A method of rapid quantification of patient-specific organ doses for CT using deep-learning-based multi-organ segmentation and GPU-accelerated Monte Carlo dose computing
A method of rapid quantification of patient-specific organ doses for CT using deep-learning-based multi-organ segmentation and GPU-accelerated Monte Carlo dose computing
复制标题
DOI：
10.1002/mp.14131
复制
发表时间：
2020-04-03
期刊：
影响因子：
3.8
通讯作者：
Xu, X. George
中科院分区：
文献类型：
--
作者：
Peng, Zhao;Fang, Xi;Xu, X. George
目的：患者特异性计算机断层扫描（CT）剂量测定的一个技术障碍是缺乏用于CT图像的自动患者特异性多器官分割和快速器官剂量定量的计算工具。当先前的CT图像可用于患者的相同身体区域时，以与放射疗法治疗规划类似的方式获得用于CT的患者特定器官剂量的能力将为个性化和前瞻性CT扫描协议打开大门。该研究旨在证明将深度学习算法用于从CT图像中自动分割多个放射敏感器官与基于GPU的Monte Carlo快速器官剂量计算相结合的可行性。方法开发并训练基于U-Net的深度卷积神经网络（CNN）用于器官分割，以自动从CT图像中描绘多个放射敏感器官。使用两个数据库：肺部CT分割挑战2017（LCTSC）数据集包含60名胸部CT扫描患者，每个患者由5个分割器官组成，胰腺CT（PCT）数据集包含43名腹部CT扫描患者，每个患者由8个分割器官组成。对两组数据进行五重交叉验证方法。Dice相似系数（DSC）用于评估分割性能对地面真相。基于GPU的蒙特卡罗剂量代码ARCHER用于计算患者特异性CT器官剂量。所提出的方法进行评估的相对剂量误差（RDES）。为了证明新方法的潜在改进，将器官剂量结果与马萨诸塞州总医院的离线剂量报告软件VirtualDose中使用的人群平均患者体模获得的结果进行比较。（右肺），0.96（左肺），0.92（心脏），0.86（脊髓），0.76（食管），沿着为0.96（脾）、0.96（肝）、0.95（肝）对于PCT数据集，0.90（胃）、0.87（胆囊）、0.80（胰腺）、0.75（食道）和0.61（十二指肠）。与来自群体平均体模的器官剂量结果相比，新的患者特异性方法实现了更小的绝对RDE（平均值+/-标准差）：1.8% +/- 1.4%（vs 16.0% +/- 11.8%）肺，0.8% +/- 0.7%（vs 34.0% +/- 31.1%）心脏，1.6% +/- 1.7%（vs 45.7% +/- 29.3%）食管，0.6% +/- 1.2%（vs 15.8% +/- 12.7%）脾脏，1.2% +/- 1.0%胰腺（vs 18.1% +/- 15.7%），0.9% +/- 0.6%（vs 20.0% +/- 15.2%）左肾，1.7% +/- 3.1%（vs 19.1% +/- 9.8%），肝脏0.3% +/- 0.3%（vs 24.2% +/- 18.7%），胃1.6% +/- 1.7%（vs 19.3% +/- 13.6%）。该工作显示了以临床可接受的准确度和效率执行CT图像的患者特定多器官自动分割和基于GPU的快速蒙特卡罗剂量量化的可行性。
Purpose One technical barrier to patient-specific computed tomography (CT) dosimetry has been the lack of computational tools for the automatic patient-specific multi-organ segmentation of CT images and rapid organ dose quantification. When previous CT images are available for the same body region of the patient, the ability to obtain patient-specific organ doses for CT - in a similar manner as radiation therapy treatment planning - will open the door to personalized and prospective CT scan protocols. This study aims to demonstrate the feasibility of combining deep-learning algorithms for automatic segmentation of multiple radiosensitive organs from CT images with the GPU-based Monte Carlo rapid organ dose calculation.Methods A deep convolutional neural network (CNN) based on the U-Net for organ segmentation is developed and trained to automatically delineate multiple radiosensitive organs from CT images. Two databases are used: The lung CT segmentation challenge 2017 (LCTSC) dataset that contains 60 thoracic CT scan patients, each consisting of five segmented organs, and the Pancreas-CT (PCT) dataset, which contains 43 abdominal CT scan patients each consisting of eight segmented organs. A fivefold cross-validation method is performed on both sets of data. Dice similarity coefficients (DSCs) are used to evaluate the segmentation performance against the ground truth. A GPU-based Monte Carlo dose code, ARCHER, is used to calculate patient-specific CT organ doses. The proposed method is evaluated in terms of relative dose errors (RDEs). To demonstrate the potential improvement of the new method, organ dose results are compared against those obtained for population-average patient phantoms used in an off-line dose reporting software, VirtualDose, at Massachusetts General Hospital.Results The median DSCs are found to be 0.97 (right lung), 0.96 (left lung), 0.92 (heart), 0.86 (spinal cord), 0.76 (esophagus) for the LCTSC dataset, along with 0.96 (spleen), 0.96 (liver), 0.95 (left kidney), 0.90 (stomach), 0.87 (gall bladder), 0.80 (pancreas), 0.75 (esophagus), and 0.61 (duodenum) for the PCT dataset. Comparing with organ dose results from population-averaged phantoms, the new patient-specific method achieved smaller absolute RDEs (mean +/- standard deviation) for all organs: 1.8% +/- 1.4% (vs 16.0% +/- 11.8%) for the lung, 0.8% +/- 0.7% (vs 34.0% +/- 31.1%) for the heart, 1.6% +/- 1.7% (vs 45.7% +/- 29.3%) for the esophagus, 0.6% +/- 1.2% (vs 15.8% +/- 12.7%) for the spleen, 1.2% +/- 1.0% (vs 18.1% +/- 15.7%) for the pancreas, 0.9% +/- 0.6% (vs 20.0% +/- 15.2%) for the left kidney, 1.7% +/- 3.1% (vs 19.1% +/- 9.8%) for the gallbladder, 0.3% +/- 0.3% (vs 24.2% +/- 18.7%) for the liver, and 1.6% +/- 1.7% (vs 19.3% +/- 13.6%) for the stomach. The trained automatic segmentation tool takes This work shows the feasibility to perform combined automatic patient-specific multi-organ segmentation of CT images and rapid GPU-based Monte Carlo dose quantification with clinically acceptable accuracy and efficiency.