计算机应用 ›› 2020, Vol. 40 ›› Issue (1): 77-82.DOI: 10.11772/j.issn.1001-9081.2019061113

• 人工智能 • 上一篇    下一篇

基于不同超声成像的甲状腺结节良恶性判别

武宽1,2, 秦品乐1,2, 柴锐1,2, 曾建朝1,2   

  1. 1. 山西省医学影像与数据分析工程研究中心(中北大学), 太原 030051;
    2. 中北大学 大数据学院, 太原 030051
  • 收稿日期:2019-06-27 修回日期:2019-09-09 出版日期:2020-01-10 发布日期:2019-10-15
  • 通讯作者: 曾建朝
  • 作者简介:武宽(1995-),男,山西运城人,硕士研究生,主要研究方向:计算机视觉、医学影像、机器学习;秦品乐(1978-),男,山西长治人,副教授,博士,主要研究方向:医学影像、机器视觉、三维重建;柴锐(1985-),男,山西运城人,讲师,博士,主要研究方向:医学影像处理;曾建潮(1963-),男,山西太原人,教授,博士,博士生导师,主要研究方向:医学影像、复杂系统的维护决策。
  • 基金资助:
    山西省自然科学基金资助项目(2015011045)。

Benign and malignant diagnosis of thyroid nodules based on different ultrasound imaging

WU Kuan1,2, QIN Pinle1,2, CHAI Rui1,2, ZENG Jianchao1,2   

  1. 1. Shanxi Medical Imaging and Data Analysis Engineering Research Center(North University of China), Taiyuan Shanxi 030051, China;
    2. College of Big Data, North University of China, Taiyuan Shanxi 030051, China
  • Received:2019-06-27 Revised:2019-09-09 Online:2020-01-10 Published:2019-10-15
  • Supported by:
    This work is partially supported by the Natural Science Foundation of Shanxi Province (2015011045).

摘要: 为实现更为准确的甲状腺结节良恶性超声图像诊断,避免不必要的穿刺或活检手术,提出了一种基于卷积神经网络(CNN)的常规超声成像和超声弹性成像的特征结合方法,提高了甲状腺结节良恶性分类准确率。首先,卷积网络模型在大规模自然图像数据集上完成预训练,并通过迁移学习的方式将特征参数迁移到超声图像域用以生成深度特征并处理小样本。然后,结合常规超声成像和超声弹性成像的深度特征图形成混合特征空间。最后,在混合特征空间上完成分类任务,实现了一个端到端的卷积网络模型。在1156幅图像上进行实验,所提方法的准确率为0.924,高于其他单一数据源的方法。实验结果表明,浅层卷积共享图像的边缘纹理特征,高层卷积的抽象特征与具体的分类任务相关,使用迁移学习的方法可以解决数据样本不足的问题;同时,弹性超声影像可以对甲状腺结节的病灶硬度进行客观的量化,结合常规超声的纹理轮廓特征,二者融合的混合特征可以更全面地描述不同病灶之间的差异。所提方法可以高效准确地对甲状腺结节进行良恶性分类,减轻患者痛苦,给医生提供更为准确的辅助诊断信息。

关键词: 图像分类, 迁移学习, 特征融合, 深度学习, 超声影像, 弹性超声

Abstract: In order to achieve more accurate diagnosis of benign and malignant of thyroid nodule ultrasound images and avoid unnecessary puncture or biopsy surgery, a feature combining method of conventional ultrasound imaging and ultrasound elastography based on Convolutional Neural Network (CNN) was proposed to improve the accuracy of benign and malignant classification of thyroid nodules. Firstly, large-scale natural image datasets were used by the convolutional network model for pre-training, and the feature parameters were transferred to the ultrasound image domain by transfer learning to generate depth features and process small samples. Then, the depth feature maps of conventional ultrasound imaging and ultrasound elastography were combined to form a hybrid feature space. Finally, the classification task was completed in the hybrid feature space, and an end-to-end convolution network model was constructed. The experiments were carried out on 1156 images, the proposed method had the accuracy of 0.924, which was higher than that of other single data source methods. The experimental results show that, the edge and texture features of the image are shared by the shallow convolutions, the abstract features of the high-level convolutions are related to the specific classification tasks, and the transfer learning method can solve the problem of insufficient data samples. At the same time, the elastic ultrasound image can objectively quantify the lesion hardness of thyroid nodules, and with the combination of the texture contour features of conventional ultrasound image, the mixed features can more fully describe the differences between different lesions. Therefore, this method can effectively and accurately classify the thyroid nodules, reduce the pain of patients, and provide doctors with more accurate auxiliary diagnostic information.

Key words: image classification, transfer learning, feature fusion, deep learning, ultrasound imaging, elastic ultrasound

中图分类号: