Skip to main content
QUICK REVIEW

[论文解读] Accurate Weakly Supervised Deep Lesion Segmentation on CT Scans: Self-Paced 3D Mask Generation from RECIST

Jinzheng Cai, Youbao Tang|arXiv (Cornell University)|Jan 25, 2018
Medical Imaging Techniques and Applications参考文献 5被引用 12
一句话总结

本文提出一种弱监督3D病灶分割方法,利用CT扫描中的RECIST直径作为弱监督信号,结合自步长CNN框架与堆叠生成对抗网络(GAN)实现超分辨率增强,以生成精确的体积分割掩膜。该方法在RECIST切片上达到93%的平均Dice系数,在3D空间中达到76%的性能,且获得92.4%的放射科医生认可度,优于基于RECIST的体积估计算法。

ABSTRACT

Volumetric lesion segmentation via medical imaging is a powerful means to precisely assess multiple time-point lesion/tumor changes. Because manual 3D segmentation is prohibitively time consuming and requires radiological experience, current practices rely on an imprecise surrogate called response evaluation criteria in solid tumors (RECIST). Despite their coarseness, RECIST marks are commonly found in current hospital picture and archiving systems (PACS), meaning they can provide a potentially powerful, yet extraordinarily challenging, source of weak supervision for full 3D segmentation. Toward this end, we introduce a convolutional neural network based weakly supervised self-paced segmentation (WSSS) method to 1) generate the initial lesion segmentation on the axial RECIST-slice; 2) learn the data distribution on RECIST-slices; 3) adapt to segment the whole volume slice by slice to finally obtain a volumetric segmentation. In addition, we explore how super-resolution images (2~5 times beyond the physical CT imaging), generated from a proposed stacked generative adversarial network, can aid the WSSS performance. We employ the DeepLesion dataset, a comprehensive CT-image lesion dataset of 32,735 PACS-bookmarked findings, which include lesions, tumors, and lymph nodes of varying sizes, categories, body regions and surrounding contexts. These are drawn from 10,594 studies of 4,459 patients. We also validate on a lymph-node dataset, where 3D ground truth masks are available for all images. For the DeepLesion dataset, we report mean Dice coefficients of 93% on RECIST-slices and 76% in 3D lesion volumes. We further validate using a subjective user study, where an experienced radiologist accepted our WSSS-generated lesion segmentation results with a high probability of 92.4%.

研究动机与目标

  • 通过临床CT扫描中弱标注的RECIST直径实现精确的3D病灶分割。
  • 通过生成完整的体积分割测量结果,克服RECIST方法仅提供2D直径估计的局限性。
  • 开发一种自步长学习框架,通过在RECIST切片上训练的CNN,逐步优化各轴向切片上的病灶分割性能。
  • 利用堆叠生成对抗网络(SGAN)生成的超分辨率图像,模拟放射科医生的放大观察行为,以提升模型性能。
  • 通过定量指标与主观放射科医生评估,验证该方法在临床中的可行性。

提出的方法

  • 该方法首先在RECIST切片上使用GrabCut等无监督方法初始化病灶分割。
  • 在RECIST切片上训练深度卷积神经网络(CNN),学习病灶外观特征,并迭代优化分割结果。
  • 模型以自步长方式将分割结果传播至相邻切片,逐步扩展至完整的3D体积。
  • 采用堆叠生成对抗网络(SGAN)将图像分辨率提升2–5倍,以提升对小病灶的检测能力。
  • 利用优化后的分割结果进一步微调CNN,实现在病灶体积范围内的自适应学习。
  • 该方法在包含完整3D真实标签的DeepLesion数据集和淋巴结数据集上进行了验证。

实验结果

研究问题

  • RQ1RECIST直径能否被有效用作弱监督信号,以在CT扫描中生成精确的3D病灶分割?
  • RQ2与标准弱监督方法相比,自步长学习在病灶分割性能上提升程度如何?
  • RQ3超分辨率图像生成在多大程度上提升了病灶边界检测与分割的准确性?
  • RQ4与基于RECIST的椭球体积估计算法相比,所提出的WSSS方法在追踪病灶体积随时间变化时的性能如何?
  • RQ5专家放射科医生对WSSS生成的分割结果的临床可接受度如何?

主要发现

  • 该方法在RECIST切片上实现了93%的平均Dice系数,表明在初始病灶定位方面具有高精度。
  • 3D病灶体积分割的平均Dice分数达到76%,表明在完整体积分割重建中表现优异。
  • 放射科医生对WSSS生成分割结果的主观接受率为92.4%,接近手动分割98.7%的接受率。
  • WSSS方法与手动体积测量的相关系数(R² = 0.78)高于基于RECIST的椭球估计算法(R² = 0.62)。
  • 通过堆叠GAN实现的超分辨率处理使平均表面距离(AVD)降低15%,从0.230降至0.189,表明边界勾画能力显著提升。
  • 在直接对比中,放射科医生在72例中更偏好WSSS分割,在645例中认为两者无差异,在269例中更偏好手动分割,证实该方法具有高度临床相关性。

更好的研究,从现在开始

从阅读论文到最终审阅,大幅缩短您的研究时间。

无需绑定信用卡

本解读由 AI 生成,并经人工编辑审核。