1964

Effect of inflow and in-plane saturation in SASHA and MOLLI T1 and T1* maps in a perfusion phantom and in-vivo
Ingo Hermann1, Tanja Uhrig1, Jorge Chacon-Caldera1, Mehmet Akçakaya2,3, Lothar R. Schad1, and Sebastian Weingärtner1,2,3

1Computer Assisted Clinical Medicine, University Medical Center Mannheim, Heidelberg University, Mannheim, Germany, 2Electrical and Computer Engineering, University of Minnesota, Minneapolis, MN, United States, 3Center for Magnetic Resonance Research, University of Minnesota, Minneapolis, MN, United States

Synopsis

Aim of this work was to evaluate the e ffect of flow on blood T1 measurement considering inflow of non-inverted spins. Experiments using a flow phantom demonstrate shorter T1 for increasing flow velocity, with high reproducibility. In-vivo measurements show major variations throughout the cardiac cycle, validating the flow sensitivity observed in phantom measurements.

Introduction:

Blood T1 times are widely used for myocardial tissue characterization, as part of ECV mapping[1] and synthetic Hct[2]. However, conventional myocardial T1 mapping methods are known to be susceptible to flow eff ects, due to 1) in-flow of non-inverted spins[3] and 2) partial saturation of spins caused by previous imaging pulses[4]. In this work we study flow eff ects on commonly used myocardial T1 mapping techniques, using a perfusion phantom and velocity controlled T1 measurements in the aorta.

Methods:

A perfusion phantom with adjustable flow velocities was used for the flow dependent T1 measurement of gadolinium doped water ( Fig. 1a). T1 and T1* measurements were performed using 5(3s)3 MOLLI[5] for bSSFP and GRE readouts (MOLLI / MOLLIGRE) with and without Look-Locker correction (MOLLI nLL / MOLLIGRE nLL) and SASHA with 2 and 3 parameter fit (SASHA 2P / SASHA). The common sequence parameters were FOV = 240 x 240 mm2, matrix size = 192 x 192, slice thickness = 8 mm, GRAPPA factor = 2, partial fourier = 6/8. For bSSFP readout TR/TE/ α = 3.6 ms/1.8 ms/60° and for GRE readout TR/TE/ α = 2.9 ms/1.7 ms/8° respectively. Inversion recovery turbo spin-echo measurements were performed as reference T1 and T2 quanti fication in the absence of flow. All sequences were acquired in anterior and posterior slice in the dialysis fi lter, yielding di fferent amount of non-inverted spin in-flow despite identical through-plane flow velocities. Additionally, in-vivo T1 and T1* measurements for bSSFP (α = 45°) and GRE (α = 8°) readout were performed in six healthy subjects (26 ± 5 years, 66.6% male) in the descending aorta. Velocity encoded CINE measurements were performed to quantify the flow velocity in the dialysis fi lter and throughout the cardiac cycle. During systole more T1 maps were acquired due to quicker change in flow velocity compared with the diastole.

Results:

T1 times are decreased by up to 200 ms in the presence of flow in posterior slice in the dialysis fi lter for positive flow direction (water reservoir to dialysis filter). The same trend is observed for the anterior slice for opposite flow direction ( Fig. 1b/c). Flow-effects were substantially decreased for reversed flow direction. T1* maps and SASHA 2P showed a smaller increase in T1 and resulted in consistently shorter T1 times especially in the absence of flow. SASHA 2P and MOLLI without Look-Locker correction, suffered from increased T1 times for slow flow velocities, irrespective of the flow direction. However, this was alleviated using Look-Locker correction or a 3 parameter model respectively. Reference probe measurements showed good reproducibility of T1. In-vivo measurements confirmed the flow dependency of T1. ( Fig. 2a). In the absence of flow the longest aortic T1 times were observed comparable to T1 times in the left ventricle. T1 maps lacked in precision and homogeneity as seen in Fig. 3a resulting in standard deviations up to 500 ms for peak flow velocities. Difference in peak flow and no flow T1 times around 100-200 ms in T1 maps and 300-600 ms in T1* maps are depict in Table 1.

