A segmented modified look-locker inversion recovery (MOLLI) sequence for high heart rate T1 mapping of mice

Maryam Nezafat, Markus Henningsson, Christian Stehning, Mehmet Akçakaya, Andrea Protti, René BotnarJournal of Cardiovascular Magnetic Resonance 2015 · DOI

Abstract

Quantitative T1 mapping provides myocardial tissue characterization for assessment of various cardiomyopathies. The Modified Look-locker (MOLLI) sequence is widely used for mapping the T1 quantification, where multiple single-shot images are acquired along the T1 recovery curve after an inversion pulse. However, single-shot imaging becomes infeasible for mouse imaging at high heart rates due to motion artifacts and requirements of resolution/coverage. Additionally, typical MOLLI sampling schemes (3-3-5) and the pauses between blocks have to be adapted to the high heart rates in mice. In this work, we propose a segmented acquisition scheme for T1 mapping of mouse at high heart rates. After an initial inversion pulse, we acquire segments for 20 images in subsequent heartbeats followed by 20 pause heartbeats to allow for full magnetization recovery. The complete k-space is acquired in this fashion over 5 segments per image. Experiments were performed with a T1 phantom by simulating high heart rates to evaluate the accuracy of the proposed sequence. Proof of concept T1 maps were also acquired in one healthy mouse.

Introduction

T1 mapping is a vital MRI technique used to characterize myocardial tissue, helping clinicians and researchers identify various cardiomyopathies. The "gold standard" for this is the MOLLI sequence. However, MOLLI was originally designed for humans, whose heart rates typically range from 60 to 100 bpm.

Mice are essential models for cardiovascular research, but their hearts beat incredibly fast—often between 400 and 600 bpm. At these speeds, standard "single-shot" imaging (where an entire image is captured in one go) fails. It results in severe motion artifacts and lacks the spatial resolution needed to visualize small mouse hearts clearly. Furthermore, the standard timing of pulses and pauses in MOLLI doesn't allow enough time for the mouse's heart magnetization to recover between measurements.

Methods

The researchers developed a segmented acquisition scheme to overcome the limitations of single-shot imaging at high heart rates. Instead of capturing a full image in a single heartbeat, the "k-space" (the raw data grid of an image) is divided into five segments.

Figure 1: Diagram of the segmented ECG trigger MOLLI sequence showing the 20-acquisition and 20-pause heartbeat cycle.

The Pulse Sequence Workflow: 1. Inversion Pulse: A magnetic pulse "flips" the magnetization to initiate the T1 recovery process. 2. Acquisition Phase: The scanner acquires segments for 20 images over 20 subsequent heartbeats. 3. Recovery Phase: A pause of 20 heartbeats follows to allow the magnetization to fully recover before the next cycle. 4. Repetition: This process repeats until all 5 segments for each of the 20 images are collected.

The study utilized a 1.5T Philips Achieva scanner with a 32-element cardiac coil. The sequence was validated using a T1 phantom (14 vials with T1 values from 200ms to 1500ms) and simulated ECG heart rates of 60, 100, 250, 300, and 400 bpm.

Results and Key Insights

By spreading data collection over multiple heartbeats, the researchers achieved a higher in-plane resolution ($2 \times 2 \text{ mm}^2$) without the blurring caused by rapid cardiac motion. The choice of 20 acquisitions followed by 20 pauses was specifically tuned for high heart rates, ensuring sufficient "dead time" for longitudinal magnetization to reset, which is critical for accurate T1 calculation.

Figure 2: Proof-of-concept T1 map acquired in vivo from a healthy mouse heart at 310 bpm.

Key Findings: * Accuracy: At a simulated heart rate of 300 bpm, the difference between the proposed segmented MOLLI and the reference spin-echo was statistically non-significant ($p=0.3$). * Error Margins: For short T1 values, the relative difference was only 0.5–4.1%. For longer T1 values, the error increased slightly to 7.1–10%. * In Vivo Success: The team successfully generated a T1 map of a living mouse heart at 310 bpm. The measured T1 of the mouse myocardium was $792.9 \pm 95.1 \text{ ms}$, demonstrating the sequence's viability in a biological environment.

Discussion and Limitations

The primary limitation is the underestimation of higher T1 values, a known systematic bias in MOLLI-based sequences that becomes more pronounced at higher heart rates. As a proof-of-concept study, the in vivo testing was limited to a single healthy mouse.

Future work will involve larger cohorts, including mice with induced heart conditions, to evaluate the sequence's sensitivity in detecting diseased tissue. Additionally, there is potential to further optimize the number of segments and pauses to reduce total scan time while maintaining high accuracy.

Summary updated

Abstract

Background Quantitative T1 mapping provides myocardial tissue characterization for assessment of various cardiomyopathies. The Modified Look-locker (MOLLI) sequence is widely used for mapping the T1quantification, where multiple single-shot images are acquired along the T1 recovery curve after an inversion pulse. However, single-shot imaging becomes infeasible for mouse imaging at high heart rates due to motion artifacts and requirements of resolution/coverage. Additionally, typical MOLLI sampling schemes [1] (3-3-5) and the pauses between blocks have to be adapted to the high heart rates in mice. In this work, we propose a segmented acquisition scheme for T1 mapping of mouse at high heart rates. After an initial inversion pulse we acquire segments for 20 images in subsequent heartbeats followed by 20 pause heartbeats to allow for full magnetization recovery. The complete k-space is acquired in this fashion over 5 segments per image. Experiments were performed with a T1 phantom by simulating high heart rates to evaluate the accuracy of the proposed sequence. Proof of concept T1 maps were also acquired in one healthy mouse. Methods
BibTeX
@article{nezafat2015segmented,
  author = {Maryam Nezafat and Markus Henningsson and Christian Stehning and Mehmet Akçakaya and Andrea Protti and René Botnar},
  title = {{A segmented modified look-locker inversion recovery (MOLLI) sequence for high heart rate T1 mapping of mice}},
  journal = {Journal of Cardiovascular Magnetic Resonance},
  year = {2015},
  doi = {10.1186/1532-429x-17-s1-p120},
}
Figure