Abstract
There are a growing number of patients undergoing nuclear medicine examination each year. It is estimated that the committee effective doses of patients contributed from nuclear medicine examinations are as high as those contributed from sectional computed tomography (CT) scan. Thus, the estimation dose of nuclear medicine examinations is an important issue. At present, the Medical Internal Radiation Dose Committee system (MIRDOSE) and OLINDA/EXM v1.0 (Organ Level INternal Dose Assessment/EXponential Modeling) are commonly toolkits for calculating radiation dose to selected organs and whole body from internally administered radioisotopes. The MIRDOSE calculates the dose based on referenced human body and modeled tissue organs. They are not suitable for individual dose estimation. In this dissertation, we developed and validated a patient-specific dosimetry system namely SimDOSE that enabled the estimation of the body dose based on the static nuclear image. The time-activity curves (TAC) of all source organs are required before one can estimate the body dose during the period of examination. However, TAC is known to be difficult to obtain for each individual. Therefore, we propose a novel method to estimate the total dose from the dose measured outside the body without resorting to the TAC information. The method involves two steps. First, to compute the S values on the body surface for each source organ based on nuclear medicine image using SimDOSE. Second, to measure the dose externally from the TLD placed on the body surface during nuclear examination. Since the doses in TLD are contributed by the radiations from all source organs, they can be expressed by simultaneous equations with the S values as known variables and the cumulative activities of source organs unknown. Solving the simultaneous equations, the cumulative activities of all source organs can be obtained and subsequently the total body dose can be computed. The ORNL mathematical phantom with TAC adapted from MIRD Report 19 was simulated at eight time point (15, 30, 45, 60, 120, 180, 240 and 300 min). Eighty TLDs were placed on the phantom and the TLD readings were employed to estimate the doses at various organs. The percent sum of square errors in the estimation of cumulative activity and organ dose rate were within 4% and 1.5%, respectively. The results demonstrate the effectiveness of this method. The proposed method can be used to estimate patient-specific dose of nuclear medicine examination in PET/CT and should be used to monitor the TAC of radiopharmaceutical in tumor for the treatment dose in nuclear medicine therapy. At last, a phantom study was performed to prove the feasibility of the proposed method. A NEMA-like phantom with three cylinder inserts was administered with 99mTc pertechnetate (which activity was measured by dose calibrator). After administration, three or four sets of TLDs were placed on the surface of the phantom under 30 min exposure of radioisotope. The TLD readings (nC) were calibrated to be the absorbed dose. With a correction of factor that activity measured by dose calibrator divided by that calculated by our proposed method, the results of PSSE for the four experiments are 5.8%, 6.1%, 3.5%, 4.2%, respectively. The trend between references and our method are quite correlative and the factor (about 30%) should result from the absolute dose measurement of TLD. Preliminary results of the phantom experiments have proved the feasibility of the proposed method and yet there are still rooms for improvement. In conclusion, a novel method has proposed in the dissertation and it can provide TAC information by using Monte Carlo method combined with external dose measurement without sequential scans by PET or SPECT.