properties of the epithelial cells and d) molecular interactions at focal adhesion sites. In addition, biochemical responses (i.e. protein/gene expression) to these different hydrodynamic conditions may also contribute to lung injury. Our laboratory utilizes a combination of in-vitro cell culture experiments, advanced imaging techniques and computational models to investigate the relative importance of these different factors. The goal is to obtain a better understanding of the biophysical and biological mechanisms of cell/tissue injury during the reopening of fluid-filled airways/alveoli.


In this presentation we will first present a brief overview of previous studies which have elucidated the multiphase fluid mechanics of airway/alveolar reopening. We will then present recent experimental data which demonstrate how changes in reopening dynamics, cellular morphology, cell mechanics and cytoskeletal structure influence cell injury during airway reopening. We will also discuss the development of multi-scale, computational models of cellular deformation during airway reopening. In addition to providing valuable insight into the mechanism of cell injury, these computational models have been used to explain several counter-intuitive experimental results including increased cell death at slow reopening velocities and decreased death in cells with a lower elastic modulus (i.e. a more flexible cell). Finally, we will discuss future challenges in modeling cellular responses to complex hydrodynamic and surface tension forces including the incorporation of realistic airway/cellular geometries, non-linear membrane mechanics, heterogeneous viscoelastic cell properties and molecular interactions at focal adhesion sites.

01:50 PM
02:10 PM
Wayne Mitzner - Modeling airway distensibility in the lung

The lung is an inflatable organ that must incorporate three separate tree structures. Thus, in order for the lung to inflate, all of these trees must stretch both axially and radially. Indeed, it was noted almost a century ago that if the airway tree could not lengthen, then the lung likely could not inflate. In this presentation we will discuss the first distensibility of airways, and then how this distensibility interacts with that of the surrounding lung parenchyma.


In modeling the lung an implicit assumption is often made that airway diameter varies as the cube root of lung volume. While there has been some published evidence to support this under certain conditions, the assumption is not supported by what is known about the airway structure. The role of the relatively nondistensible basement membrane is generally ignored. However, the fact that the length of this membrane is relatively fixed has been used histologically to compare airways of comparable size in asthmatic and normal subjects. So if the basement membrane is rigid, how can the airway diameter vary as the cube root of lung volume? Either this fixed length is larger than the largest size the airway ever achieves, or the assumption is wrong. We will show that the latter reason is the correct one.


We will show that pressure-area curves of relaxed airways in vivo are highly nonlinear, reaching a maximal size at a very low pressure (less than 10 cmH2O). The limit is consistent with a rigid basement membrane. With smooth muscle tone, the relation between airway area and pressure becomes highly variable depending on the level of tone and other uncertain factors. This variability in tone even under baseline conditions is what often causes misinterpretation of the nature of the distensibility. All dogs we have studied have a variable degree of baseline tone, averaging from 50-70% of maximally relaxed size. Although most clinical research suggests that there is little baseline tone in normal healthy subjects, this is not true. With sufficient relaxation of the airway smooth muscle, we can show that even normal humans also live with airways at about 705 of maximal size.


These results stress the importance of understanding the link between airway structure and their functional distensibility. The way airways respond to inflation may have important clinical manifestations in the asthmatic pathology.

03:10 PM
03:30 PM
Akira Tsuda - Folding and mixing

N/A

03:30 PM
03:50 PM
Wednesday, November 8, 2006
Time Session
09:00 AM
09:20 AM
Jeff Fredberg - Cytoskeletal remodeling and slow dynamics in the living cell: Are we built of glass?

With every beat of the heart, inflation of the lung, or peristalsis of the gut, cell types of diverse function are subjected to substantial stretch. New data show that cell responses to a transient stretch exhibit remarkable physical similarities to fluidization observed in jammed inert matter, including colloids, pastes, emulsions, and foams, and thus implicate mechanisms mediated not only by specific signaling intermediates, as is usually presumed, but also by nonspecific actions of a slowly evolving network of physical forces. These results support the idea that the cell interior is at once a crowded chemical space and a fragile soft material in which the effects of biochemistry, molecular crowding, and physical forces are complex and inseparable, yet conspire nonetheless to yield remarkably simple phenomenological laws. These laws appear to be universal and thus comprise a striking point of contact between the worlds of cell signaling biology and soft matter physics.

09:20 AM
09:40 AM
Ben Fabry - Nonlinear viscoelasticity of living cells

The linear viscoelastic behavior of adherent cells and biological tissue is characterized by a wide distribution of relaxation times and shows a power-law creep response or a power-law viscoelastic spectrum over several decades of time or frequency. In addition, single cells and tissue exhibit a highly nonlinear stress-strain relationship.  The viscoelastic behavior of cells in the non-linear regime is unknown, however, but is of particular interest to test different conflicting theories.  At high stress, soft glassy rheology for example predicts a speed-up of relaxation processes due to yielding, whereas the theory of semiflexible polymer networks predicts a slow-down due to reduced thermal bending fluctuations in low-frequency modes.  We measured the viscoelastic behavior of different cell types (fibroblasts, epithelial cells, endothelial cells, lung and breast carcinoma cells, vinculin mutant cells) in the linear and non-linear regime using magnetic tweezers with real-time force feedback.  We imposed a staircase-like sequence of 1 nN force steps up to a maximum force of 10 nN onto 4.5 µm fibronectin-coated magnetic beads bound to the cytoskeleton via integrins. For each stress level s, the differential creep response of single cells followed a power law: J(s,t) = J0(s) ta(s), however the differential creep modulus J0(s) decreased with stress, equivalent to a linear stress stiffening. The power-law creep exponent a showed stress dependence that was cell type specific.  We observed both increasing and decreasing exponents with increasing force. Stress reversal and repeated measurement on the same cell showed that creep recovery was incomplete and the creep modulus approached a steady state after several cycles. This behavior resembles pre-conditioning in tissue and could possibly arise from force-induced permanent structural changes in the cytoskeleton. In summary, living cells exhibit a nonlinear viscoelastic behavior remarkably similar to connective tissue. 

