文献速递腹部CT

基于深度学习构建局部原发性胃肠道间质瘤核分裂计数恶性潜能二分类预测模型

Development of a Malignancy Potential Binary Prediction Model Based on Deep Learning for the Mitotic Count of Local Primary Gastrointestinal Stromal Tumors.
PubMed 文献 · Korean J Radiol · 2020-10-21

回顾性收集148例病理确诊胃肠道间质瘤的腹部增强CT图像,由经验丰富的放射科医师勾画肿瘤区域,以术后病理核分裂计数为金标准(高计数>5/50高倍视野,低计数≤5/50高倍视野),基于VGG16卷积神经网络训练二分类模型(训练集108例、验证集20例、测试集20例)。测试集图像水平敏感度85.7%、特异度67.5%、AUC 0.771,患者水平敏感度90.0%、特异度70.0%、AUC 0.800,模型显示出良好的预测性能。

要点

  • 模型基于腹部增强CT图像预测GIST核分裂计数高低,图像水平AUC 0.771、患者水平AUC 0.800。
  • 高核分裂计数定义为>5/50高倍视野,低核分裂计数定义为≤5/50高倍视野,以术后病理为金标准。
  • 模型可依据Armed Forces Institute of Pathology标准预测肿瘤风险分级。

诊断标准与关键数值

  • 高核分裂计数定义为>5/50高倍视野,低核分裂计数定义为≤5/50高倍视野,以术后病理为金标准。
  • 测试集图像水平:敏感度85.7%(95%CI 0.834-0.877)、特异度67.5%(95%CI 0.636-0.712)、阳性预测值82.1%(95%CI 0.797-0.843)、阴性预测值73.0%(95%CI 0.691-0.766)、AUC 0.771(95%CI 0.750-0.791)。
  • 测试集患者水平:敏感度90.0%(95%CI 0.541-0.995)、特异度70.0%(95%CI 0.354-0.919)、阳性预测值75.0%(95%CI 0.428-0.933)、阴性预测值87.5%(95%CI 0.467-0.993)、AUC 0.800(95%CI 0.563-0.943)。
  • 根据Armed Forces Institute of Pathology标准预测肿瘤风险分级。

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

OBJECTIVE: The mitotic count of gastrointestinal stromal tumors (GIST) is closely associated with the risk of planting and metastasis. The purpose of this study was to develop a predictive model for the mitotic index of local primary GIST, based on deep learning algorithm.

MATERIALS AND METHODS: Abdominal contrast-enhanced CT images of 148 pathologically confirmed GIST cases were retrospectively collected for the development of a deep learning classification algorithm. The areas of GIST masses on the CT images were retrospectively labelled by an experienced radiologist. The postoperative pathological mitotic count was considered as the gold standard (high mitotic count, > 5/50 high-power fields [HPFs]; low mitotic count, ≤ 5/50 HPFs). A binary classification model was trained on the basis of the VGG16 convolutional neural network, using the CT images with the training set (n = 108), validation set (n = 20), and the test set (n = 20). The sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated at both, the image level and the patient level. The receiver operating characteristic curves were generated on the basis of the model prediction results and the area under curves (AUCs) were calculated. The risk categories of the tumors were predicted according to the Armed Forces Institute of Pathology criteria.

RESULTS: At the image level, the classification prediction results of the mitotic counts in the test cohort were as follows: sensitivity 85.7% (95% confidence interval [CI]: 0.834-0.877), specificity 67.5% (95% CI: 0.636-0.712), PPV 82.1% (95% CI: 0.797-0.843), NPV 73.0% (95% CI: 0.691-0.766), and AUC 0.771 (95% CI: 0.750-0.791). At the patient level, the classification prediction results in the test cohort were as follows: sensitivity 90.0% (95% CI: 0.541-0.995), specificity 70.0% (95% CI: 0.354-0.919), PPV 75.0% (95% CI: 0.428-0.933), NPV 87.5% (95% CI: 0.467-0.993), and AUC 0.800 (95% CI: 0.563-0.943).

CONCLUSION: We developed and preliminarily verified the GIST mitotic count binary prediction model, based on the VGG convolutional neural network. The model displayed a good predictive performance.

来源:Korean J Radiol,查看原文。版权归原作者所有。

引用

  1. Yang J, Chen Z, Liu W, et al. Development of a Malignancy Potential Binary Prediction Model Based on Deep Learning for the Mitotic Count of Local Primary Gastrointestinal Stromal Tumors. Korean J Radiol. 2020. PMID: 33169545. doi:10.3348/kjr.2019.0851

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。