부산대학교 도서관

Patient-specific internal dosimetry with high accuracy is the most significant issue in the field of nuclear medicine. In recent researches has been a dramatic alter in different methods to compute the correct organ doses according to injected radioactivity distribution, admittedly Monte Carlo simulation has brought high accuracy results in voxel-based dosimetry techniques.In this study patient datasets who were injected F18-FDG (2 subjects, normal cases, 60±8.2 kg, injected activity:10 ±0.5 Mbq) were acquired in 3 times sequences (30,60,90 min post-injection). The CT and PET images were used as attenuation maps and activity distribution of patient phantom respectively. Hence the CT and PET images have been registered by 3D Slicer software to achieve the same matrix and pixel sizes. The regions of interest were segmented on the CT images that entailed kidneys, spleen, bladder, lung, pancreas, liver, stomach, gallbladder organs. The segmenting ROI of CT images were reconstructed by MATLAB codes and the voxelized phantom and voxelized source of each patient at specific times were generatedBy way of conclusion, the F18-FDG dose distribution in patient-specific phantom has investigated and bladder, spleen, kidney absorbed the most activity as we expected by PET images, the mean S- factors and absorbed dose were computed as 4.02e-05 mGy/Mbq & 0.67±1.2 mGy, 5.29e-07 mGy/Mbq & 1.67±0.45mGy, 2.23e-07 mGy/Mbq & 0.57±0.52 mGy for bladder, kidney and spleen organs. Monte Carlo methods are shown the best results aspect of accuracy but the essential issue that prevents clinical uses is excessive computational cost and time.