10:40 AM
11:00 AM
Hiroko Kitaoka - Morphogenesis-based 4D model of the alveolar structure

There has been no consensus how the alveolar shape changes during breathing due to difficulties in observing directly alveoli in vivo. A computational model of dynamical 3D model, that can be called a 4D model, would be useful for understanding such an invisible phenomenon. The living organ changes its shape during morphogenetic process as well as during physiologic motion. Therefore, an algorithm for constructing a 3D model of the organ should be consistent with its morphogenetic process, too. In other words, the fourth dimension of the organ model should possess both of physiological and morphogenetic time scales. The alveolar morphogenesis is roughly classified into four steps from a geometric point of view: formation of a fluid pathway, enlargement of intra-acinar pathway, bellows-like arrangement of primary septa by capillary invasions, and growth of secondary septa whose edges form alveolar mouths containing abundant elastin fibers. Those steps are accompanied with functional maturation: making a duct transport system, utilize full space in the organ, increasing surface area, and finally acquiring deformability. We have constructed a 4D alveolar model according to the above morphogenetic process, where the alveolar deformation is modeled by a combination of spring (elastin fibers at alveolar mouths) and hinge (septal junctions). The model includes a hypothesis that alveolar mouths are closed at minimum volume and that closed alveoli are stabilized by the alveolar lining liquid film containing surfactant. The validity of the model has been verified not fully but satisfactorily by comparing it with real experimental results; in vivo microscopic observation of subpleural alveoli (courtesy of Nieman, GF), histologic sections of rapidly frozen lungs (in literatures), and morphometric investigations by light scattering and intravascular fixation-dehydration method (in literatures).



 
11:00 AM
11:20 AM
11:00 AM
11:20 AM
01:30 PM
01:50 PM
Eric Hoffman - Lung structure from imaging

N/A

01:50 PM
02:10 PM
Michael Sanderson - Video imaging of parenchymal mechanics

N/A

03:10 PM
03:30 PM
Joe Anderson - Impact of Airway Gas Exchange on the Multiple Inert Gas Elimination Technique

The multiple inert gas elimination technique (MIGET) provides a method for estimating the efficiency of alveolar gas exchange. Six soluble inert gases with blood solubility spanning six orders of magnitude are infused into a peripheral vein of an animal. After gas equilibration, the exhaled, venous, and arterial partial pressures of these six gases are measured and interpreted by using a mathematical model of alveolar gas exchange. MIGET has been very successful in improving our understanding of alveolar gas exchange in both health and disease.


Developed in the early 1970s, MIGET assumed that none of the six chosen gases exchanged with the airways. However, this assumption is no longer considered to be valid. Over the past 20 years, our experimental and theoretical investigations on airway gas exchange have shown that the location of gas exchange (airway versus alveoli) depends primarily on gas solubility in both blood and water (as measured by a partition coefficient). Gases with both blood-air and water-air partition coefficients greater than 10 exchange primarily in the airways. Meeting this criteria, two of the most soluble MIGET gases, ether and acetone, exchange primarily in the airways, violating a basic assumption for application of MIGET.


To examine the impact of airway gas exchange on the MIGET predictions, we replaced ether and acetone in the MIGET analysis with two gases of similar blood solubility but decreased water solubility (toluene for ether and m-DCB for acetone) and, thereby, theoretically minimized the influence of airway gas exchange on MIGET. A comparison of the MIGET predictions using ether and acetone versus using the two replacement gases provided a straightforward means for evaluating the effect of airway gas exchange on the MIGET predictions. We found that airway gas exchange mostly affects predictions of dead space, mean of the ventilation distribution and standard deviation of the ventilation distribution. We conclude that revision of the protocol for the MIGET method by substituting gases with lower water solubility would improve MIGET predictions.

03:30 PM
03:50 PM
Kim Prisk - Gas mixing in the periphery of the lung: Insights from modeling and microgravity

Gas entering the lung encounters a rapidly dividing network of airways with a total cross-sectional area that increases extremely rapidly from a few cm2 in the trachea, to an area of 50-100 m2 at the gas exchange surface. As a consequence, the forward velocity of the inspired gas falls precipitously as the airway tree is traversed, and in the periphery of the lung, diffusion is the dominant gas transport process.


Understanding gas transport in the periphery of the lung rests on modeling the interaction between diffusive and convective transport processes. Extensive modeling studies have shown that asymmetry in the vicinity of the acinar entrance, results in an inhomogeneous distribution of gas within the lung, even if the lung expands homogeneously. This degree of structural asymmetry in the lung at the point at which convective transport and diffusive transport are of a similar magnitude greatly affects the behavior of the exhaled gas profile. Further, variability in the degree of structural asymmetry is a critical component of how inhomogeneously inspired gas is distributed within the lung.


By using gases of different diffusivity (typically He and SF6), studies of gas mixing in the periphery of the lung may be performed, since the convective component of the transport is largely the same for both gases. When these studies were performed in sustained microgravity, the difference in Phase III slope between SF6 and He (normally positive in ground studies) was abolished. These results suggested a change in the conformation of the acinus of the lung (the functional gas exchange unit of the lung). Identical studies performed in the transient microgravity of parabolic flight produced different results to those in sustained microgravity, showing that the time constant of the changes in conformation was relatively long. Ground-based studies suggest that changes in the amount of extra-vascular water in the pulmonary interstitium, the early stages of pulmonary edema, may be responsible.

