Abstract
This paper presents a novel MRI pulse sequence for simultaneous mapping of myocardial $T_1$ and $T_2$ relaxation times within a single breath-hold. By applying a saturation pulse in every R-R interval to eliminate magnetization history, followed by a recovery period ($T_{sat}$) and a $T_2$-preparation pulse ($TE_{prep}$), the sequence generates interleaved $T_1$ and $T_2$ weighting. This approach ensures perfectly registered maps and heart-rate independence, providing a highly efficient diagnostic tool for cardiac assessment.
Introduction
In clinical cardiac MRI, $T_1$ and $T_2$ mapping are essential for characterizing myocardial tissue. $T_1$ mapping is used to identify focal and diffuse fibrosis, while $T_2$ mapping is the gold standard for detecting edema and inflammation. Traditionally, these parameters are acquired using separate pulse sequences.
This conventional approach suffers from two major drawbacks:
- Spatial Misalignment: Patient motion or breathing between separate scans often results in maps that do not line up perfectly, complicating joint analysis.
- Reduced Efficiency: Acquiring two separate sets of images requires multiple breath-holds, which can be challenging for patients.
While joint estimation techniques have been proposed previously, they often require long acquisition times or are sensitive to heart rate variations. This work introduces a saturation-recovery based sequence designed to overcome these limitations.
Methods
The proposed technique utilizes a saturation-recovery approach to ensure the signal is independent of the patient's heart rate.
The Pulse Sequence
The sequence consists of four main components per heartbeat:
- Saturation Pulse: Resets the magnetization history to zero at the start of every R-R interval.
- Recovery ($T_{sat}$): Longitudinal magnetization recovers for a specific duration.
- $T_2$-Preparation: A $T_2$-prep pulse with a specific echo length ($TE_{prep}$) adds $T_2$ weighting.
- Readout: A single-shot Steady-State Free Precession (SSFP) image is acquired.
Acquisition Strategy
The mapping sequence acquires a total of 13 images over 13 heartbeats. The first image is an "anchor" acquired without magnetization preparation. The subsequent 12 images vary both $T_{sat}$ and $TE_{prep}$ values to sample the relaxation curves effectively.
Parameter Estimation
$T_1$ and $T_2$ values are estimated simultaneously via voxel-wise least-squares fitting to a 4-parameter signal model: $$S \approx A \cdot (1 - \exp(-T_{sat}/T_1)) \cdot \exp(-TE_{prep}/T_2) + B$$
Results
The technique was validated through phantom studies and in-vivo testing on five healthy subjects.
Phantom Validation
In a study of 14 vials with diverse relaxation properties, the joint mapping technique showed excellent agreement with gold-standard references (Inversion-Recovery for $T_1$ and CPMG spin-echo for $T_2$), with no significant differences observed ($P > 0.4$).
In-vivo Performance
In healthy subjects, the joint maps provided values nearly identical to established single-parameter sequences in the myocardial septum:
- $T_1$: $1210 \pm 92$ ms (Joint) vs. $1211 \pm 82$ ms (SASHA).
- $T_2$: $47.3 \pm 6.5$ ms (Joint) vs. $49.0 \pm 5.8$ ms (Conventional).
The entire acquisition was completed within 13 heartbeats, fitting comfortably within a single breath-hold.
Discussion
The proposed joint mapping sequence offers several key advantages. First, it provides inherent registration; because both parameters are derived from the same raw data, the resulting maps are perfectly aligned by design. Second, the use of saturation pulses ensures heart-rate independence, making the scan robust to arrhythmias. Finally, the sequence achieves "two-for-one" efficiency, providing both $T_1$ and $T_2$ maps in the time typically required for a single $T_1$ map.
Limitations and Future Work
While promising, this study is limited by its small sample size of healthy volunteers. Future research should evaluate the sequence in patients with myocardial pathologies, such as infarction or myocarditis, to assess performance in diseased tissue. Additionally, further optimization of the $T_{sat}$ and $TE_{prep}$ combinations could enhance the signal-to-noise ratio (SNR) and overall map quality.
Summary updated