Electronic Thesis and Dissertation Repository

Thesis Format



Master of Science


Medical Biophysics


Lee, Ting Yim

2nd Supervisor

Hajdok, George



Targeted radionuclide therapy (TRT) is an effective treatment for metastatic prostate cancer. This thesis develops a convolution-based dosimetric method for TRT and validates its performance against a Monte Carlo (MC) simulation method, egs_mird, developed in our laboratory.

The research completed includes: 1) validating egs_mird by generating dose point kernels (DPKs) for 90Y, 131I and 177Lu and comparing them with MCNP4C, PENELOPE and GEANT4 DPK in literature, and 2) comparing TRT dosimetry of prostate cancer patients obtained using the convolution method and egs_mird.

egs_mird DPKs for 90Y and 131I agreed with those from other MC codes. The discrepancy between 177Lu DPK and literature was due to differences in emission spectra used in the MC simulation. The mean doses in the prostate and critical organs as evaluated by the convolution method were 5-7% lower than egs_mird due to density inhomogeneities.

Summary for Lay Audience

Prostate cancer (PCa) is the most commonly diagnosed cancer in men and the second leading cause of cancer-related deaths in Canada. Besides standard treatment methods such as surgery, hormone therapy, and chemotherapy, targeted radionuclide therapy (TRT) is a newer effective method to treat PCa that has spread to other parts of the body. TRT is a form of “internal” radiotherapy where the radiation is guided to the tumor by molecules that specifically target cancer cells. As in radiotherapy TRT requires an accurate calculation of the radiation dose to the tumor and surrounding healthy tissues to maximize the success of killing the tumor and minimize the chances of complications. Current TRT dose calculation uses a whole tumor/organ approach which is not accurate. The more accurate approach, called Monte Carlo simulation, requires a long calculation time so it is not practicable in clinical use. This thesis develops a new dose calculation method, called 3D convolution, which is faster and has similar accuracy as the MC simulation method.

The accuracy of the 3D convolution method was first tested in the idealized situation of calculating the dose in a material of the same density surrounding a radiation source. The 3D convolution results agreed with those obtained with MC simulation. Then dose calculations in PCa patients were compared. The mean dose in the prostate and healthy organs as evaluated by the 3D convolution method was 5-7% lower than the MC simulation. The likely explanation for this discrepancy is that MC simulation is able to account for density variations within the human body whereas the 3D convolution method assumes no density variations.