06:30 PM
07:30 PM
Peter Macklam - A Physician's View of Complexity, the Origins of Order, Health and Disease

Two relatively unexplored features characterize physiologic systems: 1) They are complex, non-linear and dynamic which results in emergent phenomena that can neither be predicted nor explained by examining their component parts in isolation; 2) they become highly ordered during fetal development and throughout the course of Darwinian evolution in apparent violation of the second law of thermodynamics. It follows that interconnections among the parts must play a role in emergent phenomena and the origin of order. How this is accomplished through the nature and number of interconnections has been explored by Kauffman(1). Explanation of increasing order in spite of the second law was achieved by Prigogine(2) who showed that order can spontaneously appear in systems close to thermodynamic equilibrium if they are made to dissipate energy which increases order by displacing them far from equilibrium and decreasing entropy production rate. The approaches of Kauffman and Prigogine have not been combined or reconciled and this needs to be done in order to have a more complete understanding of health and how it breaks down in disease. If energy dissipation in a system is too little or too much and/or if the nature or number of a system's interconnections is altered malfunction results. Although how this occurs is rather obscure, fluctuations in time and space are a common feature of complex systems. Many ways have been used to characterize these fluctuations but few have yet proven beneficial to medicine. Another common feature of complex systems is that, unlike many physical systems, the future can only be assessed by statistical probability. Physicians deal inadequately with uncertainty. Prognosis is part of the art of medicine and is the least scientific part of our profession. Yet the development of statistical mechanics to quantify probabilities in quantum mechanics has the potential of making prognosis more precise. Of the many ways to characterize fluctuations in complex systems, power laws are ubiquitous(3). They have powerful predictive properties; e.g., the Gutenberg Richter Law can predict the probability of an earthquake of any magnitude occurring over any region of the earth's surface over any given time interval with a high degree of certainty. Can power laws make prognosis quantitative? Although the future of physiology is uncertain, I predict our understanding of health will depend on uncovering the secrets of energy-dissipating, interconnected complex biological systems. Precise knowledge of how abnormal interconnections and energy dissipation leads to dysfunction is essential in the understanding of disease and should lead to more precise prognostication.



  1. S. Kauffman. The Origins of Order: Self-Organization and Selection in Evolution. New York: Oxford University Press, 1993.

  2. I. Prigogine and I Stengers. Order Out of Chaos: Man's New Dialogue with Nature. New York: Bantm Books, 1984

  3. P. Bak. How Nature Works: The Science of Self-Organized Criticality. New York: Springer-Verlag, 1996.

Thursday, November 9, 2006
Time Session
09:00 AM
09:20 AM
Ken Lutchen - Modeling lung mechanical function

N/A

09:20 AM
09:40 AM
Jose Vengas - Paradoxical airway response to bronchoprovocation, a manifestation of complex system behavior

Heterogeneity in airway constriction and in ventilation are cardinal features of asthma and understanding them is important as they affect overall mechanical obstruction, gas exchange efficiency, intrapulmonary delivery of therapeutic agents, and interpretation of diagnostic tests.


The current paradigm in asthma research assumes that knowledge about individual components of the lung (airways, tissues, or cells) can be extrapolated to predict full organ behavior. This approach is of limited quantitative value when complex interactions amongst components of the system are involved. In this presentation we will review experimental data challenging the current appraoch and will present modeling results demonstrating that interdependence amongst lung components may be the cause for those experimental findings


Using an integrative model of the airway tree we have demonstrate that, smooth muscle activation cause heterogeneous ventilation characterized by large and contiguous clusters of highly constricted terminal bronchioles similar to ventilation defects observed with MRI, and PET. The heterogeneous behavior of the model is remarkable because it is exhibited by a virtually symmetric airway tree, with uniform parameters at each generation and homogeneous smooth muscle activation.


In this report we evaluate the interdependent behavior amongst the airways of such a model and present the following results: 1) Smooth muscle activation above a critical level yields a heterogeneous response of the airways including dilation of some and constriction of others. 2) A reduction of end-expiratory lung volume during breathing reduces the critical level of smooth muscle activation triggering clusters of severe constriction. 3) Smooth muscle activation of a single central airway leads to its full closure although stronger but uniform activation of the fill tree does not. And 4) A progressive relaxation from maximal smooth muscle activation leads to a progressive reduction in the extent of ventilation defects and a simultaneous constriction of central airways.


We postulate that these results are manifestations of interdependent behavior in a complex system which may explain apparently paradoxical experimental findings.

10:40 AM
11:00 AM
Brent McParland - Determinants of hyperresponsiveness: Real Media

N/A

01:30 PM
01:50 PM
Bela Suki - Hierarchical force transmission in the lung: some modeling results relevant to the normal and emphysematous parenchyma

The lung tissue is constantly under a preexisting tensile stress also called prestress which results from the distension of the lung by the transpulmonary pressure. The regional distribution of the prestress is determined by the hydrostatic pressure in the pleural space and the shape of the lung. Superimposed on this prestress are additional stresses due to breathing which change cyclically and irregularly. The prestress in the alveolar wall is transferred through the extracellular matrix (ECM) to the adhering cells with important consequences on cellular biophysics, biochemistry and phenotype which can modulate connective tissue homeostasis itself. The interaction between the ECM and cellular biochemistry also has important implications for the biomechanical properties of the connective tissues of the lung. Recently, we have argued that collagen plays a major role in transmitting the transpulmonary pressure to lung cells in the alveolar septa through a hierarchical transmission of mechanical stimuli from the level of the whole lung down to single cells with various possible feedback loops controlling ECM remodeling and ultimately organ level mechanics. In this multiple loop system, the alveolar wall network plays an important role since it must respond to any changes in local stiffness. We will discuss several general modeling approaches that are appropriate to describe this hierarchical force transmission in the normal and diseased lungs. As specific examples, we will describe several models of the parenchyma using two- or three-dimensional spring networks. By connecting different length scales, these models are able to account for many functional properties of the normal and the emphysematous lung including remodeling due to stretch, the deterioration of lung function due to rupture of the alveolar septa, or the effects of lung volume reduction surgery on survival rate. Finally, based on the results, we will speculate on how small- and large-scale heterogeneities necessarily develop during the progression of emphysema and possibly other diseases.

