Dual attention enhancement feature fusion network for segmentation and quantitative analysis of paediatric echocardiography

被引:55
作者
Guo, Libao [1 ]
Lei, Baiying [1 ]
Chen, Weiling [2 ]
Du, Jie [1 ]
Frangi, Alejandro F. [3 ,4 ,5 ,6 ]
Qin, Jing [7 ]
Zhao, Cheng [1 ]
Shi, Pengpeng [2 ]
Xia, Bei [2 ]
Wang, Tianfu [1 ]
机构
[1] Shenzhen Univ, Sch Biomed Engn,Marshall Lab Biomed Engn, Hlth Sci Ctr,Guangdong Key Lab Biomed Measurement, AI Res Ctr Med Image Anal & Diag,Natl Reg Key Tec, Shenzhen 518060, Peoples R China
[2] Hosp Shantou Univ, Shenzhen Children Hosp, Dept Ultrasound Dept, Shenzhen 518050, Peoples R China
[3] Univ Leeds, Sch Comp, Ctr Computat Imaging & Simulat Technol Biomed CIS, Leeds LS2 9JT, W Yorkshire, England
[4] Univ Leeds, Sch Med, Leeds Inst Data Analyt, Leeds LS2 9JT, W Yorkshire, England
[5] Katholieke Univ Leuven, Dept Cardiovasc Sci, B-3000 Leuven, Belgium
[6] Katholieke Univ Leuven, Dept Elect Engn, B-3000 Leuven, Belgium
[7] Hong Kong Polytech Univ, Ctr Smart Hlth, Sch Nursing, Hong Kong, Peoples R China
基金
中国国家自然科学基金;
关键词
Paediatric echocardiography segmentation; quantitative analysis; Attention mechanism; Feature fusion; Dual attention enhancement; CONGENITAL HEART-DISEASE; AMERICAN SOCIETY; QUANTIFICATION; RECOMMENDATIONS; ULTRASOUND; UNET;
D O I
10.1016/j.media.2021.102042
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Paediatric echocardiography is a standard method for screening congenital heart disease (CHD). The segmentation of paediatric echocardiography is essential for subsequent extraction of clinical parameters and interventional planning. However, it remains a challenging task due to (1) the considerable variation of key anatomic structures, (2) the poor lateral resolution affecting accurate boundary definition, (3) the existence of speckle noise and artefacts in echocardiographic images. In this paper, we propose a novel deep network to address these challenges comprehensively. We first present a dual-path feature extraction module (DP-FEM) to extract rich features via a channel attention mechanism. A high-and low-level feature fusion module (HL-FFM) is devised based on spatial attention, which selectively fuses rich semantic information from high-level features with spatial cues from low-level features. In addition, a hybrid loss is designed to deal with pixel-level misalignment and boundary ambiguities. Based on the segmentation results, we derive key clinical parameters for diagnosis and treatment planning. We extensively evaluate the proposed method on 4,485 two-dimensional (2D) paediatric echocardiograms from 127 echocardiographic videos. The proposed method consistently achieves better segmentation performance than other state-of-the-art methods, whichdemonstratesfeasibility for automatic segmentation and quantitative analysis of paediatric echocardiography. Our code is publicly available at https://github.com/end- of- the-century/Cardiac . (c) 2021 Elsevier B.V. All rights reserved.
引用
收藏
页数:15
相关论文
共 58 条
[1]   Mitral Annulus Segmentation Using Deep Learning in 3-D Transesophageal Echocardiography [J].
Andreassen, Borge Solli ;
Veronesi, Federico ;
Gerard, Olivier ;
Solberg, Anne H. Schistad ;
Samset, Eigil .
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS, 2020, 24 (04) :994-1003
[2]  
[Anonymous], 2011, WHO I Global atlas on cardiovascular disease prevention and control
[3]   Generalizable fully automated multi-label segmentation of four-chamber view echocardiograms based on deep convolutional adversarial networks [J].
Arafati, Arghavan ;
Morisawa, Daisuke ;
Avendi, Michael R. ;
Amini, M. Reza ;
Assadi, Ramin A. ;
Jafarkhani, Hamid ;
Kheradvar, Arash .
JOURNAL OF THE ROYAL SOCIETY INTERFACE, 2020, 17 (169)
[4]  
Ba J. L., 2016, J. Mach. Learn. Res.
[5]   SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image Segmentation [J].
Badrinarayanan, Vijay ;
Kendall, Alex ;
Cipolla, Roberto .
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2017, 39 (12) :2481-2495
[6]  
Cao Y, 2019, IEEE ICC
[7]   An Attentive Survey of Attention Models [J].
Chaudhari, Sneha ;
Mithal, Varun ;
Polatkan, Gungor ;
Ramanath, Rohan .
ACM TRANSACTIONS ON INTELLIGENT SYSTEMS AND TECHNOLOGY, 2021, 12 (05)
[8]  
Chen L.C., 2014, Semantic image segmentation with deep convolutional nets and fully connected CRFs, DOI DOI 10.48550/ARXIV.1412.7062
[9]  
Chen L.-C., 2017, arXiv preprint arXiv:2310.0182
[10]  
Cheng H.K., 2020, P IEEE CVF C COMP VI, P8890