Related Experiment Video
Updated: Jun 9, 2026
06:53
3D Whole-heart Myocardial Tissue Analysis
Published on: April 12, 2017
A strategic approach for cardiac MR left ventricle segmentation
Sarada Prasad Dakua1, J S Sahambi
1Department of Electronics and Communication Engineering, Indian Institute of Technology, Guwahati, India. sarada@iitg.ernet.in
Cardiovascular Engineering (Dordrecht, Netherlands)
|September 3, 2010
Summary
This study introduces an automated method for outlining the left ventricle in cardiac magnetic resonance (CMR) images, improving cardiac function analysis. The novel approach enhances accuracy in segmenting myocardial walls, particularly for patients with severe conditions.
Area of Science:
- Medical Imaging
- Cardiovascular Imaging
- Image Segmentation
Background:
- Quantitative evaluation of cardiac function using cardiac magnetic resonance (CMR) necessitates precise identification of myocardial walls.
- Manual outlining of left ventricular contours in CMR images is challenging, especially for patients with severe diseases, often requiring interactive tracing by clinicians.
- Accurate segmentation of the left ventricle is crucial for reliable cardiac function assessment.
Purpose of the Study:
- To propose and evaluate an automated approach for outlining the left ventricular contour in CMR images.
- To improve the efficiency and accuracy of myocardial wall segmentation in challenging CMR datasets.
- To facilitate quantitative evaluation of cardiac function by automating a critical image analysis step.
Main Methods:
- A combined segmentation strategy integrating two distinct approaches for left ventricle segmentation.
- Introduction of a Difference of Gaussian (DoG) weighting function within the random walk algorithm for blood pool (inner contour) extraction.
- Application of a modified active contour method, using the extracted blood pool boundary as the initial contour, for myocardial wall (outer contour) segmentation.
Main Results:
- The proposed automated method successfully segmented the left ventricle in CMR images.
- Experimental results demonstrated promising performance in outlining both inner and outer myocardial contours.
- The approach showed potential for accurate segmentation, even in CMR images from subjects with serious diseases.
Conclusions:
- The developed automated outlining approach for the left ventricle shows significant potential for improving cardiac function analysis from CMR images.
- The combination of Difference of Gaussian weighting in random walks and modified active contours offers an effective solution for challenging left ventricular segmentation tasks.
- This method can aid clinicians by automating a difficult segmentation process, leading to more consistent and efficient quantitative evaluations.
