Cape Town - 2026 ISMRM-ISMRT Annual Meeting and Exhibition • 09-14 May 2026
|
467-04-001.
3D myocardial LGE imaging in under 4 minutes using a self-supervised deep learning-based motion compensated reconstruction
Impact: High-resolution 3D LGE with short,
predictable scan times is achieved. In patient data, myocardial scarring is confirmed
using 2D LGE. The proposed method enables rapid 3D LGE as a potential
alternative to conventional 2D imaging.
|
||
|
467-04-002.
Unsupervised Patch-Based Cardiac Cine MRI Reconstruction Using Learnable Tensor Function with Implicit Neural Representation
Impact: This framework shows strong potential for faster
and more reliable dynamic cardiac MRI in clinical applications.
|
||
|
467-04-003.
ESPIRiT guided Implicit Neural Representations with Multi-Resolution Hash Encoding for Self-supervised Cardiac Cine MRI Recon
Impact: Fast INR-based
reconstructions enable shorter breath-holds, making cardiac CINE feasible for
patients with limited breath-holding capacity or difficulty following
instructions, without needing fully-sampled ground truth training data.
|
||
|
467-04-004.
FID-self-navigation to track physiological motion with high temporal resolution in 3D radial MRI of the heart
Impact: This work demonstrates the use of an FID readout for physiological
motion extraction at high temporal resolution, enabling motion-resolved free-breathing
cardiac MRI without external sensors, thereby improving motion-robustness
of free-breathing scans.
|
||
|
467-04-005.
S³-MUSIC: Physics-guided Self-Supervised 4D Ferumoxytol-based Steady-State Cardiac MRI Reconstruction with 32× Acceleration
Impact: 4D multiphase steady-state imaging with ferumoxytol-enhancement
reconstructed via a self-supervised spatiotemporal deep learning model (S3-MUSIC)
enabled highly accelerated cardiac imaging (up to 32×) with preserved image fidelity.
The framework can advance efficient and reliable ferumoxytol-enhanced MR
angiography and cardiac MRI.
|
||
|
467-04-006.
2D radial self-navigation for free-breathing wideband 3D stack-of-stars LGE in patients with an ICD: rotating vs. fixed view
Impact: Wideband 3D-isotropic stack-of-stars
LGE with 2D radial self-navigation using fixed-view angles enhanced image
quality compared with that using continuously rotating-view angles, while improving
image reconstruction efficiency, enabling faster and more practical LGE imaging
in patients with implantable cardioverter-defibrillators.
|
||
|
467-04-007.
Real-time Tiny Golden Angle Phase Contrast using bSSFP (PC-SSFP) and model consistency condition (MOCCO) Reconstruction
Impact: Real-time PC-SSFP with MOCCO reconstruction
offers high-SNR, free-breathing flow imaging with accurate visualization of E/A
filling patterns, providing a promising approach for robust and rapid
assessment of diastolic function in clinical cardiac MRI.
|
||
|
467-04-008.
Respiratory and Cardiac Self-navigated Free-Breathing Fatty Acid Composition MRI of Human Epicardial Adipose Tissue
Impact: Self-navigated
motion-resolved free-breathing radial MRI enabled reliable, high-resolution EAT
FAC quantification without breath-holding. These methods promise to facilitate
the investigation of EAT composition in cardiovascular disease.
|
||
|
467-04-009.
Improved Sharpness and Quality in Dyspnea and Arrhythmia by Free Breathing Compressed Sensing Single-Shot MOCO Averaged LGE
Impact: Leveraging the speed and SNR of compressed sensing, our free
breathing MOCO averaged delayed enhancement (LGE) technique allows dyspneic and arrhythmic
patients to experience image sharpness and quality previously reserved for segmented
breath held imaging, thereby clinically benefiting these patients.
|
||
|
467-04-010.
Evaluation of Quantitative High-Resolution Spiral First-Pass Perfusion Imaging Using Motion-Corrected Subspace Reconstruction
Impact: Subspace-MOCO
yielded high-fidelity quantitative perfusion values consistent with reference
measurements, along with enhanced image sharpness and clear myocardial
delineation.
|
||
|
467-04-011.
Free-breathing fast 3D whole-heart cardiac CEST imaging at 3T
Impact: A whole-heart 3D cardiac CEST approach that delivers
high-quality creatine/phosphocreatine mapping within 9 minutes is demonstrated.
This capability broadens the clinical and research use of metabolic cardiac
imaging for disease characterization, therapy monitoring, facilitating free-breathing,
robust and motion-resolved acquisition.
|
||
|
467-04-012.
Computationally efficient respiratory motion-informed reconstruction for 4D flow MRI
Impact: We
propose a deep learning-based 4D flow MRI reconstruction framework, running on commodity
CPU/GPU hardware, that incorporates respiratory motion to improve velocity
quantification for highly accelerated aortic 4D flow MRI.
|
||
|
467-04-013.
3D radial 4D flow MRI with rigid motion detection and correction
Impact: This study combines a 3D radial 4D flow acquisition with eddy
current corrections and rigid motion detection and correction, allowing motion-
and undersampling-robust imaging that may improve fetal 4D flow MRI
acquisitions for early detection of congenital heart disease.
|
||
|
467-04-014.
Free-breathing cardiac T1 mapping using frequency-modulated continuous-wave radar for prospective respiratory tracking
Impact: Our approach enables free-breathing cardiac T1 mapping via contact-free radar slice-tracking, which improves subject comfort, scan throughput, and reproducibility.
|
||
|
467-04-015.
Cardiac triggering based on ultrasound sensor signals
Impact: The proposed ultrasound sensor system was validated at 3T for its ability to provide reliable cardiac gating signals. This system is expected to be particularly helpful in situations where electrocardiograms are not available, such as in high-field environments.
|
||
|
467-04-016.
Automatic Coil Movement Detection and Triggering Prescan Reacquisition Using 3D Hall Sensor
Impact: This work enables detection of coil position changes caused by, e.g., operator adjustment or
patient motion, using 3D Hall sensor, triggering prescan data reacquisition to
prevent image artifacts from incorrect prescan data. It guarantees consistent image quality for clinical diagnosis.
|
© 2026 International Society for Magnetic Resonance in Medicine