Features: Feb. 25, 2002

Tracking Tumors for Treatment
Patients who require radiation treatment may soon have less of their healthy tissue exposed to the harmful procedure.

Rich Radke, assistant professor of electrical, computer, and systems engineering, in collaboration with the Department of Radiation Therapy at the Massachusetts General Hospital in Boston, is using advanced algorithms to create more accurate images of tumors or growths in patients from the time of diagnosis to the time of treatment.

Two CT scans of the body taken when the breath is held, and then expelled.

According to Radke, doctors can accurately aim radiation beams wherever they want, but the problem is finding the exact location within the body that needs to be irradiated. Often doctors will irradiate a larger-than-necessary area because they cannot take into account the movement of the patient during X-rays or MRIs or even the movement of the tumor itself without using invasive techniques, such as inserting a metal bead in the tumor or surrounding tissue.

Variations in breathing, eating habits, or any movement can change the location of the tumor and make it difficult to pinpoint its location. Radke hopes to "train" computers to recognize images and movement from data and line things up to a template for treatment. Advances in computing technology over the past five years now make this possible. Radke envisions doctors using a high-speed computer system that reads the images, analyzes them through algorithms, and sends an accurate location directly to the radiation beam.

Radke hopes that by the end of this year he will be able to match up two-dimensional images (such as CAT scans) and three-dimensional images (such as MRIs) and that in another year he will be able to match images that take into account regular breathing patterns. The work is part of the Center for Subsurface Sensing and Imaging Systems (CenSSIS).