Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2020 Oct:12266:468-477.
doi: 10.1007/978-3-030-59725-2_45. Epub 2020 Sep 29.

A Semi-supervised Joint Network for Simultaneous Left Ventricular Motion Tracking and Segmentation in 4D Echocardiography

Kevinminh Ta  1 , Shawn S Ahn  1 , John C Stendahl  2 , Albert J Sinusas  2   3 , James S Duncan  1   4   3
Affiliations

A Semi-supervised Joint Network for Simultaneous Left Ventricular Motion Tracking and Segmentation in 4D Echocardiography

Kevinminh Ta et al. Med Image Comput Comput Assist Interv. 2020 Oct.

Abstract

This work presents a novel deep learning method to combine segmentation and motion tracking in 4D echocardiography. The network iteratively trains a motion branch and a segmentation branch. The motion branch is initially trained entirely unsupervised and learns to roughly map the displacements between a source and a target frame. The estimated displacement maps are then used to generate pseudo-ground truth labels to train the segmentation branch. The labels predicted by the trained segmentation branch are fed back into the motion branch and act as landmarks to help retrain the branch to produce smoother displacement estimations. These smoothed out displacements are then used to obtain smoother pseudo-labels to retrain the segmentation branch. Additionally, a biomechanically-inspired incompressibility constraint is implemented in order to encourage more realistic cardiac motion. The proposed method is evaluated against other approaches using synthetic and in-vivo canine studies. Both the segmentation and motion tracking results of our model perform favorably against competing methods.

Keywords: Echocardiography; Motion tracking; Segmentation.

PubMed Disclaimer

Figures

Fig. 1.
Fig. 1.
Architecture of our proposed joint network: The motion branch (top) and the segmentation branch (bottom).
Fig. 2.
Fig. 2.
A short-axis view of the displacement vectors for a normal (healthy) synthetic sequence using different methods
Fig. 3.
Fig. 3.
Epicardium (green) and endocardium (red) segmentations for a normal (healthy) synthetic sequence using different methods
Fig. 4.
Fig. 4.
A short-axis view of the displacement vectors for a normal (healthy) in vivo sequence using different methods: A) crystal derived displacement, B) nonrigid registration (NRR), C) Lucas-Kanade Optical Flow (LK), D) Shape Tracking (ST), E) Unsupervised w/ shape regularizer (Unsup+Shape), F) Unsupervised G) Proposed Model
Fig. 5.
Fig. 5.
Epicardium (green) and endocardium (red) segmentations for a normal (healthy) in vivo sequence using different methods: A) manual, B) dynamic appearance model (DAM), C) trained with crystal-generated labels (Seg-CD) (LK), D) nonrigid registration-generated labels (Seg-NRR), E) Lucas-Kanade generated labels (Seg-LK), F) Unsupervised motion generated labels (Seg-Unsup) G) Proposed Model
See this image and copyright information in PMC

References

    1. Ahn SS, Ta K, Lu A, Stendahl JC, Sinusas AJ, Duncan JS: Unsupervised motion tracking of left ventricle in echocardiography. In: Medical Imaging 2020: Ultrasonic Imaging and Tomography, p. 113190Z (2020) - PMC - PubMed
    1. Alessandrini M, et al.: A pipeline for the generation of realistic 3D synthetic echocardiographic sequences: Methodology and open-access database. IEEE Trans. Med. Imaging 34, 1436–1451 (2015) - PubMed
    1. Balakrishnan G, et al.: An unsupervised learning model for deformable medical image registration. In: The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2018)
    1. Cheng J, et al.: Segflow: joint learning for video object segmentation and optical flow. In: IEEE International Conference on Computer Vision (ICCV) (2017)
    1. Compas C, et al.: Radial basis functions for combining shape and speckle tracking in 4D echocardiography. IEEE Trans. Med. Imaging 33, 1275–1289 (2014) - PMC - PubMed

LinkOut - more resources