Related Experiment Video
Updated: Feb 21, 2026
09:57
Development and Evaluation of 3D-Printed Cardiovascular Phantoms for Interventional Planning and Training
Published on: January 18, 2021
4.7K
Isotropic 3D cardiac cine MRI allows efficient sparse segmentation strategies based on 3D surface reconstruction
Freddy Odille1,2, Aurélien Bustin1,3,4, Shufang Liu1,3,4
1IADI, INSERM U947 and Université de Lorraine, Nancy, France.
Magnetic Resonance in Medicine
|October 4, 2017
Summary
Optimizing cardiac MRI analysis, this study shows sparse segmentation of 3D cine datasets improves precision for volumetric analysis. This advanced technique enhances cardiac function assessment using fewer slices.
Area of Science:
- Medical Imaging
- Cardiovascular MRI
- Image Segmentation
Background:
- Cardiac cine MRI segmentation is crucial for volumetric analysis of cardiac function.
- Conventional methods use 2D short-axis (SAX) image stacks with thick slices, potentially limiting workflow efficiency.
- There is a need for optimized segmentation strategies for 3D isotropic cardiac cine datasets.
Purpose of the Study:
- To develop and evaluate an optimized manual segmentation workflow for isotropic 3D cine cardiac MRI.
- To assess the feasibility of using sparse segmentation strategies on reformatted slices from 3D datasets.
- To compare the accuracy and precision of sparse segmentation against conventional SAX segmentation.
Main Methods:
- Isotropic 3D cine datasets were generated using nonrigid motion correction (cine-GRICS) and super-resolution from free-breathing SAX and long-axis (LAX) cine stacks.
- Manual segmentation strategies were compared, including conventional SAX, LAX-only, and combined SAX/LAX sparse slice segmentation.
- An implicit B-spline surface reconstruction algorithm was employed for left ventricular cavity surface reconstruction from sparse 2D contours.
Main Results:
- All tested sparse segmentation strategies achieved high agreement (Dice scores > 0.9) using fewer slices (3-6) compared to conventional methods (8-10 slices).
- Stroke volumes from sparse segmentation (4-6 slices) showed improved precision (SD 5.4 mL vs. 6.1 mL) but slightly lower accuracy (bias -1.2 mL vs. 0.2 mL) than conventional SAX segmentation.
- Functional parameters, including end-diastolic volumes, end-systolic volumes, and ejection fractions, demonstrated a trend towards improved precision.
Conclusions:
- The postprocessing workflow for 3D isotropic cardiac imaging can be optimized using sparse segmentation and 3D surface reconstruction.
- This approach enhances the efficiency and precision of manual segmentation for cardiac volumetric analysis.
- Sparse segmentation offers a viable alternative for optimizing cardiac MRI analysis workflows.
