Title:
Use of Large Scale Parallel Computing for Patient Specific Dosimetry in Radionuclide Therapy

Speaker:
Yuni Dewaraja, PhD Research Assistant Professor Division of Nuclear Medicine University of Michigan

Date/Time:
June 7, 3pm

Location:
SDSC Auditorium

Abstract :
The Monte Carlo method is ideal for modeling Nuclear Medicine imaging and for accurate calculation of radiation adsorbed dose because of the stochastic nature of radiation emission, transport and detection processes. Our application is in tumor/organ absorbed dose calculation following I-131 radioimmunotherapy, which is showing great promise in the treatment of non-Hodgkins Lymphoma. If our goal of highly accurate dosimetry is achieved in the future oncologists can potentially modify existing radioimmunotherapy schemes to tailor the treatment to a specific patient, based on radiation absorbed dose. To work towards the goal of highly accurate dosimetry we have extensively used Monte Carlo methods to evaluate and implement patient specific techniques for I-131 activity quantification based on single photon emission computed tomography (SPECT). In these studies parallel processing has been essential to obtain statistically acceptable SPECT simulations for evaluations using clinically realistic voxel-based phantoms or for patient specific calculation. We have recently implemented a Monte Carlo algorithm to carry out highly patient specific 3D calculation of absorbed dose distribution based on the patients measured SPECT activity distribution and anatomical information from co-registered x-ray CT images. The first clinical dose distribution images from these calculations will be presented and the role of Monte Carlo and parallel computing in our application will be discussed.