Department of Radiation Oncology, Washington University in Saint Louis, St. Louis, MO, USA.
Med Phys. 2019 Oct;46(10):4490-4501. doi: 10.1002/mp.13726. Epub 2019 Aug 7.
To automatically and precisely detect a large quantity of landmark pairs between two lung computed tomography (CT) images to support evaluation of deformable image registration (DIR). We expect that the generated landmark pairs will significantly augment the current lung CT benchmark datasets in both quantity and positional accuracy.
A large number of landmark pairs were detected within the lung between the end-exhalation (EE) and end-inhalation (EI) phases of the lung four-dimensional computed tomography (4DCT) datasets. Thousands of landmarks were detected by applying the Harris-Stephens corner detection algorithm on the probability maps of the lung vasculature tree. A parametric image registration method (pTVreg) was used to establish initial landmark correspondence by registering the images at EE and EI phases. A multi-stream pseudo-siamese (MSPS) network was then developed to further improve the landmark pair positional accuracy by directly predicting three-dimensional (3D) shifts to optimally align the landmarks in EE to their counterparts in EI. Positional accuracies of the detected landmark pairs were evaluated using both digital phantoms and publicly available landmark pairs.
Dense sets of landmark pairs were detected for 10 4DCT lung datasets, with an average of 1886 landmark pairs per case. The mean and standard deviation of target registration error (TRE) were 0.47 ± 0.45 mm with 98% of landmark pairs having a TRE smaller than 2 mm for 10 digital phantom cases. Tests using 300 manually labeled landmark pairs in 10 lung 4DCT benchmark datasets (DIRLAB) produced TRE results of 0.73 ± 0.53 mm with 97% of landmark pairs having a TRE smaller than 2 mm.
A new method was developed to automatically and precisely detect a large quantity of landmark pairs between lung CT image pairs. The detected landmark pairs could be used as benchmark datasets for more accurate and informative quantitative evaluation of DIR algorithms.
自动、准确地检测大量肺 CT 图像之间的 landmark 对，以支持形变图像配准（DIR）的评估。我们期望生成的 landmark 对在数量和位置精度方面都能极大地扩充现有的肺 CT 基准数据集。
在肺的 4DCT 数据集的呼气末（EE）和吸气末（EI）相位之间，大量 landmark 对在肺内被检测到。数千个 landmark 是通过对肺血管树概率图应用 Harris-Stephens 角点检测算法检测到的。通过将 EE 和 EI 相位的图像进行参数图像配准（pTVreg），建立初始 landmark 对应关系。然后，开发了一个多流伪孪生（MSPS）网络，通过直接预测三维（3D）位移来进一步提高 landmark 对的位置精度，以最优地将 EE 中的 landmark 与其 EI 中的对应点对齐。使用数字体模和公开的 landmark 对评估了检测到的 landmark 对的位置精度。
10 个 4DCT 肺数据集检测到了密集的 landmark 对集，平均每个病例有 1886 个 landmark 对。对于 10 个数字体模病例，98%的 landmark 对的目标配准误差（TRE）小于 2mm，平均 TRE 为 0.47 ± 0.45mm。在 10 个肺 4DCT 基准数据集（DIRLAB）中使用 300 个手动标记的 landmark 对进行的测试，产生的 TRE 结果为 0.73 ± 0.53mm，97%的 landmark 对的 TRE 小于 2mm。
开发了一种新方法，可自动、准确地检测肺 CT 图像对之间的大量 landmark 对。检测到的 landmark 对可用作基准数据集，以更准确和有意义地评估 DIR 算法。
