Imagine Magazine - Johns Hopkins - January/February 2009 - (Page 15)

Nolan ran his simulation on six computers, which stayed on his dining room table for the entire month of January. the weekends in December and January, designing the project and learning to program the simulation. I ran it approximately 18,000 times on six computers set up on my dining room table. Engineering with LEGOs Being new to programming, I had used NetLogo software, a basic simulation program, to avoid having to design the entire simulation architecture. NetLogo uses a particle-based model to approximate social interactions and population mixing. In this model, each person was represented as a pixel on a grid. The pixels were separated into different regions within the overall grid to simulate different sub-groups within a given population, and the different regions were connected by smaller, more densely packed connection areas where many of the particles interacted—much like infected people might interact at supermarkets or malls. When a person was within a given distance of another person, the program would see if either person was infected. If one was infected and the other was not, then a probability run would determine if the infection spread. The most fascinating finding was that although elderly people died of the avian flu more often than other age groups, vaccinating school-age children would prevent the most deaths because of children’s high degree of social contact. My project was selected for the Hawaii District Science Fair and won first place in the senior division. I was awarded a trip to attend the International Science and Engineering Fair (ISEF) in Albuquerque. Dr. Glass visited me at my booth at ISEF, and we talked about my work. I spent most of the next summer thinking about the project’s flaws and possible improvements. If I could design a simulation in a more common language such as C++ or Java, I could better predict outcomes in a given epidemic. I decided to completely redefine my modeling program for the tenth grade science fair. With Dr. Erdogan’s help, I ended up with a rather lengthy C++ program that was by far more accurate in terms of simulating an epidemic. It used a social network model, in which each person is represented as a “node” of a graph, and their social connections as “links” between nodes. This allowed me to directly observe interactions and possible opportunities for infection, rather than estimate rates through particle collisions. But just like the particlebased simulation, it concluded that children are the most important group to target with intervention. It was immensely more rewarding to have programmed January/February 009 The stages of infection that an individual may pass through. my own model from scratch. The difference between the NetLogo model and the C++ model is like the difference between engineering with LEGO blocks and “real” engineering. With the malleability of my new model, I was able to show that if a disease kills people quickly—as does the avian flu—the overall mortality rate will be lower than with seasonal influenza. This is likely because there are fewer sick or immune people around to spread viral particles. My project placed second in the senior division at the state science fair. Once again I was able to attend ISEF in Atlanta, where I won a second place grand award in the Medicine and Health category. Real-Life Applications Even more than the prizes, I value what I learned. I can see how my work can lead to proactive decisions that can save many lives. Recently, I submitted a paper on my work to the Pacific Symposium. I am currently elaborating on my model by accounting for localized variables, such as the presence of hospitals or schools. I also hope to include multiple cities over a span of multiple years to investigate how the mutability of influenza and other fast-mutating diseases would affect epidemic dynamics. Eventually, I plan to attend medical school and pursue a career in medical research. i Nolan Kamitaki is a junior at Waiakea High School in Hilo, HI. In addition to simulating pandemics, Nolan plays the piano and is active in student government, Science Bowl, and Acadec. He is also host of “Nolan Knows,” a monthly segment on the local cable TV show Living in Paradise. imagine 15

Table of Contents for the Digital Edition of Imagine Magazine - Johns Hopkins - January/February 2009

Imagine Magazine - John Hopkins - January/February 2009
Contents
Letters
Big Problems
In My Own Words
Witness to a Pandemic
An Ounce of Prevention
Of Math and Measles
Predicting the Next Pandemic
Medicine, Medicine Everywhere
Food Matters
Looking for a Challenge? Try Summer College!
CTY: The Real Deal
Hot Topic
Selected Opportunities & Resources
Off the Shelf
Word Wise
Middle Ground
One Step Ahead
Exploring Career Options
Planning Ahead for College
Students Review
Creative Minds Imagine
Sudoku
Knossos Games

Imagine Magazine - Johns Hopkins - January/February 2009