Des Moines Register: Simulator's Aim Is Aiding Breast Cancer Detection

Wednesday, December 10, 2008

By Dawn Sagario
Des Moines Register

A University of Iowa professor is developing a device that he hopes will better train doctors in detecting breast cancer.

Geb Thomas' research involves creating prosthetic breasts made with simulated tumors embedded in them, which doctors can use when practicing how to perform clinical breast exams. The lumps are made of special balloons filled with water, and each can be inflated and deflated in real-time to create lumps of varying size and depth within the simulated tissue.

"We call these 'dynamic tumors,' " said Thomas, associate professor of industrial engineering in the U of I's Department of Mechanical and Industrial Engineering.

This training device is distinct from other "static" models, with fixed tumors, Thomas said. That's because once doctors find the tumors, they've already memorized the locations on the prosthetic device.

A more reliable exam

"The idea behind this work is that if doctors got to practice more with a device like this, they could detect more tumors, so it could increase the reliability (of the clinical breast exam)," said Thomas, who has conducted research in virtual reality and robotics. "Mostly what I'm doing is trying to take this and make it practical enough so it can actually be used."

Thomas, along with U of I colleagues from the Family Medicine and Biomedical Engineering departments, published a study in 2003 in the journal Cancer Detection and Prevention. The results found that training on the dynamic breast model led to participants finding more lumps compared with training on the static model.

The dynamic breast model in the study included lumps that varied in size, depth of placement and mobility. Each lump could be individually inflated.

There have been five prototypes in the last five years, Thomas said. The current model has 10 balloons.

Gregory Gerling, whose master's thesis at the U of I was on the dynamic breast model, was the lead author of the study. Gerling has since moved to the University of Virginia in Char- lottesville, where he is an assistant professor of systems and information engineering.

For breast cancer, the five-year survival rate is 98 percent if tumors are detected before they grow to 2 centimeters in diameter, Gerling said. "We're trying to train people to detect 0.5 centimeter, at the smallest," he said.

The value of touch

Gerling's research is currently focused on a prostate simulator, which is being used at the medical school at the University of Virginia, he said.

The trend seems to be moving away from tests, such as the clinical breast exam, where physicians touch the patient, Thomas said.

"One of the things that keeps me curious is that the brain has a huge amount of matter devoted to the sense of touch," he said. "But we as people devote so much to the sense of sight ... that sometimes I think we completely underestimate the power of what we feel. It may be that we're overlooking some possibilities that just require manual skill."

Dr. Roberta Wattleworth, chairwoman of the Department of Family Medicine at Des Moines University, said the device could be especially helpful to new physicians or those in training.

Students currently practice their skills using different breast models, many made of rubber, Wattleworth said. "When they (students) get to their third and fourth year, they're doing more live palpation, with a preceptor following them, so they don't use the models as much anymore," she said.

Wattleworth also thinks the device could be a helpful educational tool in the community, teaching women what a cyst or cancerous tumor might feel like.

"For those women who don't get routine mammograms, they find the tumor more often than I do on the mammogram," Wattleworth said of women she saw in her rural practice. At least 60 percent of patients she saw with a mass found it during a self-exam.

Medical students may be using the dynamic model in a couple of years, Thomas said, but the device first needs to be made affordable, reliable and realistic enough to be an effective simulator before it's made widely available.