Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

DDSB: An Unsupervised and Training-free Method for Phase Detection in Echocardiography

Zhenyu Bu, Yang Liu, Jiayu Huo, Jingjing Peng, Kaini Wang, Guangquan Zhou, Rachel Sparks, Prokar Dasgupta, Alejandro Granados, Sebastien Ourselin

TL;DR

This work proposes an unsupervised and training-free method that leverages unsupervised segmentation to enhance fault tolerance against segmentation inaccuracies and effectively reduce dependence on the accuracy of initial segmentation images and enhance fault tolerance, all while improving robustness.

Abstract

Accurate identification of End-Diastolic (ED) and End-Systolic (ES) frames is key for cardiac function assessment through echocardiography. However, traditional methods face several limitations: they require extensive amounts of data, extensive annotations by medical experts, significant training resources, and often lack robustness. Addressing these challenges, we proposed an unsupervised and training-free method, our novel approach leverages unsupervised segmentation to enhance fault tolerance against segmentation inaccuracies. By identifying anchor points and analyzing directional deformation, we effectively reduce dependence on the accuracy of initial segmentation images and enhance fault tolerance, all while improving robustness. Tested on Echo-dynamic and CAMUS datasets, our method achieves comparable accuracy to learning-based models without their associated drawbacks. The code is available at https://github.com/MRUIL/DDSB

DDSB: An Unsupervised and Training-free Method for Phase Detection in Echocardiography

TL;DR

This work proposes an unsupervised and training-free method that leverages unsupervised segmentation to enhance fault tolerance against segmentation inaccuracies and effectively reduce dependence on the accuracy of initial segmentation images and enhance fault tolerance, all while improving robustness.

Abstract

Accurate identification of End-Diastolic (ED) and End-Systolic (ES) frames is key for cardiac function assessment through echocardiography. However, traditional methods face several limitations: they require extensive amounts of data, extensive annotations by medical experts, significant training resources, and often lack robustness. Addressing these challenges, we proposed an unsupervised and training-free method, our novel approach leverages unsupervised segmentation to enhance fault tolerance against segmentation inaccuracies. By identifying anchor points and analyzing directional deformation, we effectively reduce dependence on the accuracy of initial segmentation images and enhance fault tolerance, all while improving robustness. Tested on Echo-dynamic and CAMUS datasets, our method achieves comparable accuracy to learning-based models without their associated drawbacks. The code is available at https://github.com/MRUIL/DDSB
Paper Structure (16 sections, 7 equations, 3 figures, 2 tables)

This paper contains 16 sections, 7 equations, 3 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Method
  3. Unsupervised cavity segmentation
  4. Anchor points picker
  5. Temporal expansion-contraction discriminator
  6. Dataset and Metrics
  7. Dataset
  8. Metrics
  9. Results and Discussion
  10. Comparisons with State-of-the-art Methods
  11. In-dataset Evaluation.
  12. Cross-dataset Evaluation.
  13. Ablation Study
  14. Effects of $k$ directions.
  15. Effects of change threshold $\alpha$.
  16. ...and 1 more sections

Figures (3)

  • Figure 1: Example of cardiac ultrasound sequence with ED in Red and ES in Green.
  • Figure 2: The overview of our DDSB for unsupervised and training-free ED/ES detection.
  • Figure 3: Comparison of the size-based method with our DDSB.