利用蒙地卡羅技術評估三層乳房假體於乳房攝影之乳腺劑量

Estimation of normalized glandular doses using a three layer mammographic phantom: a Monte Carlo simulation study

陳若嵐(Ruo-lan Chen)

指導教授 ： 吳杰

中臺科技大學/健康科學院/放射科學研究所/碩士(2010年)

摘要

臨床上常利用平均乳腺劑量（Normalized average glandular dose, DgN）評估乳房X光攝影所造成的能量沉積,其中轉換因子的取得,多為模擬均勻混合假體的乳房劑量,然而,乳房組織並非均勻分佈,會造成轉換因子的評估錯誤。本研究依據乳房組織的分佈開發了三層混合假體進行模擬,使得轉換因子的評估更準確。
研究中採用MCNP程式（Monte Carlo N-Particles code）建立照影環境的幾合型狀,其中乳房X光攝影射源參照Boone於1997年提出的能譜,並選擇臨床常用的靶極組合（Mo/Mo、Mo/Rh、Rh/Rh）,管電壓（25~35kVp）與對應的六個半值層等條件,並在不同乳房厚度（2~10 cm）的三層假體中,分別模擬25%、50%和75%腺體比例共二十七種排列組合在不同腺體比例分佈下對轉換因子的影響。實驗結果發現Mo/Mo、Mo/Rh在腺體比例為0.1%、50%和100%的情況下,模擬的轉換因子與ACR差異於6%內,Rh/Rh差異於9%內。三層假體在Mo/Mo、管電壓27kVp、半值層0.34mmAl與乳房厚度4公分的模擬中,腺體比例皆為50%的情況下,與均勻假體的轉換因子差異為13.8%。根據模擬的結果以迴歸分析找出三層假體中0~100%腺體比例下所對應的校正因子,並計算出在不同腺體組織分佈下的轉換因子。
乳腺劑量的評估會因乳房中腺體分佈的不同,而造成轉換因子的差異,藉由三層乳房假體的開發與轉換因子模擬的結果,提供臨床更快速的劑量計算並有效的提升評估的準確度。

關鍵字

平均乳腺劑量 ； 蒙地卡羅 ； 乳房X光攝影

並列摘要

The average glandular dose（DgN）estimates the energy deposition of mammography in common clinical. The Monte Carlo simulations uses uniformly mixed phantom for calculating the conversion factor. However, breast tissues are un-uniformly distributed, leading to errors of conversion factor estimates. This study used a three layer phantom that estimated more accurate of average glandular dose.
In this study, MCNP code (Monte Carlo N-Particles code) was used to created the simulation parameters and geometric structure. We simulated three type of target/filter combinations (Mo/Mo, Mo/Rh, Rh/Rh), six voltages (25 ~ 35 kVp), six HVL parameters and, nine breast phantom thickness (2 ~ 10 cm) for the three layer phantom. The conversion factor for glandular weights were 25%, 50% and 75% all of 27 type.
Experiment result found that the glandular weights were 0.1%, 50% and 100%. The conversion factor compared with the results of the American College of Radiology (ACR) was within 6%. For Rh/Rh, the difference was within 9%. The difference between the 50% average glandular weight and the uniform phantom was 7.1% ~ -6.7% at the Mo/Mo combination, voltage at 27 kVp, half value layer 0.34 mmAl, and breast thickness of 4 cm. According to simulation results, the regression analysis found that the three layer phantom at 0% ~ 100% glandular weight can calculate conversion factors accurately. The different in glandular tissue distribution leads to errors of conversion factor calculates. The three layer phantom can provide accuracy estimates of glandular doses in clinical practices.

並列關鍵字

Mammography ； Monte Carlo simulation ； Normalized average glandular dose

參考文獻

ACR 1999 Committee on Quality Assurance in Mammography Medical Physicist’s Section

行政院國民健康局 2009 民國98年死因結果摘要表

Bailar J C 1977 Screening for early breast cancer: Pros and cons Cancer 39 2783-95

Bakic P R, Albert M, Brzakovic D and Maidment A D A 2002a Mammogram synthesis using a 3D simulation. I. Breast tissue model and image acquisition simulation Medical Physics 29 2131-9

Bakic P R, Albert M, Brzakovic D and Maidment A D A 2002b Mammogram synthesis using a 3D simulation. II. Evaluation of synthetic mammogram texture Medical Physics 29 2140-51