|
Cover Story Sidebar |
|
|
Calculating Dosage
Computational methods will allow by T.J. BECKER | |
|
RESEARCHERS AT GEORGIA TECH and Emory University are combining their expertise in nuclear engineering and medical imaging to advance dosage accuracy in radiation therapy.
“Calculations in radiation dose have not kept pace with improvements in delivery systems,” says Farzad Rahnema, chairman of Georgia Tech’s Nuclear and Radiological Engineering/Medical Physics Program in the Woodruff School of Mechanical Engineering.
|
Rahnema is working with Tim Fox, director of Emory’s Medical Physics Division in the Department of Radiation Oncology, to develop a new computational tool that will more precisely calculate radiation dose and where it will be deposited.
Existing methods don’t account for the intricate geometry of the human body; different tissues, air cavities and bones affect how radiation deposits its energy. Also, a patient’s posture and diet prior to treatment can also influence the location of tumor.
“If we can track tumors more accurately, then the number of treatments can be reduced, which would allow radiation dose to be escalated,” Fox explains. “This would improve tumor control while reducing dose to normal tissue structures.”
With funding from the Georgia Cancer Coalition, Rahnema and Fox are developing a tool that combines the strengths of two different computational methods – stochastic (Monte Carlo) and advanced deterministic (coarse mesh transport) methods.
Their idea: Divide patient anatomy and the radiation beam head into subregions. Stochastic algorithms, which can handle complex geometric problems, will be used to predict how each subregion responds to the flow of radiation and how photons and electrons exit the area.
Because stochastic methods require long running times to reduce error from statistical noise,
This illustration shows shaped radiation beams used to treat a brain tumor with external beam radiation. Researchers are developing a new computational tool that will more precisely calculate radiation dose in procedures that use this technology for cancer treatment. (300-dpi JPEG version - 507k)
“We’ve had good success using these methods in 2-D nuclear criticality problems,” Rahnema says. “Now we’ll extend them to 3-D geometry to handle the radiation therapy problem.”
One challenge will be working with a different type of radiation. “With nuclear reactors, we’re only dealing with neutrons,” Rahnema explains. “With radiation therapy, we’re working with electrons and photons, which have different characteristics. Since the radiation types are different, it might create different numerical problems.”
Yet Rahnema and Fox are optimistic that their system could reduce dose errors to less than 2 percent. (Current methods are inaccurate by 5 to10 percent, depending on where the tumor is located in the body.)
In addition to increasing dose accuracy, this software tool would enable treatment planning to become more efficient.
“That’s significant because improved accuracy and speed don’t usually go hand in hand,” Fox adds.
this data will be computed in advance and stored in a data library. Then deterministic methods, which are more efficient, will use the library to tie subregions together and calculate a precise dose distribution in the tumor and surrounding tissues.
courtesy of Tim Fox
For more information, contact Farzad Rahnema at 404-894-3731 or farzad.rahnema@nre.gatech.edu; or Tim Fox at 404-778-2304 or tim@radonc.emory.org.
Contents
Research Horizons
GT Research News
GTRI
Georgia Tech
Send questions and comments regarding these pages to
webadmin@edi.gatech.edu
Last updated: July 7, 2004