01:50 PM
02:10 PM
James Butler - Scale-free methods from physics

N/A

Friday, November 10, 2006
Time Session
Name Email Affiliation
Aguda, Baltazar bdaguda@gmail.com MBI - Long Term Visitor, Bioinformatics Institute, Singapore
Albert, Mitch malbert@bwh.harvard.edu Brigham and Women's Hospital, Harvard University
Alencar, Adriano aalencar@hsph.harvard.edu Department of Environmental Health, Harvard University
Anderson, Joe clarkja@u.washington.edu Bioengineering, University of Washington
Annapragada, Ananth ananth.annapragada@uth.tmc.edu Health Information Sciences, University of Texas
Audi, Said said.audi@marquette.edu Department of Biomedical Engineering, Marquette University
Bates, Jason jason.h.bates@uvm.edu Medicine; Physiology; Biophysics, University of Vermont
Bentil, Daniel dbentil@uvm.edu Math & Stats., Molecular Physiology, Biology, University of Vermont
Best, Janet jbest@mbi.osu.edu
Butler, James jbutler@hsph.harvard.edu Physiology, Environmental Health, Harvard University
Castile, Robert castiler@pediatrics.osu.edu Center for Perinatal Research, The Ohio State University
Clanton, Thomas clanton.1@osu.edu Internal Medicine, Davis Heart and Lung, The Ohio State University
Clough, Anne clough@mscs.mu.edu Biomedical Engineering, Marquette University
Darquenne, Chantal cdarquenne@ucsd.edu Dept. of Medicine, University of California, San Diego
Davis, Ian davis.2448@osu.edu Veterinary Biosciences, The Ohio State University
Djordjevic, Marko mdjordjevic@mbi.osu.edu Mathematical Biosciences Institute (MBI), The Ohio State University
Enciso, German German_Enciso@hms.harvard.edu Mathematical Biosciences Institute (MBI), The Ohio State University
Fabry, Ben Physical Medical Technology, Universität Erlangen-Nürnberg
Fredberg, Jeff jeffrey_fredberg@harvard.edu Bioengineering and Physiology, Harvard University
George, Steve scgeorge@uci.edu Biomedical Engineering, University of California, Irvine
Ghadiala, Samir sag3@lehigh.edu Mechanical Engineering, Lehigh University
Grajdeanu, Paula pgrajdeanu@mbi.osu.edu Mathematical Biosciences Institute (MBI), The Ohio State University
Gunst, Susan sgunst@iupui.edu Physiology and Biophysics, Indiana University--Purdue University
Hammersley, Jeff jhammersley@mco.edu Pulmonary, Critical Care Medicine, The Medical University of Ohio
Henry, Frank fhenry@hsph.harvard.edu Harvard University
Hoffman, Eric eric-hoffman@uiowa.edu Physiologic Imaging, Radiology, University of Iowa
Kim, Yangjin ykim@mbi.osu.edu Mathematical Biosciences Institute (MBI), The Ohio State University
Kitaoka, Hiroko kitaoka@imed3.med.osaka-u.ac.jp Graduate School of Medicine, Osaka University
Knoell, Daren daren.knoell@osumc.edu Davis HLRI, The Ohio State University
LaPrad, Adam alaprad@bu.edu Respiratory and Physiological Systems ID Lab, Boston University
Lin, Ching-Long ching-long-lin@uiowa.edu Mechanical and Industrial Engineering, University of Iowa
Long, Frederick flong@chi.osu.edu Radiology, The Ohio State University
Lou, Yuan lou@math.ohio-state.edu MBI - Long Term Visitor, The Ohio State University
Lutchen, Ken klutch@bu.edu Department of Biomedical Engineering, Boston University
Ma, Baoshun bma@bu.edu Respiratory and Physiological Systmes, Boston University
Macklam, Peter peter.macklem@mcgill.ca Montreal Chest Institute, McGill University, Macdonald Campus
Macksym, Geof gmaksym@dal.ca Biomedical Engineering, Dalhousie University
Majumdar, Arnab arnab@buphy.bu.edu Biomedical Engineering, Boston University
McParland, Brent BrentMcParland@med.usyd.edu.au Pharmacology, University of Sydney
Mijailovich, Srboljub smijailo@hsph.harvard.edu Environmental Health, Harvard University
Mitzner, Wayne wmitzner@jhsph.edu Physiology, Johns Hopkins University
Moldovan, Nicanor moldovan.6@osu.edu Internal Medicine, Davis Heart and Lung, The Ohio State University
Nevai, Andrew anevai@mbi.osu.edu Mathematical Biosciences Institute (MBI), The Ohio State University
Ng, Bart bng@math.iupui.edu Department of Mathematical Sciences, Indiana University--Purdue University
Noble , Peter noblep01@cyllene.uwa.edu.au Biomedical, Biomolecular & Chemical Science, University of Western Australia
Oliver, Ryan oliver.201@osu.edu Internal Medicine, The Ohio State University
Olson, Lynne olson.3@osu.edu Veterinary Biosciences, The Ohio State University
Oster, Andrew aoester@mbi.osu.edu Mathematical Biosciences Institute (MBI), The Ohio State University
Polak, Adam adam.polak@pwr.wroc.pl Electronic and Photonic Methology, Wroclaw University of Technology
Prisk, Kim kprisk@ucsd.edu Medicine, University of California, San Diego
Rempe, Michael mrempe@mbi.osu.edu Mathematical Biosciences Institute (MBI), The Ohio State University
Sanderson, Michael michael.sanderson@umassmed.edu Physiology, University of Massachusetts
Schugart, Richard richard.schugart@wku.edu Mathematical Biosciences Institute (MBI), The Ohio State University
Seow, Chun cseow@mrl.ubc.ca Cellular and Physiological Sciences, University of British Columbia
Simon, Brett bsimon@jhmi.edu Medicine, Anesthesiology, Johns Hopkins University
Srinivasan, Partha p.srinivasan35@csuohio.edu Mathematical Biosciences Institute (MBI), The Ohio State University
Stamenovic, Dimitrije dimitrij@bu.edu Biomedical Engineering, Boston University
Stigler, Brandy bstigler@mbi.osu.edu Mathematical Biosciences Institute (MBI), The Ohio State University
Suki, Bela bsuki@bu.edu Biomedical Engineering, Boston University
Szomolay, Barbara b.szomolay@imperial.ac.uk Mathematical Biosciences Institute (MBI), The Ohio State University
Tepper, Rob rtepper@iupui.edu Department of Pediatrics, Indiana University--Purdue University
Tian, Paul tianjj@mbi.osu.edu Mathematical Biosciences Institute (MBI), The Ohio State University
Tsuda, Akira atsuda@hsph.harvard.edu Environmental Health, Physiology, Harvard Medical School
Vengas, Jose jvenegas@vqpet.mgh.harvard.edu Biomedical Engineering, Massachusetts General Hospital
Verbanck, Sylvia sylvia.verbanck@az.vub.ac.be Academic Hosptical (AZ VUB), Vrije Universiteit Brussel
Winkler, Tilo tilo@habanero.mgh.harvard.edu Depts. of Anesthesia and Critical Care, Massachusetts General Hospital
Wright, Valerie wright.14@osu.edu Internal Medicine, The Ohio State University
Zhang, Linghai liz5@lehigh.edu Department of Mathematics, Lehigh University
Impact of Airway Gas Exchange on the Multiple Inert Gas Elimination Technique

