文献速递神经CT

基于生成对抗网络的无监督深度学习用于随访CT卒中病灶分割

Unsupervised Deep Learning for Stroke Lesion Segmentation on Follow-up CT Based on Generative Adversarial Networks.
PubMed 文献 · AJNR Am J Neuroradiol · 2022-07-28

本研究利用3项荷兰急性缺血性卒中试验中820例患者的基线和随访NCCT数据,开发并评估生成对抗网络在随访NCCT上分割梗死和出血性卒中病灶。24小时和1周随访梗死灶的中位Dice相似系数分别为0.31和0.59,预测病灶体积的组内相关系数分别为0.83和0.90;出血转化和原发性出血病灶的Dice相似系数很低。结论认为无监督生成对抗网络可获得中等Dice相似系数和良好体积一致性的自动梗死病灶分割。

要点

  • 24小时和1周随访梗死灶的中位Dice相似系数分别为0.31和0.59。
  • 24小时和1周随访梗死灶预测体积的组内相关系数分别为0.83和0.90。
  • 出血转化和原发性出血病灶的中位Dice相似系数分别为0.02和0.08。

核心结论

  • 无监督生成对抗网络可用于获得自动梗死病灶分割,具有中等Dice相似系数和良好体积一致性。

研究设计与样本

  • 研究为开发并评估生成对抗网络的研究,训练数据来自3项荷兰急性缺血性卒中试验的820例患者,包含基线和随访NCCT。
  • 生成对抗网络经优化将带病灶的随访扫描转换为生成的无病灶基线扫描,通过生成差值图并从随访扫描中减去来实现。
  • 生成的差值图用于自动提取病灶分割。
  • 评估对象包括原发性出血病灶、缺血性卒中出血转化以及24小时和1周随访梗死病灶,参照专家标注,采用Dice相似系数、Bland-Altman分析和组内相关系数进行评价。

诊断标准与关键数值

  • 24小时随访梗死灶中位Dice相似系数为0.31(四分位距0.08-0.59),1周随访梗死灶为0.59(四分位距0.29-0.74)。
  • 出血转化中位Dice相似系数为0.02(四分位距0-0.14),原发性出血病灶为0.08(四分位距0.01-0.35)。
  • 24小时随访梗死灶预测体积偏倚3 mL,一致性界限-64至59 mL,组内相关系数0.83(95%CI 0.78-0.88)。
  • 1周随访梗死灶预测体积偏倚-4 m,一致性界限-66至58 mL,组内相关系数0.90(95%CI 0.83-0.93)。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

英文原文(摘要)· 点击展开

BACKGROUND AND PURPOSE: Supervised deep learning is the state-of-the-art method for stroke lesion segmentation on NCCT. Supervised methods require manual lesion annotations for model development, while unsupervised deep learning methods such as generative adversarial networks do not. The aim of this study was to develop and evaluate a generative adversarial network to segment infarct and hemorrhagic stroke lesions on follow-up NCCT scans.

MATERIALS AND METHODS: Training data consisted of 820 patients with baseline and follow-up NCCT from 3 Dutch acute ischemic stroke trials. A generative adversarial network was optimized to transform a follow-up scan with a lesion to a generated baseline scan without a lesion by generating a difference map that was subtracted from the follow-up scan. The generated difference map was used to automatically extract lesion segmentations. Segmentation of primary hemorrhagic lesions, hemorrhagic transformation of ischemic stroke, and 24-hour and 1-week follow-up infarct lesions were evaluated relative to expert annotations with the Dice similarity coefficient, Bland-Altman analysis, and intraclass correlation coefficient.

RESULTS: The median Dice similarity coefficient was 0.31 (interquartile range, 0.08-0.59) and 0.59 (interquartile range, 0.29-0.74) for the 24-hour and 1-week infarct lesions, respectively. A much lower Dice similarity coefficient was measured for hemorrhagic transformation (median, 0.02; interquartile range, 0-0.14) and primary hemorrhage lesions (median, 0.08; interquartile range, 0.01-0.35). Predicted lesion volume and the intraclass correlation coefficient were good for the 24-hour (bias, 3 mL; limits of agreement, -64-59 mL; intraclass correlation coefficient, 0.83; 95% CI, 0.78-0.88) and excellent for the 1-week (bias, -4 m; limits of agreement,-66-58 mL; intraclass correlation coefficient, 0.90; 95% CI, 0.83-0.93) follow-up infarct lesions.

CONCLUSIONS: An unsupervised generative adversarial network can be used to obtain automated infarct lesion segmentations with a moderate Dice similarity coefficient and good volumetric correspondence.

来源:AJNR Am J Neuroradiol,查看原文。版权归原作者所有。

引用

  1. van Voorst, Konduri, van Poppel, et al. Unsupervised Deep Learning for Stroke Lesion Segmentation on Follow-up CT Based on Generative Adversarial Networks. AJNR Am J Neuroradiol. 2022. PMID: 35902122. doi:10.3174/ajnr.A7582

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。