Discussion:

T1 times are decreased by increasing flow velocity due to inflow of non-inverted spins in positive flow direction in anterior slice and in negative flow direction in posterior slice. Flow sensitivity irrespective of the flow direction was only observed for slow velocities and when using a two-parameter model, or MOLLI without Look-Locker correction. This and the reduced flow sensitivity for reversed flow direction indicates that the effect of partially saturated spins by imaging pulses is largely compensated for using the Look-Locker correction.

In-vivo measurements confirmed the effect of decreasing T1 due to non-inverted inflowing spin at peak velocities. In 2 subjects in-plane saturation effect is measured in the absence of flow for T1 maps. T1* lacked in precision and high standard deviations due to fast and non homogeneous flow in the aorta compared to the dialysis filter.

Conclusion:

Blood T1 times measured with myocardial T1 mapping sequences display major variation due to inflow of non-inverted spins. Flow-e ffects caused by partial saturation of spins in the imaging slice, seem to be compensated for using a Look-Locker correction or a 3 Parameter model for MOLLI and SASHA, respectively. These results demonstrate that the use of blood T1 times measured with conventional myocardial T1 mapping methods, might be confounded by factors varying the blood circulation, including the subjects stroke volume, total blood volume and size.

Acknowledgements

No acknowledgement found.

References

[1] Kellman, P., Wilson, J. R., Xue, H., Ugander, M. & Arai, A. E. Extracellular volume fraction mapping in the myocardium, part 1: evaluation of an automated method. Journal of Cardiovascular Magnetic Resonance 14, 63 (2012).

[2] Treibel, T. A. et al. Automatic Measurement of the Myocardial Interstitium: Synthetic Extracellular Volume Quanti cation Without Hematocrit Sampling. JACC: Cardiovascular Imaging 9, 54-63 (2016).

[3] Kellman, P. & Hansen, M. S. T1-mapping in the heart: accuracy and precision. Journal of Cardio-vascular Magnetic Resonance 16, 2 (2014).

[4] Kelvin, C. et al. Saturation recovery single-shot acquisition (SASHA) for myocardial T1 mapping. Magnetic Resonance in Medicine 71, 2082-2095 (2014).

[5] Messroghli, D. R. et al. Modi ed Look-Locker inversion recovery (MOLLI) for high-resolution T1 mapping of the heart. Magnetic Resonance in Medicine 52, 141-146 (2004).

Figures

Table 1: Aortic blood T1 times for MOLLI and MOLLIGRE with and without LL correction for two healthy subjects. T1 times for the time points in the cardiac cycle with no and maximum flow velocities are provided.

Figure 1: (a) Perfusion phantom. A peristaltic pump (Watson-Marlow-Bredel) circulates gadolinium doped water from a reservoir to the dialysis filter through a pipe outside the bore and back to the reservoir (positive flow-direction) in the opposite direction (negative flow-direction). Two slices are placed in the dialysis fi lter (flow compartment). The posterior slice additionally comprises a reference tube. (b,c,d) T1 times measured in the the dialysis filter and probe for various pump speeds. MOLLI T1 and T1* (orange crosses/dots), MOLLIGRE T1 and T1* (yellow crosses/dots), SASHA with 3 and 2 parameter fit (blue crosses/dots) and reference measurements (black) are plotted against the flow velocity.

Figure 2: (a) Comparison of blood T1 values in the aorta measured at various time points of the cardiac cycle in one healthy subject. Flow velocity measured in a separate VENC scan is depicted in blue. MOLLI T1 and T1* (violet/bright violet) and T1 and T1* (orange/yellow) are depicted. Substantially reduced T1 times are observed in the presence of flow compared to the time point of no flow. (b) Example T1 map acquired in aortic view and (c) example VENC baseline image of the corresponding slice location.

Figure 3: Example T1 and T1* maps for GRE and bSSFP readout in the absence of flow and for peak flow velocities in a healthy subject.

Proc. Intl. Soc. Mag. Reson. Med. 27 (2019)
1964