Prediction of ipsilateral axillary lymph node metastases based on high-throughput sonographic features of invasive breast carcinoma

HU Yuzhou; LI Jiawei; GUO Yi; WANG Yuanyuan; YU Jinhua; CHANG Cai

doi:10.19732/j.cnki.2096-6210.2019.02.001

您当前的位置：

首页 >

文章列表页 >

Prediction of ipsilateral axillary lymph node metastases based on high-throughput sonographic features of invasive breast carcinoma

更新时间：2025-12-15

- Prediction of ipsilateral axillary lymph node metastases based on high-throughput sonographic features of invasive breast carcinoma
- Vol. 28, Issue 2, Pages: 65-71(2019)
- 作者机构：
  
  1. 复旦大学电子工程系,上海,200433
  2. 复旦大学附属肿瘤医院超声科，复旦大学上海医学院肿瘤学系,上海,200032
- 作者简介：
- 基金信息：
- DOI：10.19732/j.cnki.2096-6210.2019.02.001
  CLC：
- Published Online：24 May 2019，
  
  Published：24 May 2019
- 稿件说明：
移动端阅览
胡雨舟,李佳伟,郭翌,等. 浸润性乳腺癌超声高通量图像特征预测同侧腋窝淋巴结转移[J]. 肿瘤影像学, 2019, 28(2): 65-71 https://doi.

org/10.19732/j.cnki.2096-6210.2019.02.001
胡雨舟,李佳伟,郭翌,等. 浸润性乳腺癌超声高通量图像特征预测同侧腋窝淋巴结转移[J]. 肿瘤影像学, 2019, 28(2): 65-71 https://doi. DOI： 10.19732/j.cnki.2096-6210.2019.02.001.

org/10.19732/j.cnki.2096-6210.2019.02.001 DOI：

摘要

目的：

探讨利用浸润性乳腺癌超声高通量图像特征预测同侧腋窝淋巴结转移的价值。

方法：

为表征乳腺肿瘤的特性，设计基于乳腺影像报告与数据系统（Breast Imaging Reporting and Data System，BI-RADS）的高通量特征、尺度不变特征转换（scale-invariant feature transform，SIFT）的特征和卷积神经网络（convolutional neural network，CNN）的特征，并结合3个重要临床信息，共获得10 703个特征，构成相应的特征体系。利用十折法随机重复采样100次，先用单因素t检验筛选出

＜0.05的特征，再使用最小化绝对值收缩和选择算子（least absolute shrinkage and selection operator，LASSO）回归模型进一步筛选，将出现次数最多的100个特征作为最优特征集，输入一个使用十倍交叉验证法的线性核支持向量机（support vector machine，SVM）分类器中进行分类。

结果：

对380例女性浸润性乳腺癌患者进行研究，分类器受试者工作特征（receiver operating characteristic，ROC）曲线的曲线下面积（area under curve，AUC）为0.903，准确率为82.6%，灵敏度为90.6%，特异度为69.9%。

结论：

该研究提取的乳腺超声图像特征能较好地预测淋巴结转移，可为医师进行乳腺淋巴结检查提供参考。

Abstract

Objective:

To investigate the ipsilateral axillary lymph node metastases based on high-throughput sonographic features of invasive breast carcinoma.

Methods:

In order to represent the characteristics of breast tumors

the high-throughput features based on Breast Imaging Reporting and Data System (BI-RA

DS)

features based on scale-invariant feature transform (SIFT) and features based on convolutional neural network (CNN) were extracted. Combined with three important clinical features

a total of 10 703 features were obtained to constitute a feature system. A 10-fold feature selection method using t-test and least absolute shrinkage and selection operator (LASSO) regression model was bootstrapped for 100 times. The 100 most frequent features were input into a linear-kernel support vector machine (SVM) classifier in 10-fold cross validation experiment.

Results:

For the included 380 breast cancer cases

the area under curve (AUC) of the receiver operating characteristic (ROC) curve

accuracy

sensitivity

specificity were 0.903

82.6%

90.6%

69.9%

respectively.

Conclusion:

The sonographic features extracted from breast ultrasound images can predict lymph node metastases and provide the possibility of lymph node metastases for a sonographer.

关键词

Keywords

references

Views

1617

下载量

CNKI被引量

Alert me when the article has been cited

提交

Tools

Publicity Resources

Computer-aided design of a novel HER2-targeted cyclic peptide PET probe and its evaluation for noninvasive imaging of breast cancer

Prediction of treating response for breast cancer by multi-phase MRI histogram arrays

Risk assessment of sentinel lymph node metastasis in breast cancer using a radiomics model based on B-mode ultrasound and color doppler ultrasound

Value of deep learning ultrasound radiomics in predicting axillary lymph node metastasis of breast cancer

Multimodal ultrasound features combined with serum CEA and CK19 to predict axillary lymph node metastasis of breast cancer

Related Author

GUO Xu

GAO Xiaomin

YAO Ying

ZHU Haitao

LI Xiaoting

QU Yuhong

SUN Yingshi

Yicheng ZHU

Related Institution

Department of Pharmacy, Wuxi Maternal and Child Health Hospital Affiliated to Nanjing Medical University

Wuxi Medical College, Jiangnan University

Department of Pharmacy, Wuxi Maternal and Child Health Care Hospital

Department of Radiology, Peking University Cancer Hospital

Department of Radiology, Affiliated Beijing Chaoyang Hospital of Capital Medical University

AI问答

⁰