交错 3D-CNN 用于头颈部 CT 图像中小体积结构的联合分割。
Interleaved 3D-CNNs for joint segmentation of small-volume structures in head and neck CT images.
机构信息
School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai, 200030, China.
Department of Radiology and BRIC, University of North Carolina at Chapel Hill, Chapel Hill, NC, 27599, USA.
出版信息
Med Phys. 2018 May;45(5):2063-2075. doi: 10.1002/mp.12837. Epub 2018 Mar 23.
PURPOSE
Accurate 3D image segmentation is a crucial step in radiation therapy planning of head and neck tumors. These segmentation results are currently obtained by manual outlining of tissues, which is a tedious and time-consuming procedure. Automatic segmentation provides an alternative solution, which, however, is often difficult for small tissues (i.e., chiasm and optic nerves in head and neck CT images) because of their small volumes and highly diverse appearance/shape information. In this work, we propose to interleave multiple 3D Convolutional Neural Networks (3D-CNNs) to attain automatic segmentation of small tissues in head and neck CT images.
METHOD
A 3D-CNN was designed to segment each structure of interest. To make full use of the image appearance information, multiscale patches are extracted to describe the center voxel under consideration and then input to the CNN architecture. Next, as neighboring tissues are often highly related in the physiological and anatomical perspectives, we interleave the CNNs designated for the individual tissues. In this way, the tentative segmentation result of a specific tissue can contribute to refine the segmentations of other neighboring tissues. Finally, as more CNNs are interleaved and cascaded, a complex network of CNNs can be derived, such that all tissues can be jointly segmented and iteratively refined.
RESULT
Our method was validated on a set of 48 CT images, obtained from the Medical Image Computing and Computer Assisted Intervention (MICCAI) Challenge 2015. The Dice coefficient (DC) and the 95% Hausdorff Distance (95HD) are computed to measure the accuracy of the segmentation results. The proposed method achieves higher segmentation accuracy (with the average DC: 0.58 ± 0.17 for optic chiasm, and 0.71 ± 0.08 for optic nerve; 95HD: 2.81 ± 1.56 mm for optic chiasm, and 2.23 ± 0.90 mm for optic nerve) than the MICCAI challenge winner (with the average DC: 0.38 for optic chiasm, and 0.68 for optic nerve; 95HD: 3.48 for optic chiasm, and 2.48 for optic nerve).
CONCLUSION
An accurate and automatic segmentation method has been proposed for small tissues in head and neck CT images, which is important for the planning of radiotherapy.
目的
准确的三维图像分割是头颈部肿瘤放射治疗计划的关键步骤。这些分割结果目前是通过组织的手动轮廓来获得的，这是一个繁琐且耗时的过程。自动分割提供了一种替代方法，然而，由于小组织（即头颈部 CT 图像中的视交叉和视神经）体积小且外观/形状信息高度多样化，因此对于小组织通常很难实现。在这项工作中，我们提出了一种交错多个 3D 卷积神经网络（3D-CNN）的方法，以实现对头颈部 CT 图像中小组织的自动分割。
方法
设计了一个 3D-CNN 来分割每个感兴趣的结构。为了充分利用图像的外观信息，提取多尺度斑块来描述所考虑的中心体素，然后将其输入到 CNN 架构中。接下来，由于在生理和解剖学方面，相邻组织通常高度相关，我们交错用于各个组织的 CNN。通过这种方式，特定组织的初步分割结果可以有助于细化其他相邻组织的分割。最后，随着更多的 CNN 交错和级联，可以得到一个复杂的 CNN 网络，从而可以联合分割和迭代细化所有组织。
结果
我们的方法在一组 48 张 CT 图像上进行了验证，这些图像来自于医学图像计算和计算机辅助干预（MICCAI）挑战赛 2015。使用 Dice 系数（DC）和 95%Hausdorff 距离（95HD）来衡量分割结果的准确性。与 MICCAI 挑战赛的获胜者（DC：视交叉为 0.38，视神经为 0.68；95HD：视交叉为 3.48，视神经为 2.48）相比，所提出的方法实现了更高的分割准确性（DC：视交叉为 0.58±0.17，视神经为 0.71±0.08；95HD：视交叉为 2.81±1.56mm，视神经为 2.23±0.90mm）。
结论
提出了一种用于头颈部 CT 图像中小组织的准确自动分割方法，这对头颈部肿瘤放射治疗计划非常重要。
Zotero 插件