An experimental model of a lung cell.
One way to appreciate the way engineers do research, says Hannah Dailey, is to stand a few inches away from an impressionist painting.
Viewed up close, a Monet or a Renoir is a maze of dots and tiny brush strokes. Only when you step back a few feet can your eye take in the big picture.
Similarly, says Dailey, engineers who develop computational models will be overwhelmed if they try to account for every detail in the systems they study.
Computational models use mathematics to describe the behavior of complicated systems. They enable scientists and engineers, seated at their computers, to predict the performance of bridges, rockets and ships, and to speculate about climate change and population dynamics. They yield valuable information while reducing the need for invasive or destructive physical tests. And after their results are compared with data from lab experiments, the models are improved.
Dailey, a Ph.D. candidate in mechanical engineering, develops computational models to investigate the tiny air sacs, called alveoli, that are located in the deep lungs. Alveoli expand and contract during breathing to bring fresh, oxygen-rich air into the lungs and to remove carbon dioxide from the blood. It is probably safe to say that most people remain happily unaware of their alveoli until a disease fills the sacs with fluid, causing difficult breathing, organ failure and even death.
Using a software program that performs fluid and solid mechanical simulations, Dailey studies the mechanical properties of the epithelial cells that line the alveoli. Alveolar epithelial cells are damaged not only by lung diseases, but also by some of the therapies, especially artificial ventilation, that are prescribed to patients suffering from those diseases.
Dailey hopes to learn more about the harm inflicted on alveolar epithelial cells by ventilation-induced lung injury, and thereby improve the treatments and therapies that are prescribed for diseased and damaged lungs.
In her models, Dailey simulates the mechanical response of the cells to two mechanical forces imposed by passing air bubbles—the downward impact of pressure and the sliding strain of shear.
Her challenge in studying alveolar epithelial cells is not much different from the task facing the impressionist art aficionado. A cell’s mechanical and physical behaviors are far too numerous for each behavior to be factored into a model. Dailey must determine which behaviors are not significant and can be excluded from consideration, while still accurately assessing how cells deform during respiratory disease or treatment.
“When you develop a computational model,” says Dailey, “you have to choose what to—and what not to—simulate. You have to decide what you expect to be significant and what you expect to be negligible, in terms of the physical phenomenon you’re most interested in. A cell has too many structures to model using the tools we have at present. So we take a step back and look at the overall structure.”
Dailey presented the results of her research last summer at the Fifth World Congress on Biomechanics in Munich, Germany. The conference, held every four years, is the leading event of its kind. Dailey was one of three researchers, out of 40 invited, to win a best poster award.
Her success is one of several enjoyed recently by researchers in Lehigh’s Respiratory Biomechanics Laboratory.
At the Annual Bioengineering Conference of the ASME (American Society of Mechanical Engineers) in Florida, Samir Ghadiali, director of the lab, and Cagatay Yalcin, a Ph.D. candidate in mechanical engineering, made presentations in the Biofluid Mechanics and Micro/Nano Fluid Mechanics sessions.
Ghadiali was also invited to speak at the World Congress in Munich and at the Fourth International Symposium on Middle Ear Mechanics and Research in Otology in Zurich, Switzerland.
A fortuitous encounter
For someone who spends her days absorbed with computer calculations, Hannah Dailey came relatively late to a love of math and science.
“I had no interest in science,” she says, “until my senior year of high school, when I took my first calculus course. It was only then that I saw how exciting math could be and I realized, ‘This
is why I took algebra and geometry—so I could go on to calculus and solve problems with real-life applications.’”
After earning a B.S. in mechanical engineering from Lehigh in 2002, Dailey went to the U.S. Naval Surface Warfare Center’s Carderock Division in Maryland, where she did computational work on submarines and ships in an effort to improve stealth technology by minimizing noise and vibration.
“It was a great environment,” Dailey says. “I worked with a group of 12 engineers, all Ph.D.s. I got a lot of responsibility really fast. But my colleagues kept telling me, ‘Go back to school now for your Ph.D.’”
