Automatic left ventricle volume and mass quantification from 2D cine-MRI: Investigating papillary muscle influence

Area of Science:

• Cardiology
• Medical Imaging
• Artificial Intelligence

Background:

• Accurate quantification of left ventricle (LV) function is crucial for early cardiovascular disease detection.
• Current methods may lack automation and precision in assessing LV parameters.

Purpose of the Study:

• To develop and validate a fully automatic framework for LV volume and mass quantification from 2D-cine MR images.
• To enable accurate detection and quantification of cardiac contraction abnormalities.

Main Methods:

• Utilized a Convolutional Neural Network (CNN) for automatic LV localization and morphological operations to refine the LV cavity.
• Employed U-Net architecture for automatic LV contour extraction.
• Integrated temporal information from myocyte motion to calculate LV volume, LV ejection fraction (LVEF), and left ventricle mass (LVM).

Main Results:

• Achieved high accuracy (99.15%) in LV localization using CNN.
• Demonstrated excellent segmentation performance with U-Net (DSC 99.78%, HD 4.468 mm).
• Showcased strong correlations between quantified parameters and physiological measures (PCC > 0.92 for LV volume, LVEF, stroke volume, and LVM).

Conclusions:

• The proposed automated framework accurately quantifies LV volume and function.
• The method shows significant potential for clinical application as a diagnostic aid for cardiac contraction abnormalities.