The multiple inert gas elimination technique (MIGET) provides a method for estimating the efficiency of alveolar gas exchange. Six soluble inert gases with blood solubility spanning six orders of magnitude are infused into a peripheral vein of an animal. After gas equilibration, the exhaled, venous, and arterial partial pressures of these six gases are measured and interpreted by using a mathematical model of alveolar gas exchange. MIGET has been very successful in improving our understanding of alveolar gas exchange in both health and disease.


Developed in the early 1970s, MIGET assumed that none of the six chosen gases exchanged with the airways. However, this assumption is no longer considered to be valid. Over the past 20 years, our experimental and theoretical investigations on airway gas exchange have shown that the location of gas exchange (airway versus alveoli) depends primarily on gas solubility in both blood and water (as measured by a partition coefficient). Gases with both blood-air and water-air partition coefficients greater than 10 exchange primarily in the airways. Meeting this criteria, two of the most soluble MIGET gases, ether and acetone, exchange primarily in the airways, violating a basic assumption for application of MIGET.


To examine the impact of airway gas exchange on the MIGET predictions, we replaced ether and acetone in the MIGET analysis with two gases of similar blood solubility but decreased water solubility (toluene for ether and m-DCB for acetone) and, thereby, theoretically minimized the influence of airway gas exchange on MIGET. A comparison of the MIGET predictions using ether and acetone versus using the two replacement gases provided a straightforward means for evaluating the effect of airway gas exchange on the MIGET predictions. We found that airway gas exchange mostly affects predictions of dead space, mean of the ventilation distribution and standard deviation of the ventilation distribution. We conclude that revision of the protocol for the MIGET method by substituting gases with lower water solubility would improve MIGET predictions.

Modeling emergent behavior

The lung consists of myriad disparate components that interact in a rich variety of ways. Accordingly, a quantitative understanding of how the lung functions can be pursued at many levels of scale, from that of the molecule up to the whole organ. Importantly, however, a complete understanding of lung function at the organ level does not follow automatically from knowledge of the individual characteristics of its components. Indeed, understanding the components in isolation may not even constitute the major requirement for an understanding of the lung as a whole. This is because the lung is a complex system - something with emergent properties that arise from the way its components operate as an ensemble. Emergent properties in a complex system typically bear little qualitative relationship to the individual properties of the system components, yet somehow arise from these components and their interactions in frequently subtle ways. Prediction of emergence in complex systems is well known to physicists, as it constitutes the central paradigm of statistical mechanics that began with the kinetic theory of gases more than a century ago. Emergence has been formally embraced by the biomedical community only more recently, but it impacts the life of every scientist, whether they know it or not, through the Central Limit Theorem which provides a theoretical basis for the widespread appearance of the Gaussian distribution throughout the natural world. Even the ubiquity of power law processes in nature, which is now exciting a great deal of interest in biology and physics and makes its appearance in the lung, likely represents an emergent phenomenon. Understanding the genesis of emergent behavior, however, is frequently not straightforward. Highly nonlinear interactions between system components are typical of complex biological systems, which severely limits the use of analytical methods for predicting the emergent properties arising there from. The exploration of emergent behavior in the lung is thus going to be based on computational models based constructs such as cellular automata, artificial network networks, percolation networks, and distributed recruitment mechanisms. As specific examples, we will consider how percolation and recruitment mechanisms might be invoked to account for the bulk mechanical properties of lung tissue.



 
Scale-free methods from physics

N/A

Nonlinear viscoelasticity of living cells

The linear viscoelastic behavior of adherent cells and biological tissue is characterized by a wide distribution of relaxation times and shows a power-law creep response or a power-law viscoelastic spectrum over several decades of time or frequency. In addition, single cells and tissue exhibit a highly nonlinear stress-strain relationship.  The viscoelastic behavior of cells in the non-linear regime is unknown, however, but is of particular interest to test different conflicting theories.  At high stress, soft glassy rheology for example predicts a speed-up of relaxation processes due to yielding, whereas the theory of semiflexible polymer networks predicts a slow-down due to reduced thermal bending fluctuations in low-frequency modes.  We measured the viscoelastic behavior of different cell types (fibroblasts, epithelial cells, endothelial cells, lung and breast carcinoma cells, vinculin mutant cells) in the linear and non-linear regime using magnetic tweezers with real-time force feedback.  We imposed a staircase-like sequence of 1 nN force steps up to a maximum force of 10 nN onto 4.5 µm fibronectin-coated magnetic beads bound to the cytoskeleton via integrins. For each stress level s, the differential creep response of single cells followed a power law: J(s,t) = J0(s) ta(s), however the differential creep modulus J0(s) decreased with stress, equivalent to a linear stress stiffening. The power-law creep exponent a showed stress dependence that was cell type specific.  We observed both increasing and decreasing exponents with increasing force. Stress reversal and repeated measurement on the same cell showed that creep recovery was incomplete and the creep modulus approached a steady state after several cycles. This behavior resembles pre-conditioning in tissue and could possibly arise from force-induced permanent structural changes in the cytoskeleton. In summary, living cells exhibit a nonlinear viscoelastic behavior remarkably similar to connective tissue. 

Cytoskeletal remodeling and slow dynamics in the living cell: Are we built of glass?