Dailey returned to Lehigh in 2003 as a graduate student. While filling out forms to become a teaching assistant, she met Ghadiali, a new faculty member who is now the Frank Hook Assistant Professor of Bioengineering in the department of mechanical engineering and mechanics.
The encounter was fortuitous. Dailey was hoping to work on a computational fluids research project. Ghadiali was developing models to investigate the mechanical properties of the eustachian tube, a project he is continuing with two grants from the National Institutes of Health. Ghadiali has since received a Parker B. Francis Fellowship in Pulmonary Research to study the mechanics of respiratory disorders and lung function.
In 2004, with help from Ghadiali, Dailey applied for and received a National Science Foundation Graduate Research Fellowship, a prestigious award granted to only 900 students a year in the U.S.
Dailey’s first research project, which became her M.S. thesis, focused on particle transport in the deep lung, where molecules move around and collide with each other in a chaotic manner called Brownian diffusion. Among the applications of this research are inhalable drugs, including aerosolized insulin.
Dailey and Ghadiali were the first researchers to develop a computational model based on fluid-structure interactions to study the tissues of the deep lung. Dailey gave an oral presentation on her work, titled “Brownian Diffusion and Fine Aerosol Dynamics in Pulmonary Alveoli,” at the American Society of Mechanical Engineers’ Bioengineering Conference last year in Vail, Colorado.
Comparing lab experiments and computer simulations
A computational model of a lung cell.
In her current research project, which she started a year ago, Dailey works closely with Yalcin, a fellow Ph.D. student, who is conducting lab experiments on alveolar epithelial cells to measure their response to stress.
Yalcin grows two different configurations of cells. A “confluent” group of fully grown cells contains a greater density of cells than a “subconfluent” group of not yet fully grown cells. After staining the cells with a fluorescent dye, Yalcin can determine how many cells live and how many die during lab tests that mimic pulmonary disease.
Yalcin and Dailey also use cross-sectional images of the cells taken by a confocal light microscope to generate topological maps showing the 3-D shapes of the cells. Dailey uses these 3-D models to simulate cell deformation during lung disease. She creates a color-coded map showing, across a time sequence, the degree of deformation that each group of cells undergoes.
Dailey simulates the same activities on her desktop computer (“I am a mousepad jockey,” she cheerfully admits, “a small-muscle athlete”) and then places her computational findings alongside Yalcin’s experimental data for comparison.
So far, the results have been encouraging. Both sets of findings show that the less densely packed subconfluent cells are more vulnerable to injury and death from ventilation-induced and other types of lung distress.
“It’s very exciting to see the correlation between the computational and experimental work,” Dailey says. “A good correlation, with mutually reinforcing results, gives you greater confidence that you have included the right physical phenomena in your model.
“It all goes back to what factors you decide to neglect when you construct your model. Instead of examining everything in the cell and assigning a property for every phenomenon inside the cell, we looked at the whole thing as one homogenous whole. It turns out that we included enough [factors] to get the information we wanted from this particular computation.”
Ghadiali is also pleased with the progress that Dailey and Yalcin have made.
“Our lab has combined molecular biology with mechanical engineering to show that changes in the alveolar epithelial cells’ micro-mechanical structure are an important factor in the development of ventilation-induced lung injury,” he says.
After Dailey completes her Ph.D., she hopes to continue doing computational modeling and collaborating with experimental researchers.
“I love the investigative process and discovering things that people have not yet discovered. I’m also fascinated with biomedical applications. There’s huge room for growth in this field.”
Someday, she believes, biomedical researchers will utilize computational modeling as much as auto safety engineers rely on simulated car crashes.
“Engineers have so many great software tools that we’ve used to build cars and bridges. We can also use them to study the physical behavior of biological systems. Instead of relying mostly on clinical trials and trial and error, we can couple experiments with simulation to create new treatment systems.”