Research

The availability of data to objectively evaluate and compare different reconstruction methods could expedite innovation and promote clinical translation of these methods. In this work, we introduce an open-access dataset, called OCMR, that provides multi-coil k-space data from 74 fully sampled and 212 prospectively undersampled cardiac cine series, comprising of 183 and 842 slices, respectively.

Website:https://ocmr.info/ 

Code: https://github.com/MRIOSU/OCMR 

Document: https://arxiv.org/abs/2008.03410 

Segmented cardiac phase-contrast (PC) MRI relies on ECG synchronization and breath holding (BH), which fails in patients with arrhythmia or poor respiratory control. Real-time (RT) PC-MRI overcomes these limitations; however, compressed sensing (CS)-based reconstruction methods often underestimate the peak velocity in accelerated RT PC-MRI. To overcome this issue, we propose a deep image prior (DIP) [1]-based reconstruction method, called FlowDIP.

Publication: FlowDIP: Real-Time Phase-Contrast MRI Reconstruction with Flow-Conditional Deep Image Prior   accepted for SCMR2025

Free-breathing (FB) real-time phase contrast (RT-PC) MR is very useful to resolve the beat-by-beat variations and for patients for whom breath-holding poses a challenge; however its feasibility remains to be tested at low-field. In this work, FB RT-PC was developed using a dual-density spiral readout with a modified golden-angle rotation strategy and compressed sensing reconstruction. The proposed RT PC MR technique for low-field systems could enable flow imaging for patients with arrhythmias, critical illnesses, or claustrophobia, potentially making cardiac MRI more accessible and patient-friendly in a variety of clinical settings.

Publication: Spiral Real-time Phase Contrast MR on a 0.55T MRI System  ISMRM2024

Exercise stress CMR is important since some cardiac conditions that do not present at rest may be identified via cardiac MRI stress imaging.The artifacts caused by the coil movement are also visible in the temporally averaged coil images. To suppress the motion artifacts, the k-space from certain coils can be excluded by visual inspection of the coil images. However, visual inspection is not practical for routine use. In this work, we propose a method to automatically suppress coils that introduce a high level of artifacts to the reconstructed image. 

Publication: Automatic coil selection to suppress motion artifacts for exercise real-time cine imaging  ISMRM2023 oral

To monitor both cardiac and respiratory motions during the MRI acquisition, we are developing methods based on the Pilot Tone (PT) technology. PT is a transmitter that emits electromagnetic waves close to the Larmor frequency. The transmitted signal is modulated by the physiological motions and is picked up by the receive coils.  With PT, the physiological motions are seamlessly encoded into the raw MRI data and can be separated from the image content using signal processing techniques.

Publication: Extraction of cardiac and respiratory motion from Pilot Tone—a patient study (SCMR2021)

                              Cardiac and respiratory motion extraction for MRI using pilot tone–a patient study. The International Journal of Cardiovascular Imaging, 40(1), 93-105. 

Regularized reconstruction methods, such as compressed sensing (CS), can significantly accelerate MRI data acquisition but require tuning of regularization weights. In this work, a technique, called Sparsity adaptive Composite Recovery (SCoRe) that exploits sparsity in multiple, disparate sparsifying transforms is presented. A data‐driven adjustment of the relative contributions of different transforms yields a parameter‐free CS recovery process. SCoRe is validated in a dynamic digital phantom as well as in retrospectively and prospectively undersampled cine CMR data.

Code: https://github.com/MRIOSU/SCoRe_demo 

Publication: Sparsity adaptive reconstruction for highly accelerated cardiac MRI ( https://onlinelibrary.wiley.com/doi/epdf/10.1002/mrm.27671)

 SCoRe was implemented in the Gadgetron framework and used in three applications:  real-time cine/flow (with pseudo-random Cartesian sampling) and EPI-Perfusion (with rate 2 time-interleaved sampling ). With the help of GPU, the reconstruction time is ~3 s for real-time cine (1.5 s scan,  30 frames,  matrix size: 176x134), ~6 s for real-time flow (2.5 s scan, 47 frames, matrix size: 144x108 ), and ~4 s for EPI-Perfusion (60 repetitions, matrix size: 144x136 ), which is clinically acceptable. This technique was deployed to the clinical scanners at Richard M. Ross Heart Hospital (RHH@OSU) and Nationwide Children's Hospital (NCH) in March 2019. Up to Nov. 2021, over 1000 patients at RHH have benefited from this technique.

Publications: Improved segmented-EPI cardiac perfusion imaging with compressed sensing reconstruction (SCMR2021)

   Inline real-time phase-contrast cardiac MRI with Cartesian sampling (SCMR2020)

                           High-resolution exercise stress real-time cine imaging (SCMR2019)

                Reconstructing real-time exercise stress cine images with multiple sets of sensitivity maps (ISMRM2018)

The respiratory signal is extracted by a principal component analysis method from RT cine images. Then, a two- step procedure is used to determine the directionality (sign) of the respiratory signal. With the motion in a manually selected region- of- interest as a reference, the quality of the extracted respiratory signal is assessed using multislice RT cine data from 11 volunteers and 10 patients. Using the extracted respiratory signal as a guide, we selected heartbeats at PE and PI, as well as at random respiratory phases from each slice and demonstrated both the respiration- induced variations in cardiac output and the impact of selecting heartbeats with inconsistent respiratory phase.

Code: https://github.com/MRIOSU/AERC 

Publication: Automatic Extraction and Sign Determination of Respiratory Signal in Real-Time Cardiac Magnetic Resonance Imaging 2020 ISBI

                               Ensuring respiratory phase consistency to improve cardiac function quantification in real-time CMR (https://onlinelibrary.wiley.com/doi/10.1002/mrm.29064)

Radial k-space sampling has drawn continuous interest in cardiac MRI due to its motion robustness. However, it is sensitive to the scanner imperfections, which can lead to inconsistencies in k-space trajectories and degrade image quality.  A number of methods have been proposed to perform the trajectory correction [1,2,3]; however, they do not fully suppress the artifacts, especially from the outer region of the images. In this work, we propose to use region-optimized virtual (ROVir) [4] coil selection to suppress the signal and hence the artifact originating from the areas outside of the region of interest (ROI) in cardiac real-time (RT) cine imaging.

Publication: Radial artifact reduction in cardiac cine imaging using Region-optimized Virtual coil selection  SCMR 2022