Related Experiment Video
Updated: Aug 13, 2026
11:13
Quantification of Mouse Heart Left Ventricular Function, Myocardial Strain, and Hemodynamic Forces by Cardiovascular Magnetic Resonance Imaging
Published on: May 24, 2021
Comparative evaluation of artificial intelligence-deep learning-based cine short-axis segmentation and manual
Yanghao Lin1, Wenwei Zhong1, Junfeng Yao1
1Department of Radiology, The Third Affiliated Hospital of Sun Yat-sen University Yuedong Hospital, Meizhou, China.
Quantitative Imaging in Medicine and Surgery
|August 12, 2026
Summary
The standard deep learning AI method for cardiac MRI analysis shows superior agreement with manual segmentation compared to a simplified approach. This AI tool significantly reduces analysis time while maintaining clinical accuracy for assessing left ventricular function and structure.
Area of Science:
- Cardiovascular Imaging
- Artificial Intelligence in Medicine
- Medical Image Analysis
Background:
- Cardiac magnetic resonance (CMR) is the gold standard for assessing left ventricular (LV) function and structure.
- Manual CMR post-processing is time-consuming, subjective, and lacks reproducibility.
- Deep learning artificial intelligence (AI-DL) offers automated segmentation but requires clinical validation.
Purpose of the Study:
- To compare two AI-DL segmentation strategies (simplified vs. standard) against manual analysis for CMR.
- To determine which AI-DL approach provides superior clinical agreement for LV parameter quantification.
- To evaluate the impact of AI-DL on analysis time and reproducibility.
Main Methods:
- Retrospective analysis of CMR data from 104 patients.
- Quantification of LV parameters (EF, EDV/EDVI, ESV/ESVI, SV/SI, CO/CI, LV-mass/MI) using manual and two AI-DL methods (simplified, standard).
- Comparison of analysis time, correlation, and agreement (Bland-Altman LOA) with manual results.
Main Results:
- AI-DL significantly reduced analysis time (50s vs. 217s, P<0.001).
- The standard AI-DL method showed no significant differences from manual analysis for LV parameters (P>0.05).
- The simplified AI-DL method demonstrated significant under/overestimation of parameters (P<0.001) with wider LOA compared to the standard method.
Conclusions:
- AI-DL enables rapid and reliable quantification of LV function and structure in CMR.
- The standard AI-DL segmentation method offers superior clinical agreement and reproducibility compared to the simplified method.
- The standard AI-DL approach is a clinically acceptable tool for routine cardiac functional assessment.