Department of Radiation Oncology and Winship Cancer Institute, Emory University, Atlanta, GA, 30322, USA.
Med Phys. 2020 Apr;47(4):1763-1774. doi: 10.1002/mp.14065. Epub 2020 Feb 26.
To develop an accurate and fast deformable image registration (DIR) method for four-dimensional computed tomography (4D-CT) lung images. Deep learning-based methods have the potential to quickly predict the deformation vector field (DVF) in a few forward predictions. We have developed an unsupervised deep learning method for 4D-CT lung DIR with excellent performances in terms of registration accuracies, robustness, and computational speed.
A fast and accurate 4D-CT lung DIR method, namely LungRegNet, was proposed using deep learning. LungRegNet consists of two subnetworks which are CoarseNet and FineNet. As the name suggests, CoarseNet predicts large lung motion on a coarse scale image while FineNet predicts local lung motion on a fine scale image. Both the CoarseNet and FineNet include a generator and a discriminator. The generator was trained to directly predict the DVF to deform the moving image. The discriminator was trained to distinguish the deformed images from the original images. CoarseNet was first trained to deform the moving images. The deformed images were then used by the FineNet for FineNet training. To increase the registration accuracy of the LungRegNet, we generated vessel-enhanced images by generating pulmonary vasculature probability maps prior to the network prediction.
We performed fivefold cross validation on ten 4D-CT datasets from our department. To compare with other methods, we also tested our method using separate 10 DIRLAB datasets that provide 300 manual landmark pairs per case for target registration error (TRE) calculation. Our results suggested that LungRegNet has achieved better registration accuracy in terms of TRE than other deep learning-based methods available in the literature on DIRLAB datasets. Compared to conventional DIR methods, LungRegNet could generate comparable registration accuracy with TRE smaller than 2 mm. The integration of both the discriminator and pulmonary vessel enhancements into the network was crucial to obtain high registration accuracy for 4D-CT lung DIR. The mean and standard deviation of TRE were 1.00 ± 0.53 mm and 1.59 ± 1.58 mm on our datasets and DIRLAB datasets respectively.
An unsupervised deep learning-based method has been developed to rapidly and accurately register 4D-CT lung images. LungRegNet has outperformed its deep-learning-based peers and achieved excellent registration accuracy in terms of TRE.
开发一种用于四维 CT（4D-CT）肺部图像的精确、快速的变形图像配准（DIR）方法。基于深度学习的方法有可能在几次正向预测中快速预测变形向量场（DVF）。我们已经开发了一种用于 4D-CT 肺部 DIR 的无监督深度学习方法，在配准精度、鲁棒性和计算速度方面都具有出色的性能。
使用深度学习提出了一种快速准确的 4D-CT 肺部 DIR 方法，即 LungRegNet。LungRegNet 由两个子网组成，分别是 CoarseNet 和 FineNet。顾名思义，CoarseNet 在粗尺度图像上预测大的肺运动，而 FineNet 在细尺度图像上预测局部肺运动。CoarseNet 和 FineNet 都包括一个生成器和一个鉴别器。生成器被训练直接预测 DVF 以变形运动图像。鉴别器被训练来区分变形图像和原始图像。首先训练 CoarseNet 来变形运动图像。然后将变形图像用于 FineNet 进行 FineNet 训练。为了提高 LungRegNet 的配准精度，我们在网络预测之前通过生成肺部血管概率图来生成血管增强图像。
我们在我们部门的十个 4D-CT 数据集上进行了五重交叉验证。为了与其他方法进行比较，我们还使用 DIRLAB 数据集的十个独立数据集测试了我们的方法，每个数据集为目标配准误差（TRE）计算提供 300 个手动标志点对。我们的结果表明，LungRegNet 在 TRE 方面达到了比文献中其他基于深度学习的方法更好的配准精度。与传统的 DIR 方法相比，LungRegNet 可以生成具有小于 2mm 的 TRE 的可比配准精度。将鉴别器和肺部血管增强都集成到网络中对于获得 4D-CT 肺部 DIR 的高精度是至关重要的。在我们的数据集和 DIRLAB 数据集上，TRE 的平均值和标准差分别为 1.00±0.53mm 和 1.59±1.58mm。
已经开发了一种基于无监督深度学习的方法来快速准确地配准 4D-CT 肺部图像。LungRegNet 优于其基于深度学习的同行，并在 TRE 方面达到了出色的配准精度。
