隨著科技的日新月異，近年來核子醫學之臨床應用相關技術發展迅速，並逐漸於醫院各科室之診療領域占有舉足輕重的地位。利用放射性核種衰變放出之能量射線診斷患處，進而產生功能性影像，或內照射治療為核子醫學之核心，然而患者在接受核醫藥物的治療或診斷時，皆會在體內、外產生輻射劑量，由此輻射劑量所造成之防護問題一直是核醫學範疇內相當重要的議題。本研究即分別針對核醫藥物產生之體內劑量與體外輻射曝露進行計算評估，並建立一套完整且具系統性之分析方法。 針對體內劑量評估問題，本研究以現行核醫界常用之MIRD方法為基礎，利用蒙地卡羅程式與擬人數值假體分別進行比吸收分率與S值之計算，以驗證程式與假體使用方法的可靠度。另選用OLINDA/EXM程式針對三種有興趣之放射性核種：鎝-99m、碘-131及氟-18標記藥物評估患者接受單次診療時可能獲得之全身有效劑量。 於體外輻射曝露之防護上，本研究不僅參考一般常用之幾何較為簡單的點、線射源模型及複雜卻真實之擬人假體模型，更引入一圓柱射源模型，並針對三種核醫藥物，加以考慮藥物於體內生理分佈情形與否，進行鄰近放射性患者之體外劑量率之估算結果比較；同時藉著圓柱射源模型之衰減、增建效應間之平衡關係討論以及計算於不同圓柱模型半徑與光子能量範圍下之評估結果，證實線射源模型在大部分之核醫藥物能量與病患體型範圍內，提供了一快速且有效之體外劑量率評估方法。此外，更透過臺灣地區碘-131患者之量測數據回顧，發現線射源模型計算值與測量值間之落差主要來自於環境散射劑量的貢獻，一旦將環境散射效應之影響加以評估進線射源模型之內，則可證實線射源模型於估算劑量率上的確具其準確性。 This study investigated both internal and external dose estimation in nuclear medicine. Three widely used radioactive nuclides 99mTc, 18F, and 131I were considered. For internal dose assessment, the MIRD schema provides a general approach for the dosimetry of incorporated radionuclides. Two crucial factors for dose estimation, the specific absorbed fraction and S-value, were calculated by Monte Carlo simulations with anthropomorphic phantoms and the results were found to be generally consistent with literature values. The OLINDA/EXM computer software was also used to estimate the resulting effective doses for patients administered three interested radiopharmaceuticals. For external radiation protection, various source models, ranging from simplified point, line, and cylinder sources to high-fidelity anthropomorphic phantoms were applied to calculate dose rates near a nuclear medicine patient. This systematical comparison led to the following observations and conclusions. The Monte Carlo calculated results based on detailed phantom models are realistic but time-consuming. The point source model is simple but too conservative while both the line and cylinder source models gave reasonable predictions. The line source model was found to be comparable to the cylinder source model due to the cancellation of radiation attenuation and buildup in the source volume. The effects of various source energies and cylinder sizes on the cancellation were investigated. The line source model, relatively easy in calculation and predicting dose rates slightly conservative than the phantom results by approximately 7-18% for the three radionuclides, is therefore considered the most practical method of dose calculation for patient release criteria. An application of the line source model to 51 post-thyroidectomy patients in Taiwan was demonstrated and its comparison with measurements was discussed. The agreement between calculations and measurements is reasonable after taking the room-scattering effect into account.