剂量学
内部剂量学
医学物理学
内辐射
软件
核医学
核工程
计算机科学
医学
工程类
操作系统
作者
Eero Hippeläinen,Mikko Tenhunen,Hanna Mäenpää,Jorma Heikkonen,Antti Sohlberg
出处
期刊:Nuclear Medicine Communications
[Ovid Technologies (Wolters Kluwer)]
日期:2017-05-01
卷期号:38 (5): 357-365
被引量:22
标识
DOI:10.1097/mnm.0000000000000662
摘要
The aim of this work is to validate a software package called Hermes Internal Radiation Dosimetry (HIRD) for internal dose assessment tailored for clinical practice. The software includes all the necessary steps to perform voxel-level absorbed dose calculations including quantitative reconstruction, image coregistration and volume of interest tools.The basics of voxel-level dosimetry methods and implementations to HIRD software are reviewed. Then, HIRD is validated using simulated SPECT/CT data and data from Lu-DOTATATE-treated patients by comparing absorbed kidney doses with OLINDA/EXM-based dosimetry. In addition, electron and photon dose components are studied separately in an example patient case.The simulation study showed that HIRD can reproduce time-activity curves accurately and produce absorbed doses with less than 10% error for the kidneys, liver and spleen. From the patient data, the absorbed kidney doses calculated using HIRD and using OLINDA/EXM were highly correlated (Pearson's correlation coefficient, r=0.98). From Bland-Altman plot analysis, an average absorbed dose difference of -2% was found between the methods. In addition, we found that in Lu-DOTATATE-treated patients, photons can contribute over 10% of the kidney's total dose and is partly because of cross-irradiation from high-uptake lesions close to the kidneys.HIRD is a straightforward voxel-level internal dosimetry software. Its clinical utility was verified with simulated and clinical Lu-DOTATATE-treated patient data. Patient studies also showed that photon contribution towards the total dose can be relatively high and voxel-level dose calculations can be valuable in cases where the target organ is in close proximity to high-uptake organs.
科研通智能强力驱动
Strongly Powered by AbleSci AI