Deep Dual-Stream Convolutional Neural Networks for Cardiac Image Semantic Segmentation
- 15 February 2024
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Industrial Informatics
- Vol. 20 (5) , 7440-7448
Abstract
Cardiac image segmentation is essential when applying biomedical informatics to improve industrial healthcare applications. To extract context and detailed information more efficiently and further improve cardiac image segmentation accuracy, we present a novel deep dual-stream convolutional neural network (CNN) for cardiac image semantic segmentation in this article. We use a body stream and a shape stream, respectively, in this method. First, in the body stream we propose integrating a gated fully fusion module to fuse multilevel features in the encoder and decoder paths. In addition, we integrate a feature aggregation module to extract the multiscale context. Second, in the shape stream, we propose using a gated shape CNN exploiting multilevel context to extract detailed information, such as boundary and shape features. Finally, we apply a multitask loss function to align the predicted masks with the ground truth labels. Our experiments on the public cardiac magnetic resonance image dataset show significant performance in the left and right ventricular cavities and myocardium compared to the state-of-the-art algorithms.Keywords
Funding Information
- Key Special Project for Technological Innovation and Application Development in Chongqing (CSTB2023TIAD-KPX0050)
- Fundamental Research Funds for the Central Universities (2023CDJXY-038, 2022CDJYGRH-015)
- National Natural Science Foundation of China (62176027)
- Key Projects of Basic Strengthening Plan (2022-JCJQ-ZD-018-11)
- Chongqing Talent (CQYC20220511986)
- Joint Equipment Pre Research and Key Fund Project of the Ministry of Education (8091B012207)
- Human Resources and Social Security Bureau Project of Chongqing (cx2020073)
- Guangdong Oppo Mobile Telecommunications Corporation Ltd (H20221694)
- Hong Kong GRF-RGC General Research Fund (11209819, CityU 9042816)
- Hong Kong GRF-RGC General Research Fund (11203820, CityU 9042598)
- Hong Kong Innovation and Technology Commission
- InnoHK Project Centre for Intelligent Multidimensional Data Analysis
This publication has 32 references indexed in Scilit:
- Segmentation of Cardiac Structures via Successive Subspace Learning with Saab Transform from Cine MRIAnnual International Conference of the IEEE Engineering in Medicine and Biology Society, 2021
- HigherHRNet: Scale-Aware Representation Learning for Bottom-Up Human Pose Estimation2020
- Learning Active Contour Models for Medical Image Segmentation2019
- PSANet: Point-wise Spatial Attention Network for Scene Parsing2018
- Squeeze-and-Excitation Networks2018
- 2D-3D Fully Convolutional Neural Networks for Cardiac MR Segmentation2018
- Automatic Segmentation of LV and RV in Cardiac MRI2018
- Automatic Segmentation and Disease Classification Using Cardiac Cine MR Images2018
- Pyramid Scene Parsing Network2017
- U-Net: Convolutional Networks for Biomedical Image Segmentation2015