Fighting Cancer in Four Dimensions

by Michael Mullaney on September 22, 2009

Professors George Xu and Suvranu De have been working together for some time, pairing their talent toward the shared goal of making fake medical patients more real.

There’s an easy way to explain the complex work they do: People breathe, and when we breathe our internal organs shift around a bit. Xu and De want to program a computer to accurately and reliably recreate this process.

This organ movement is a huge disadvantage for doctors who are administering radiation treatment to patients suffering from different types of cancerous tumors. Conventional medical equipment can be carefully calibrated to focus radiation on a specific spot inside the body, but the technology is not yet robust enough to refine its movements to account for the shifting of internal organs as a patient breathes.

Think of it this way. While it’s fairly easy to thread a needle, imagine trying to thread a needle that is being held by someone else – and that person has a shaky hand. Now imagine threading that needle could be a life-or-death situation.

Xu, De, and other researchers around the world have worked to develop about 20 different “phantoms” – highly detailed 3-D digital models of humans. These phantoms allow medical researchers to run simulations to accurately assess the cause-and-effect of different actions and inputs on the body. These simulations save time, money, and – one day – lives, allowing researchers to perform simulations and determine different critical data from the patient, such as, for example, calculating the best angles from which to administer radiation therapy.

Though De and Xu are primarily concerned with lung and liver cancer, they did work on a model for pregnant women, which could one day allow doctors to greatly reduce the potential effects of radiation on fetuses when treating cancer in expecting mothers. Check out a short video of Xu talking about this project.

Adding the fourth dimension of time to these models is immensely challenging. It’s not just a matter of modeling the entire chest cavity, but rather modeling each and every organ in the human body – all of which have a different shape, surface texture, and move mostly independently as you breathe.  It involves a whole lot of coding and developing algorithms to accurately simulate the moving parts of this human-sized and human-shaped puzzles.

Further reading on De’s and Xu’s research can be found here and here. A new research paper they wrote was also featured online recently by one of their industry affiliates.