Media.Player

GPU-Based Fast Dose Calculator (GFDC) for Cancer Therapy

Detailed Technology Description

A real-time radiation therapy dose calculator.

*Abstract

Challenge

In radiation therapy, atreatment plan must be calculated for each cancer treatment that describes exactlyhow the radiation is to be delivered to the patient.  Treatment plans are produced by dosecalculators that model how the radiation is going to interact with the bodybased upon different mathematical models. More precise dose calculations provide better radiotherapy treatmentplans for cancer patients by maximizing damage to the tumor while minimizing damageto healthy tissue.

Commercial radiation treatment planning systemscommonly employ dose calculators that rely on simple models that offer more inthe way of speed rather than accuracy, resulting in especially poor performancein areas of the body where many different types of tissue are present together,such as the thorax.  Research has shownthat other models, like Monte Carlo, demonstrate a superior level of accuracyin these difficult areas. However, Monte Carlo calculations often require largecomputation times that can range from several minutes (at best) to hundreds ofhours (at worst), rendering them largely impractical for routine clinical use.

Solution

Our technology uses a unique approach to this problem:a track-repeating algorithm based on Monte Carlo calculations implemented on aGraphic Processor Unit (GPU) architecture. GPU's are the specialized processing units on video cards in moderncomputers that perform all of the complex computations required to draw theimage on the screen.  By instead usingthe GPU as a computation tool, with careful planning and development, the processingpower of a super-computer can be achieved at a fraction of the price, size, andoperational cost.  By harnessing thispower, our solution is able to aptly reproduce dose calculations from MonteCarlo systems in mere seconds.  Thistechnology can be used in traditional photon radiation therapy techniques, likebrachytherapy, as well as newer approaches, such as proton therapy.

Benefits

• Betteraccuracy than traditional radiation therapy dose calculators, resulting inbetter cancer treatment
• Fasterthan other Monte Carlo dose calculators, approaching real time
• Usingthe GPU architecture removes the need for expensive super- or cluster-computing

Features

• Capableof performing dose calculations for both photon and proton therapy
• Verifiedusing real patient data provided by UT MD Anderson Cancer Center
• Desktop-sized,GPU server running this algorithm can provide real-time dose calculations

Market Potential/ Applications

This solution would serve the well-establishedphoton therapy market, along with the growing proton therapy market, includingcancer treatment centers and hospitals. By limiting the effects of radiation treatment to healthy tissue, thistechnology helps to mitigate the risk for these centers for malpractice as aresult of various radiation sicknesses or secondary cancers that might can fromimproper dosing.

Developments andLicensing Status

A provisional patent has beenfiled for this technology jointly-owned technology with University of Texas MDAnderson Cancer Center, and it is currently available for licensing in allfields, worldwide.  Currently, aprototype system has been successfully achieved.  Rice University is actively seeking partnersto commercialize this technology or to sponsor research to bring the prototypeforward to a commercial product.

 

Contact Information

Nila Bhakuni
713-348-6231
bhakuni@rice.edu

Country/Region

USA