With every beat of the heart, inflation of the lung, or peristalsis of the gut, cell types of diverse function are subjected to substantial stretch. New data show that cell responses to a transient stretch exhibit remarkable physical similarities to fluidization observed in jammed inert matter, including colloids, pastes, emulsions, and foams, and thus implicate mechanisms mediated not only by specific signaling intermediates, as is usually presumed, but also by nonspecific actions of a slowly evolving network of physical forces. These results support the idea that the cell interior is at once a crowded chemical space and a fragile soft material in which the effects of biochemistry, molecular crowding, and physical forces are complex and inseparable, yet conspire nonetheless to yield remarkably simple phenomenological laws. These laws appear to be universal and thus comprise a striking point of contact between the worlds of cell signaling biology and soft matter physics.

An Integrative Approach Towards Understanding Nitrogen Oxide Biology in the Lungs

Asthma is a chronic disease that affects approximately 10% of the population in the United States. Although our understanding of underlying mechanisms (e.g., Th2 type inflammation) has continued to increase over the past several decades, the prevalence continues to rise. Furthermore, the mainstay of monitoring (spirometry) and therapy (anti-inflammation and -agonist bronchodilation) are essentially unchanged over the same time period. Over the past ten years, the interest in using exhaled nitric oxide as a non-invasive marker of inflammation has steadily increased despite a limited understanding of the cell and molecular mechanisms. More recently, the loss of naturally occurring NO-based bronchodilators (S-nitrosothiols or SNOs) in an animal model of asthma has sparked renewed interest in the key role of NO biochemistry in modulating asthma. Our group combines advanced primary cell culture techniques, direct measurement of gas phase NO release, quantitation of enzymatic activity, exhaled nitric oxide from human subjects, and mathematical models in an integrative fashion to understand how multiscale systems interact to influence NO storage and release in the lungs. We hope this approach will significantly improve our mechanistic understanding of NO metabolism and gas phase release, and thus non-invasive inflammatory monitoring of asthma.



 
Fluid-Structure Modeling of Cellular Deformation and Injury in Pulmonary Airways

Several respiratory disorders including infant RDS, adult RDS and acute lung injury, are characterized by the fluid-occlusion of small pulmonary airways and alveoli. The reopening of these fluid-filled structures involves the movement of air-liquid interfaces and the application of complex hydrodynamic and surface tension forces on the epithelial cells which line airway and alveolar walls. Although these fluid mechanical forces may contribute to ventilator-induced lung injury, the mechanism of cellular injury during these free-surface flows is not well established. Under these conditions, the amount of cellular deformation is not known a priori and may depend on several multi-scale factors including a) reopening dynamics, b) surfactant transport to the air-liquid interface, c) morphological and/or micro-mechanical properties of the epithelial cells and d) molecular interactions at focal adhesion sites. In addition, biochemical responses (i.e. protein/gene expression) to these different hydrodynamic conditions may also contribute to lung injury. Our laboratory utilizes a combination of in-vitro cell culture experiments, advanced imaging techniques and computational models to investigate the relative importance of these different factors. The goal is to obtain a better understanding of the biophysical and biological mechanisms of cell/tissue injury during the reopening of fluid-filled airways/alveoli.


In this presentation we will first present a brief overview of previous studies which have elucidated the multiphase fluid mechanics of airway/alveolar reopening. We will then present recent experimental data which demonstrate how changes in reopening dynamics, cellular morphology, cell mechanics and cytoskeletal structure influence cell injury during airway reopening. We will also discuss the development of multi-scale, computational models of cellular deformation during airway reopening. In addition to providing valuable insight into the mechanism of cell injury, these computational models have been used to explain several counter-intuitive experimental results including increased cell death at slow reopening velocities and decreased death in cells with a lower elastic modulus (i.e. a more flexible cell). Finally, we will discuss future challenges in modeling cellular responses to complex hydrodynamic and surface tension forces including the incorporation of realistic airway/cellular geometries, non-linear membrane mechanics, heterogeneous viscoelastic cell properties and molecular interactions at focal adhesion sites.

Lung structure from imaging

N/A

Morphogenesis-based 4D model of the alveolar structure

There has been no consensus how the alveolar shape changes during breathing due to difficulties in observing directly alveoli in vivo. A computational model of dynamical 3D model, that can be called a 4D model, would be useful for understanding such an invisible phenomenon. The living organ changes its shape during morphogenetic process as well as during physiologic motion. Therefore, an algorithm for constructing a 3D model of the organ should be consistent with its morphogenetic process, too. In other words, the fourth dimension of the organ model should possess both of physiological and morphogenetic time scales. The alveolar morphogenesis is roughly classified into four steps from a geometric point of view: formation of a fluid pathway, enlargement of intra-acinar pathway, bellows-like arrangement of primary septa by capillary invasions, and growth of secondary septa whose edges form alveolar mouths containing abundant elastin fibers. Those steps are accompanied with functional maturation: making a duct transport system, utilize full space in the organ, increasing surface area, and finally acquiring deformability. We have constructed a 4D alveolar model according to the above morphogenetic process, where the alveolar deformation is modeled by a combination of spring (elastin fibers at alveolar mouths) and hinge (septal junctions). The model includes a hypothesis that alveolar mouths are closed at minimum volume and that closed alveoli are stabilized by the alveolar lining liquid film containing surfactant. The validity of the model has been verified not fully but satisfactorily by comparing it with real experimental results; in vivo microscopic observation of subpleural alveoli (courtesy of Nieman, GF), histologic sections of rapidly frozen lungs (in literatures), and morphometric investigations by light scattering and intravascular fixation-dehydration method (in literatures).



 
Multi-scale Simulation of Pulmonary Air Flow

In this presentation I will talk about the progress in developing a comprehensive computational fluid dynamic (CFD) model for pulmonary air flow which utilizes subject-specific airway geometries and employs a Computed Tomography (CT) data-driven, multi-stage approach to provide accurate predictions of regional ventilation and gas transport through the entire moving airway tree. The model is based up an in-house parallel characteristic Galerkin fractional four-step finite element method. Both direct numerical simulation (DNS) and large-eddy simulation (LES) techniques have been employed to simulate turbulent flow in the proximal conducting airways. A three-stage approach has been adopted to simulate air flow in the acinar airways. Due to the large mesh size, some CFD simulations have been carried on TeraGrid clusters. The effects of boundary conditions and subject-specific human airways on air flow have been assessed. The transient air flow in a CT-based monopodial sheep (ovine) tracheobronchial tree of up to 13 generations has also been studied numerically. Currently, a geometric model of 11 generations of human airways has been constructed and the preliminary CFD simulation has been successfully performed. The progress in developing an in-house mesh generation software which aims to create CFD mesh from the trachea to the terminal bronchioles will be reported. Toward modeling a breathing lung, the host-mesh technique in conjunction with 4D dynamic imaging, the model for soft tissue mechanics, and the fluid-structure interaction (FSI) technique have been or are being developed, which will be briefly discussed as well.



