Multi-Crop Convolutional Neural Networks for Fast Lung Nodule Segmentation

被引:28
|
作者
Chen, Quan [1 ,2 ]
Xie, Wei [3 ]
Zhou, Pan [4 ]
Zheng, Chuansheng [1 ]
Wu, Dapeng [5 ]
机构
[1] Huazhong Univ Sci & Technol, Dept Radiol, Union Hosp, Tongji Med Coll, Wuhan 430022, Peoples R China
[2] Hubei Prov Key Lab Mol Imaging, Wuhan 430022, Peoples R China
[3] Huazhong Univ Sci & Technol, Sch Elect Informat & Commun, Wuhan 430074, Peoples R China
[4] Huazhong Univ Sci & Technol, Hubei Engn Res Ctr Big Data Secur, Sch Cyber Sci & Engn, Wuhan 430074, Peoples R China
[5] Univ Florida, Dept Elect & Comp Engn, Gainesville, FL 32611 USA
来源
IEEE TRANSACTIONS ON EMERGING TOPICS IN COMPUTATIONAL INTELLIGENCE | 2022年 / 6卷 / 05期
基金
中国国家自然科学基金;
关键词
Lung; Three-dimensional displays; Feature extraction; Image segmentation; Computed tomography; Two dimensional displays; Task analysis; Convolutional neural network; loss function; lung nodule segmentation; pooling layer; MR BRAIN IMAGES; PULMONARY NODULES; AUTOMATIC SEGMENTATION; CT SCANS;
D O I
10.1109/TETCI.2021.3051910
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Computed tomography (CT) images are formally taken as an assistance of early diagnosis in lung nodule analysis. Thus the accurate lung nodule segmentation is in great need for image-driven tasks. However, as heterogeneity exists between different types of lung nodules, the similar visual appearance between the pixels of nodules and pixels of non-nodule area make it difficult for automatic lung nodule segmentation. In this article, we propose a fast end-to-end framework, called Fast Multi-crop Guided Attention (FMGA) network, to accurately segment lung nodules in CT images. Our method utilizes multi-crop nodule slices as input to aggregate contextual information (2D context from current image slice and 3D context from adjacent axial slices), and exploits a global convolutional layer for nodule pixel embedding matching. To further make use of the information from border pixels near the nodule margin for better segmentation, we develop a weighted loss function to facilitate the model training by considering a balanced class samples of pixels around the nodule margin. Moreover, we utilize a central pooling layer to facilitate the contexts feature propagation in pixel neighbors. We evaluate our method on the largest public lung CT dataset LIDC and the collected lung CT data from Wuhan local hospital, respectively. Experimental results show that FMGA achieves superior performance among the state-of-the-arts. In addition, we give an ablation study and visualization results to illustrate how each component works for accurate lung nodule segmentation.
引用
收藏
页码:1190 / 1200
页数:11
相关论文
共 50 条
  • [21] Convolutional Neural Networks for Multi-scale Lung Nodule Classification in CT: Influence of Hyperparameter Tuning on Performance
    Hernandez-Rodriguez, Jorge
    Cabrero-Fraile, Francisco-Javier
    Rodriguez-Conde, Maria-Jose
    TEM JOURNAL-TECHNOLOGY EDUCATION MANAGEMENT INFORMATICS, 2022, 11 (01): : 297 - 306
  • [22] Automatic Lung Nodule Detection in CT Images Using Convolutional Neural Networks
    Shaukat, Furcian
    Javed, Kamran
    Raja, Gulistan
    Mir, Junaid
    Shahid, Muhammad Laiq Ur Rahman
    IEICE TRANSACTIONS ON FUNDAMENTALS OF ELECTRONICS COMMUNICATIONS AND COMPUTER SCIENCES, 2019, E102A (10) : 1364 - 1373
  • [23] LUNG NODULE CLASSIFICATION BY THE COMBINATION OF FUSION CLASSIFIER AND CASCADED CONVOLUTIONAL NEURAL NETWORKS
    Sakamoto, Masaharu
    Nakano, Hiroki
    Zhao, Kun
    Sekiyama, Taro
    2018 IEEE 15TH INTERNATIONAL SYMPOSIUM ON BIOMEDICAL IMAGING (ISBI 2018), 2018, : 822 - 825
  • [24] Lung Nodule Detection Based on 3D Convolutional Neural Networks
    Fan, Lei
    Xia, Zhaoqiang
    Zhang, Xiaobiao
    Feng, Xiaoyi
    2017 INTERNATIONAL CONFERENCE ON THE FRONTIERS AND ADVANCES IN DATA SCIENCE (FADS), 2017, : 7 - 10
  • [25] Lung Nodule Detection in CT Images using Deep Convolutional Neural Networks
    Golan, Rotem
    Jacob, Christian
    Denzinger, Jorg
    2016 INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS (IJCNN), 2016, : 243 - 250
  • [26] Fast and robust segmentation of the striatum using deep convolutional neural networks
    Choi, Hongyoon
    Jin, Kyong Hwan
    JOURNAL OF NEUROSCIENCE METHODS, 2016, 274 : 146 - 153
  • [27] Unsupervised Image Segmentation using Convolutional Neural Networks for Automated Crop Monitoring
    Bhatt, Prakruti
    Sarangi, Sanat
    Pappula, Srinivasu
    ICPRAM: PROCEEDINGS OF THE 8TH INTERNATIONAL CONFERENCE ON PATTERN RECOGNITION APPLICATIONS AND METHODS, 2019, : 887 - 893
  • [28] A Fast Automatic Juxta-pleural Lung Nodule Detection Framework Using Convolutional Neural Networks and Vote Algorithm
    Tan, Jiaxing
    Huo, Yumei
    Liang, Zhengrong
    Li, Lihong
    PATCH-BASED TECHNIQUES IN MEDICAL IMAGING, PATCH-MI 2018, 2018, 11075 : 85 - 92
  • [29] Lung-Nodule Segmentation Using a Convolutional Neural Network with the U-Net Architecture
    Hernandez-Solis, Vicente
    Tellez-Velazquez, Arturo
    Orantes-Molina, Antonio
    Cruz-Barbosa, Raul
    PATTERN RECOGNITION (MCPR 2021), 2021, 12725 : 335 - 344
  • [30] Optimized convolutional neural network for automatic lung nodule detection with a new active contour segmentation
    Kumar, M. Kiran
    Amalanathan, Anthoniraj
    SOFT COMPUTING, 2023, 27 (20) : 15365 - 15381