Free-breathing T2* mapping using respiratory motion corrected averaging

Peter Kellman, Hui Xue, Bruce Spottiswoode, Christopher M. Sandino, Michael S. Hansen, Amna Abdel‐Gadir, Thomas A. Treibel, Stefania Rosmini, Christine Mancini, W. Patricia Bandettini, Laura‐Ann McGill, Peter Gatehouse, James Moon, Dudley J. Pennell, Andrew E. AraiJournal of Cardiovascular Magnetic Resonance 2015 · DOI

Abstract

A free-breathing approach to T2* mapping is demonstrated to produce consistently good quality maps in the presence of respiratory motion and arrhythmias.

Introduction

T2* mapping is the clinical gold standard for measuring iron concentration in the heart and liver, which is vital for managing conditions like Thalassemia. Traditionally, this is achieved using breath-held segmented sequences. However, these sequences are highly sensitive to patient motion and arrhythmias, often resulting in "ghosting" artifacts that render the images non-diagnostic. This is particularly problematic for pediatric, elderly, or heart failure patients who struggle to hold their breath for the required duration.

Methodology

To address these challenges, researchers developed a free-breathing T2* mapping technique utilizing motion-corrected averaging of single-shot images. Unlike segmented acquisitions, single-shot imaging captures the entire image in one go, making it immune to ghosting artifacts.

Traditional breath-held segmented acquisition sequence, which is prone to motion artifacts.

The proposed workflow involves:

  1. Multiple Repetitions: Acquiring 16 repetitions of the same slice during free breathing.
  2. Non-Rigid Motion Correction (MOCO): Aligning these repetitions to compensate for cardiac and diaphragmatic movement.
  3. Selective Averaging: Discarding the 50% of images with the most through-plane motion to improve sharpness.
  4. Automatic Truncation: Removing late-echo data points that fall into the noise floor to prevent bias.
Proposed free-breathing measurement sequence utilizing multiple repetitions for motion correction.

The entire process is integrated into the scanner using the Gadgetron framework, providing finished maps within seconds of the scan ending.

Results and Discussion

The method was validated across two cohorts: a general cardiac cohort (NIH) and a Thalassemia cohort (Royal Brompton). In the NIH cohort, while 30% of traditional scans were non-diagnostic due to artifacts, the free-breathing MOCO maps achieved a 100% diagnostic success rate.

T2* values from the free-breathing method showed exceptional correlation with the breath-held gold standard ($r^2 = 0.93$ to $>0.99$), with a negligible mean bias of less than 1 ms. Blinded observers consistently rated the MOCO maps higher in quality than traditional segmented maps.

Conclusion

This "push-button" free-breathing approach provides a robust, clinically reliable alternative to traditional T2* mapping. By converting motion artifacts from ghosting to manageable blurring and using selective averaging, it ensures high-quality diagnostic data even in challenging patient populations.

Summary updated

Abstract

A free-breathing approach to T2* mapping is demonstrated to produce consistently good quality maps in the presence of respiratory motion and arrhythmias.
BibTeX
@article{kellman2015free,
  author = {Peter Kellman and Hui Xue and Bruce Spottiswoode and Christopher M. Sandino and Michael S. Hansen and Amna Abdel‐Gadir and Thomas A. Treibel and Stefania Rosmini and Christine Mancini and W. Patricia Bandettini and Laura‐Ann McGill and Peter Gatehouse and James Moon and Dudley J. Pennell and Andrew E. Arai},
  title = {{Free-breathing T2* mapping using respiratory motion corrected averaging}},
  journal = {Journal of Cardiovascular Magnetic Resonance},
  year = {2015},
  doi = {10.1186/s12968-014-0106-9},
}
Figure