A deep learning-based method for automatic segmentation of free-running 4D cardiac MRI achieved high geometric accuracy (DSC 0.94 for LV blood pool) and excellent clinical metric agreement.
Does a deep learning-based segmentation framework provide accurate and fast anatomical and functional analysis of free-running 4D whole-heart cardiac MRI?
A deep learning-based segmentation method provides fast, accurate, and clinically reliable anatomical and functional analysis of whole-heart free-running 4D cardiac MRI.
BACKGROUND: Free-running (FR) cardiac magnetic resonance imaging (MRI) enables free-breathing electrocardiogram (ECG)-free fully dynamic five-dimensional (5D) (three-dimensional 3D spatial+cardiac+respiration dimensions) imaging but poses significant challenges for clinical integration due to the volume of data and complexity of image analysis. Existing segmentation methods are tailored to two-dimensional (2D) cine or static 3D acquisitions and cannot leverage the unique spatial-temporal wealth of FR data. The aim of this study was to develop and validate a deep learning (DL)-based segmentation framework for isotropic 3D+cardiac cycle FR cardiac MRI that enables accurate, fast, and clinically meaningful anatomical and functional analysis. METHODS: Free-running, contrast-free balanced steady-state free precession (bSSFP) acquisitions at 1.5T and contrast-enhanced gradient-recalled echo (GRE) acquisitions at 3T were used to reconstruct motion-resolved 5D datasets. From these, the end-expiratory respiratory phase was retained to yield fully isotropic four-dimensional (4D) datasets. Automatic propagation of a limited set of manual segmentations was used to segment the left and right ventricular blood pool (LVB, RVB) and left-ventricular myocardium (LVM) on reformatted short-axis (SAX) end-systolic (ES) and end-diastolic (ED) images. These were used to train a 3D nnU-Net model. Validation was performed using geometric metrics (Dice similarity coefficient DSC, relative volume difference RVD), clinical metrics (ED and ES volumes, ejection fraction EF), and physiological consistency metrics (systole-diastole LVM volume mismatch and LV-RV stroke volume agreement). To assess the robustness and flexibility of the approach, we evaluated multiple additional DL training configurations, such as using 4D propagation-based data augmentation to incorporate all cardiac phases into training. RESULTS: The main proposed method achieved automatic segmentation within a minute, delivering high geometric accuracy and consistency (DSC: 0.94 ± 0.01 LVB, 0.86 ± 0.02 LVM, 0.92 ± 0.01 RVB; RVD: 2.7%, 5.8%, 4.5%). Clinical LV metrics showed excellent agreement (ICC >0.98 for EDV/ESV/EF, bias 0.93 for EDV/ESV/EF, bias <1 mL for EDV/ESV, <1% for EF) but exhibited wider limits of agreement. Training on all cardiac phases improved temporal coherence, reducing LVM volume mismatch from 4.0% to 2.6%. CONCLUSION: This study validates a DL-based method for fast and accurate segmentation of whole-heart free-running 4D cardiac MRI. Robust performance across diverse protocols and evaluation with complementary metrics that match state-of-the-art benchmarks supports its integration into clinical and research workflows, helping to overcome a key barrier to the broader adoption of free-running imaging.
Ogier et al. (Wed,) reported a other. Deep learning-based segmentation framework (3D nnU-Net) was evaluated on Geometric metrics (DSC, RVD), clinical metrics (EDV, ESV, EF), and physiological consistency metrics. A deep learning-based method for automatic segmentation of free-running 4D cardiac MRI achieved high geometric accuracy (DSC 0.94 for LV blood pool) and excellent clinical metric agreement.