 
Modeling lung mechanical function

N/A

A Physician's View of Complexity, the Origins of Order, Health and Disease

Two relatively unexplored features characterize physiologic systems: 1) They are complex, non-linear and dynamic which results in emergent phenomena that can neither be predicted nor explained by examining their component parts in isolation; 2) they become highly ordered during fetal development and throughout the course of Darwinian evolution in apparent violation of the second law of thermodynamics. It follows that interconnections among the parts must play a role in emergent phenomena and the origin of order. How this is accomplished through the nature and number of interconnections has been explored by Kauffman(1). Explanation of increasing order in spite of the second law was achieved by Prigogine(2) who showed that order can spontaneously appear in systems close to thermodynamic equilibrium if they are made to dissipate energy which increases order by displacing them far from equilibrium and decreasing entropy production rate. The approaches of Kauffman and Prigogine have not been combined or reconciled and this needs to be done in order to have a more complete understanding of health and how it breaks down in disease. If energy dissipation in a system is too little or too much and/or if the nature or number of a system's interconnections is altered malfunction results. Although how this occurs is rather obscure, fluctuations in time and space are a common feature of complex systems. Many ways have been used to characterize these fluctuations but few have yet proven beneficial to medicine. Another common feature of complex systems is that, unlike many physical systems, the future can only be assessed by statistical probability. Physicians deal inadequately with uncertainty. Prognosis is part of the art of medicine and is the least scientific part of our profession. Yet the development of statistical mechanics to quantify probabilities in quantum mechanics has the potential of making prognosis more precise. Of the many ways to characterize fluctuations in complex systems, power laws are ubiquitous(3). They have powerful predictive properties; e.g., the Gutenberg Richter Law can predict the probability of an earthquake of any magnitude occurring over any region of the earth's surface over any given time interval with a high degree of certainty. Can power laws make prognosis quantitative? Although the future of physiology is uncertain, I predict our understanding of health will depend on uncovering the secrets of energy-dissipating, interconnected complex biological systems. Precise knowledge of how abnormal interconnections and energy dissipation leads to dysfunction is essential in the understanding of disease and should lead to more precise prognostication.



  1. S. Kauffman. The Origins of Order: Self-Organization and Selection in Evolution. New York: Oxford University Press, 1993.

  2. I. Prigogine and I Stengers. Order Out of Chaos: Man's New Dialogue with Nature. New York: Bantm Books, 1984

  3. P. Bak. How Nature Works: The Science of Self-Organized Criticality. New York: Springer-Verlag, 1996.

Determinants of hyperresponsiveness: Real Media

N/A

Modeling airway distensibility in the lung

The lung is an inflatable organ that must incorporate three separate tree structures. Thus, in order for the lung to inflate, all of these trees must stretch both axially and radially. Indeed, it was noted almost a century ago that if the airway tree could not lengthen, then the lung likely could not inflate. In this presentation we will discuss the first distensibility of airways, and then how this distensibility interacts with that of the surrounding lung parenchyma.


In modeling the lung an implicit assumption is often made that airway diameter varies as the cube root of lung volume. While there has been some published evidence to support this under certain conditions, the assumption is not supported by what is known about the airway structure. The role of the relatively nondistensible basement membrane is generally ignored. However, the fact that the length of this membrane is relatively fixed has been used histologically to compare airways of comparable size in asthmatic and normal subjects. So if the basement membrane is rigid, how can the airway diameter vary as the cube root of lung volume? Either this fixed length is larger than the largest size the airway ever achieves, or the assumption is wrong. We will show that the latter reason is the correct one.


We will show that pressure-area curves of relaxed airways in vivo are highly nonlinear, reaching a maximal size at a very low pressure (less than 10 cmH2O). The limit is consistent with a rigid basement membrane. With smooth muscle tone, the relation between airway area and pressure becomes highly variable depending on the level of tone and other uncertain factors. This variability in tone even under baseline conditions is what often causes misinterpretation of the nature of the distensibility. All dogs we have studied have a variable degree of baseline tone, averaging from 50-70% of maximally relaxed size. Although most clinical research suggests that there is little baseline tone in normal healthy subjects, this is not true. With sufficient relaxation of the airway smooth muscle, we can show that even normal humans also live with airways at about 705 of maximal size.


These results stress the importance of understanding the link between airway structure and their functional distensibility. The way airways respond to inflation may have important clinical manifestations in the asthmatic pathology.

Gas mixing in the periphery of the lung: Insights from modeling and microgravity

Gas entering the lung encounters a rapidly dividing network of airways with a total cross-sectional area that increases extremely rapidly from a few cm2 in the trachea, to an area of 50-100 m2 at the gas exchange surface. As a consequence, the forward velocity of the inspired gas falls precipitously as the airway tree is traversed, and in the periphery of the lung, diffusion is the dominant gas transport process.


Understanding gas transport in the periphery of the lung rests on modeling the interaction between diffusive and convective transport processes. Extensive modeling studies have shown that asymmetry in the vicinity of the acinar entrance, results in an inhomogeneous distribution of gas within the lung, even if the lung expands homogeneously. This degree of structural asymmetry in the lung at the point at which convective transport and diffusive transport are of a similar magnitude greatly affects the behavior of the exhaled gas profile. Further, variability in the degree of structural asymmetry is a critical component of how inhomogeneously inspired gas is distributed within the lung.


