Segmentation of biventricle in cardiac cine MRI via nested capsule dense network

Jilong Zhang; Yajuan Zhang; Hongyang Zhang; Quan Zhang; Weihua Su; Shijie Guo; Yuanquan Wang

doi:10.7717/peerj-cs.1146

Segmentation of biventricle in cardiac cine MRI via nested capsule dense network

PeerJ Comput Sci. 2022 Nov 30:8:e1146. doi: 10.7717/peerj-cs.1146. eCollection 2022.

Authors

Jilong Zhang^#¹, Yajuan Zhang^#¹, Hongyang Zhang¹, Quan Zhang^{2

3}, Weihua Su¹, Shijie Guo⁴, Yuanquan Wang^{1

4}

Affiliations

¹ School of Artificial Intelligence, Hebei University of Technology, Tianjin, China.
² School of Information and Communication Engineering, North University of China, Taiyuan, China.
³ Shanxi Provincial Key Laboratory for Biomedical Imaging and Big Data, North University of China, Taiyuan, China.
⁴ Hebei Key Laboratory of Robot Perception and Human-Robot Interaction, HeBUT, Tianjin, China.

^# Contributed equally.

Abstract

Background: Cardiac magnetic resonance image (MRI) has been widely used in diagnosis of cardiovascular diseases because of its noninvasive nature and high image quality. The evaluation standard of physiological indexes in cardiac diagnosis is essentially the accuracy of segmentation of left ventricle (LV) and right ventricle (RV) in cardiac MRI. The traditional symmetric single codec network structure such as U-Net tends to expand the number of channels to make up for lost information that results in the network looking cumbersome.

Methods: Instead of a single codec, we propose a multiple codecs structure based on the FC-DenseNet (FCD) model and capsule convolution-capsule deconvolution, named Nested Capsule Dense Network (NCDN). NCDN uses multiple codecs to achieve multi-resolution, which makes it possible to save more spatial information and improve the robustness of the model.

Results: The proposed model is tested on three datasets that include the York University Cardiac MRI dataset, Automated Cardiac Diagnosis Challenge (ACDC-2017), and the local dataset. The results show that the proposed NCDN outperforms most methods. In particular, we achieved nearly the most advanced accuracy performance in the ACDC-2017 segmentation challenge. This means that our method is a reliable segmentation method, which is conducive to the application of deep learning-based segmentation methods in the field of medical image segmentation.

Keywords: Cardiac diagnosis; Multiple codecs; Nested Capsule Dense Network.

Grants and funding

This work was supported by the National Science Foundation Program of China (NSFC) under Grants 61976241, 61871173, by major research program of National Science Foundation of China (NSFC) under Grant 91948303, by the Tianjin Science and Technology Planning Project under Grants 19ZXJRGX00080, by Science and Technology Program of Tianjin under Grant 20YDTPJC00670, and by the project of the Left and right ventricle segmentation method of cardiac MRI images, Shanxi Key Laboratory of Biomedical Imaging and Image Big Data, North University of China. There was no additional external funding received for this study. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.