文献速递腹部CT

基于机器学习的CT特征鉴别胃神经鞘瘤与胃肠间质瘤

Differentiation of gastric schwannomas from gastrointestinal stromal tumors by CT using machine learning.
PubMed 文献 · Abdom Radiol (NY) · 2020-10-20

本研究纳入49例胃神经鞘瘤和139例胃肠间质瘤患者,利用单因素分析中P<0.1的11项CT特征构建逻辑回归、决策树、随机森林和梯度提升决策树模型,并与多因素分析和放射科医师主观诊断比较。逻辑回归(M2)模型在测试集AUC为0.967,鉴别效能最优;胃体位置、外生及混合生长模式、无坏死及表面溃疡、淋巴结增大及较大强化潜能是提示神经鞘瘤的最重要CT特征。

要点

  • 胃体位置、外生及混合生长模式、无坏死及表面溃疡、淋巴结增大、较大强化潜能提示胃神经鞘瘤。
  • 逻辑回归(M2)模型测试集AUC为0.967,优于其他机器学习版本、多因素分析及放射科医师主观诊断。
  • CT特征包括平扫CT值、门静脉期CT值、门静脉期强化程度、门静脉期减动脉期CT值差、强化潜能、位置、轮廓、生长模式、坏死、表面溃疡及淋巴结增大。

核心结论

  • 逻辑回归(M2)模型在鉴别胃神经鞘瘤与胃肠间质瘤方面诊断效能最优,优于其他机器学习版本、多因素分析及放射科医师的主观诊断表现。

研究设计与样本

  • 本研究纳入经病理证实的49例胃神经鞘瘤患者和139例胃肠间质瘤患者,比较四种机器学习模型(逻辑回归、决策树、随机森林、梯度提升决策树)、多因素分析及放射科医师主观诊断的鉴别效能。
  • 将单因素分析中P<0.1的11项CT特征输入四种模型,其中LR(M1)、随机森林、决策树和梯度提升决策树模型包含全部11个变量,LR(M2)模型使用从M1中筛选出的6个最具预测价值的变量构建。

关键影像表现/征象

  • 单因素分析中P<0.1的CT特征包括:平扫期CT值(CTU)、门静脉期强化(CTV)、门静脉期强化程度(DEPP)、门静脉期减动脉期CT值差(CTV-CTA)、强化潜能(EP)、位置、轮廓、生长模式、坏死、表面溃疡及淋巴结增大。
  • 提示胃神经鞘瘤的最重要CT特征为:位于胃体、外生及混合生长模式、无坏死及表面溃疡、淋巴结增大以及较大的强化潜能。

诊断标准与关键数值

  • LR(M2)模型在测试数据集中的AUC为0.967,被认为是鉴别两种肿瘤的最优模型。

临床应用与阅片要点

  • 日常阅片鉴别胃神经鞘瘤与胃肠间质瘤时,可关注胃体位置、外生及混合生长模式、无坏死及表面溃疡、淋巴结增大及较大强化潜能等CT特征。
  • LR(M2)模型基于6个最具预测价值的变量构建,其诊断效能优于放射科医师主观诊断。

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

OBJECTIVE: To identify schwannomas from gastrointestinal stromal tumors (GISTs) by CT features using Logistic Regression (LR), Decision Trees (DT), Random Forest (RF), and Gradient Boosting Decision Tree (GBDT).

METHODS: This study enrolled 49 patients with schwannomas and 139 with GISTs proven by pathology. CT features with P < 0.1 derived from univariate analysis were inputted to four models. Five machine learning (ML) versions, multivariate analysis, and radiologists' subjective diagnostic performance were compared to evaluate diagnosis performance of all the traditional and advanced methods.

RESULTS: The CT features with P < 0.1 were as follows: (1) CT attenuation value of unenhancement phase (CTU), (2) portal venous enhancement (CTV), (3) degree of enhancement in the portal venous phase (DEPP), (4) CT attenuation value of portal venous phase minus arterial phase (CTV-CTA), (5) enhanced potentiality (EP), (6) location, (7) contour, (8) growth pattern, (9) necrosis, (10) surface ulceration, (11) enlarged lymph node (LN). LR (M1), RF, DT, and GBDT models contained all of the above 11 variables, while LR (M2) was developed using six most predictive variables derived from (M1). LR (M2) model with AUC of 0.967 in test dataset was thought to be optimal model in differentiating the two tumors. Location in gastric body, exophytic and mixed growth pattern, lack of necrosis and surface ulceration, enlarged lymph nodes, and larger EP were the most important CT features suggestive of schwannomas.

CONCLUSION: LR (M2) provided the optimal diagnostic potency among other ML versions, multivariate analysis, and radiologists' performance on differentiation of schwannomas from GISTs.

来源:Abdom Radiol (NY),查看原文。版权归原作者所有。

引用

  1. Wang J, Xie Z, Zhu X, et al. Differentiation of gastric schwannomas from gastrointestinal stromal tumors by CT using machine learning. Abdom Radiol (NY). 2020. PMID: 33083871. doi:10.1007/s00261-020-02797-9

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。