By using gases of different diffusivity (typically He and SF6), studies of gas mixing in the periphery of the lung may be performed, since the convective component of the transport is largely the same for both gases. When these studies were performed in sustained microgravity, the difference in Phase III slope between SF6 and He (normally positive in ground studies) was abolished. These results suggested a change in the conformation of the acinus of the lung (the functional gas exchange unit of the lung). Identical studies performed in the transient microgravity of parabolic flight produced different results to those in sustained microgravity, showing that the time constant of the changes in conformation was relatively long. Ground-based studies suggest that changes in the amount of extra-vascular water in the pulmonary interstitium, the early stages of pulmonary edema, may be responsible.

Video imaging of parenchymal mechanics

N/A

Clinical Perspectives: Modeling in Acute Lung Injury

Acute lung injury (ALI) and the adult respiratory distress syndrome (ARDS) are syndromes of respiratory failure with mortality in excess of 30% affecting an estimated 25-50 patients per 100,000 per year worldwide. While there are a variety of initiating events, the final common pathway is characterized by hypoxemic respiratory failure with a heterogeneous loss of aerated lung volume and reduced compliance due to flooding and collapse, surfactant inactivation, increased dead space, and inflammation. Clinical management is currently limited to supportive measures, primarily low tidal volume mechanical ventilation with positive end-expiratory pressure (PEEP), in an attempt to recruit the poorly aerated lung while reducing injurious cyclic end-expiratory airspace opening and closing and end-inspiratory regional overdistension. Many other therapeutic approaches, including pharmacologic therapy to reduce edema formation or inflammation or improve perfusion distribution, exogenous surfactant treatment, and novel mechanical ventilation strategies, have been tried or are in development but none have yet been shown to improve outcome.


Modeling of ALI/ARDS pathophysiology to date has been focused on the quasi-static pressure-volume properties of the injured lungs, primarily in an attempt to predict recruitment and optimal mechanical ventilation strategies, but these approaches have had limited success, possibly because they do not adequately reflect the very different dynamic operating conditions of the injured lung. Opportunities for modeling to contribute to improved understanding of ALI pathophysiology and management abound. Better characterization of the dynamic mechanical properties of the lung, particularly the distributed regional contributions, non-linear effects over the tidal breathing cycle, and a focus on heterogeneity may help optimize mechanical ventilation management strategies. Next, the consequences of these interventions on ventilation and perfusions distributions and matching, dead space, and vascular regulation are important in translating mechanics to improved physiology and effective gas exchange. Novel therapies such as high-frequency oscillatory ventilation, airway pressure-release ventilation (APRV), and "noisy" ventilation need to be understood in terms of their mechanisms of gas exchange and effects on lung mechanics and V/Q. ALI/ARDS represents an area in which there is an opportunity for model-based advances to very rapidly achieve translation to the clinical arena and thus have great potential to improve patient care and outcomes.

Hierarchical force transmission in the lung: some modeling results relevant to the normal and emphysematous parenchyma

The lung tissue is constantly under a preexisting tensile stress also called prestress which results from the distension of the lung by the transpulmonary pressure. The regional distribution of the prestress is determined by the hydrostatic pressure in the pleural space and the shape of the lung. Superimposed on this prestress are additional stresses due to breathing which change cyclically and irregularly. The prestress in the alveolar wall is transferred through the extracellular matrix (ECM) to the adhering cells with important consequences on cellular biophysics, biochemistry and phenotype which can modulate connective tissue homeostasis itself. The interaction between the ECM and cellular biochemistry also has important implications for the biomechanical properties of the connective tissues of the lung. Recently, we have argued that collagen plays a major role in transmitting the transpulmonary pressure to lung cells in the alveolar septa through a hierarchical transmission of mechanical stimuli from the level of the whole lung down to single cells with various possible feedback loops controlling ECM remodeling and ultimately organ level mechanics. In this multiple loop system, the alveolar wall network plays an important role since it must respond to any changes in local stiffness. We will discuss several general modeling approaches that are appropriate to describe this hierarchical force transmission in the normal and diseased lungs. As specific examples, we will describe several models of the parenchyma using two- or three-dimensional spring networks. By connecting different length scales, these models are able to account for many functional properties of the normal and the emphysematous lung including remodeling due to stretch, the deterioration of lung function due to rupture of the alveolar septa, or the effects of lung volume reduction surgery on survival rate. Finally, based on the results, we will speculate on how small- and large-scale heterogeneities necessarily develop during the progression of emphysema and possibly other diseases.

Folding and mixing

N/A

Paradoxical airway response to bronchoprovocation, a manifestation of complex system behavior

Heterogeneity in airway constriction and in ventilation are cardinal features of asthma and understanding them is important as they affect overall mechanical obstruction, gas exchange efficiency, intrapulmonary delivery of therapeutic agents, and interpretation of diagnostic tests.


The current paradigm in asthma research assumes that knowledge about individual components of the lung (airways, tissues, or cells) can be extrapolated to predict full organ behavior. This approach is of limited quantitative value when complex interactions amongst components of the system are involved. In this presentation we will review experimental data challenging the current appraoch and will present modeling results demonstrating that interdependence amongst lung components may be the cause for those experimental findings


Using an integrative model of the airway tree we have demonstrate that, smooth muscle activation cause heterogeneous ventilation characterized by large and contiguous clusters of highly constricted terminal bronchioles similar to ventilation defects observed with MRI, and PET. The heterogeneous behavior of the model is remarkable because it is exhibited by a virtually symmetric airway tree, with uniform parameters at each generation and homogeneous smooth muscle activation.


In this report we evaluate the interdependent behavior amongst the airways of such a model and present the following results: 1) Smooth muscle activation above a critical level yields a heterogeneous response of the airways including dilation of some and constriction of others. 2) A reduction of end-expiratory lung volume during breathing reduces the critical level of smooth muscle activation triggering clusters of severe constriction. 3) Smooth muscle activation of a single central airway leads to its full closure although stronger but uniform activation of the fill tree does not. And 4) A progressive relaxation from maximal smooth muscle activation leads to a progressive reduction in the extent of ventilation defects and a simultaneous constriction of central airways.


We postulate that these results are manifestations of interdependent behavior in a complex system which may explain apparently paradoxical experimental findings.