文献速递腹部技术与安全CT

基于CT图像的全自动深度学习分层肌肉分割用于肌少症评估

Fully automated deep-learning section-based muscle segmentation from CT images for sarcopenia assessment.
PubMed 文献 · Clin Radiol · 2022-03-05

本研究开发了基于全卷积神经网络的全自动深度学习流程,在无病例排除标准的腹部CT上自动检测L3椎体层面并分割竖脊肌、腰大肌和腹直肌,用于肌少症的机会性筛查。在三折交叉验证中,各肌肉分割Dice重叠为0.94–0.97,联合肌肉面积Dice为0.96,留出测试集28/31例中肌肉衰减/面积的R²为0.95/0.98,全自动处理每例CT不足1秒。

要点

  • 全自动流程无需手动步骤,可在无排除标准的腹部CT上检测L3层面并分割三组肌肉。
  • 竖脊肌、腰大肌、腹直肌分割Dice分别为0.97、0.95、0.94,联合肌肉面积Dice为0.96。
  • 留出测试集28/31例中肌肉衰减和面积的R²分别为0.95和0.98,自动输出与第二标注者无统计学差异。

核心结论

  • 全卷积神经网络流程可准确高效地从无筛选的腹部CT体积中自动分割L3椎体层面的肌肉,无需任何手动处理步骤;该方法有望作为机会性衰弱筛查的通用工具。

研究设计与样本

  • 伦理批准的回顾性研究,使用公开可用及机构内无筛选的腹部CT图像(训练集n=1,070,测试集n=31),采用基于UNet类架构的全卷积神经网络,分两步:从CT体积检测层面,再在单层面上分割肌肉,输出L3椎体层面的肌肉面积,并与专家手动分割对照,采用三折交叉验证评估模型。

关键影像表现/征象

  • 输出为L3椎体层面的肌肉面积,肌肉被分割为竖脊肌、腰大肌和腹直肌肌群。
  • 输入数据包含多种视野、层厚、遮挡、伪影和解剖变异。

诊断标准与关键数值

  • 层面检测交叉验证误差为1.41 ± 5.02(以层面计)。
  • 分割交叉验证Dice重叠:竖脊肌0.97 ± 0.02,腰大肌0.95 ± 0.04,腹直肌0.94 ± 0.04,联合肌肉面积0.96 ± 0.02。
  • 在28/31例留出测试病例中,肌肉衰减和面积的R²分别为0.95和0.98。
  • 自动输出与第二标注者之间未发现统计学差异。
  • 全自动处理每例CT检查耗时不足1秒。

临床应用与阅片要点

  • 该流程可在无病例排除标准的常规腹部CT上自动完成L3层面肌肉分割,无需手动处理步骤。
  • 作者认为该方法有望作为机会性衰弱筛查的通用工具。

局限性

  • 原文未报告局限性。

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

AIM: To develop a fully automated deep-learning-based approach to measure muscle area for assessing sarcopenia on standard-of-care computed tomography (CT) of the abdomen without any case exclusion criteria, for opportunistic screening for frailty.

MATERIALS AND METHODS: This ethically approved retrospective study used publicly available and institutional unselected abdominal CT images (n=1,070 training, n=31 testing). The method consisted of two sequential steps: section detection from CT volume followed by muscle segmentation on single-section. Both stages used fully convolutional neural networks (FCNN), based on a UNet-like architecture. Input data consisted of CT volumes with a variety of fields of view, section thicknesses, occlusions, artefacts, and anatomical variations. Output consisted of segmented muscle area on a CT section at the L3 vertebral level. The muscle was segmented into erector spinae, psoas, and rectus abdominus muscle groups. Output was tested against expert manual segmentation.

RESULTS: Threefold cross-validation was used to evaluate the model. Section detection cross-validation error was 1.41 ± 5.02 (in sections). Segmentation cross-validation Dice overlaps were 0.97 ± 0.02, 0.95 ± 0.04, and 0.94 ± 0.04 for erector spinae, psoas, and rectus abdominus, respectively, and 0.96 ± 0.02 for the combined muscle area, with R2 = 0.95/0.98 for muscle attenuation/area in 28/31 hold-out test cases. No statistical difference was found between the automated output and a second annotator. Fully automated processing took <1 second per CT examination.

CONCLUSIONS: A FCNN pipeline accurately and efficiently automates muscle segmentation at the L3 vertebral level from unselected abdominal CT volumes, with no manual processing step. This approach is promising as a generalisable tool for opportunistic screening for frailty on standard-of-care CT.

来源:Clin Radiol,查看原文。版权归原作者所有。

引用

  1. Islam, Kanavati, Arain, et al. Fully automated deep-learning section-based muscle segmentation from CT images for sarcopenia assessment. Clin Radiol. 2022. PMID: 35260232. doi:10.1016/j.crad.2022.01.036

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。