1345151 - Characterization of Quantitative Susceptibility Mapping (QSM) in the Heart

Wednesday, January 25, 2023

9:10 AM – 9:20 AM

Location: Silver Pearl 1

Presenting Author(s)

AT

Andrew Tyler, PhD

Research Associate
King's College London, United Kingdom

Primary Author(s)

AT

Andrew Tyler, PhD

Research Associate
King's College London, United Kingdom

Co-Author(s)

LH

Li Huang, PhD

Research Associate
King's College London
London, England, United Kingdom
RN

Radhouene Neji, PhD

Siemens Research Scientist
King's College London, United Kingdom
RM

Ronald Mooiweer, PhD

MRI Scientist & Research Associate
Siemens Healthcare & King's College London, United Kingdom
Pier-Giorgio Masci, MD, PhD

Consultant Cardiologist
|School of Biomedical Engineering & Imaging Sciences Faculty of Life Sciences & Medicine | King’s College London
London, United Kingdom
SR

Sébastien Roujol, PhD

Reader in Medical Imaging
King's College London
London, England, United Kingdom

Disclosure(s):

Pier-Giorgio Masci, MD, PhD: No relevant disclosure to display

Sébastien Roujol, PhD: No financial relationships to disclose

Background: Cardiac quantitative susceptibility mapping (QSM) shows promise for the assessment of a number of cardiac conditions [1]. For example, QSM may offer better specificity to iron compared to T₂* imaging, which is influenced by a range of pathophysiological factors including edema, collagen, and fat deposition, potentially cancelling out the signal from iron deposits. In hemorrhagic infarcts cardiac QSM outperformed T2* imaging by identifying and quantifying the iron deposits in the core of infarcted tissue, where susceptibility was elevated by up to 1 ppm [2]. Despite the potential of cardiac QSM, characterization of this technique is currently lacking. In this study, we evaluate the accuracy, precision, and repeatability of cardiac QSM using phantom and healthy volunteer studies.

Methods:

All experiments were performed on a 1.5T scanner (MAGNETOM Aera, Siemens Healthcare, Erlangen, Germany). QSM was performed using a prototype free-breathing diaphragmatic respiratory navigated multi-echo 3D GRE sequence with 5 echoes (TE₁=3.2ms, ΔTE=2.9ms, FOV=288x384x100mm³, voxel size=1.5x1.5x5mm³, FA=15°, BW=543Hz/Px, segments=10, GRAPPA factor=2, partial Fourier=75%). In vivo scans were in short-axis orientation and gated to mid-diastolic cardiac phase.

QSM maps were reconstructed from magnitude and phase images using the MEDI toolbox [3] in Matlab (R2019b). Phase images were unwrapped with a region growing algorithm [4] and the local field was extracted with the projection onto dipole fields algorithm [5]. For the MEDI reconstruction, λ=1000 and zero-padding (to isotropic voxel size) were used.
The accuracy of this technique was evaluated in a phantom study consisting of 6 tubes containing solutions of gadobutrol with theoretical susceptibilities of (0, 0.2, 0.5, 1.0, 2.0, 3.0 ppm). In-vivo precision and repeatability was characterized in 10 healthy volunteers (5F/5M, Age=31±7, BMI=25±4). Each subject was scanned twice with a break outside of the scan room. Precision (spatial standard deviation) and repeatability (signed difference of both repeats) were measured using an AHA-16 segment model.

Results:

In phantom (Figure 1), measured susceptibility had a strong linear dependence on gadolinium concentration (R²=0.99) with significant streaking artefacts only appearing above the expected range of physiological susceptibility shifts [2].
In-vivo results are summarized in Figure 2 and 3. Across all segments the precision and repeatability were less (better) than the expected effect size for hemorrhagic infarct (0.5-1.0ppm[2]). Both metrics however, were affected by field inhomogeneities with segments near the heart-lung-liver interface (segments #5 and #11) showing degraded precision and repeatability.

Conclusion: Cardiac QSM is feasible and has an excellent accuracy in phantom study. In-vivo precision and repeatability measures are promising despite a decline in performance in specific cardiac regions due to field inhomogeneity. Further clinical evaluation of the technique is now warranted.