Rapid cardiac T1 mapping within two heartbeats

Introduction

Late gadolinium enhancement (LGE) imaging is an important CMR method than can detect salvageable myocardium after myocardial infarction [1-2]. Recently, T₂-weighted-imaging has gained a significant interest to assess myocardial edema [3]. However, clinical interpretation of T₂-weighted-imaging could be hindered by surface coil effects which yield non-uniform signals. Multi-point T₁ mapping approaches, such as Modified Look-Locker inversion recovery (MOLLI) [4], have been proposed to measure myocardial T₁, but, as a multiple heartbeat acquisition, it may be sensitive to cardiac motion and arrhythmia. We propose to develop a 2-second cardiac T₁ mapping pulse sequence for assessment of myocardial edema (pre-contrast) and infarction (post contrast) in patients with acute myocardial infarction.

Purpose

To develop and validate a cardiac T₁-mapping technique.

Methods

The proposed T₁-mapping acquisition consists of 2 TurboFLASH images with centric k-space ordering: proton density-weighted (PDw) image in the first heartbeat and saturation recovery (SR) T₁w acquisition in the second heartbeat. A robust non-selective saturation pulse [5] was used to achieve uniform saturation of magnetization. A long delay time=500ms was used to achieve adequate signal-to-noise ratio. The T₁w-image was normalized by the PDw image to correct for unknown equilibrium magnetization and receiver coil sensitivity. T₁ was calculated algebraically assuming an ideal saturation-recovery equation based on the Bloch equation [6]. Eight healthy volunteers (32±13y.o.) were imaged in a short-axis basal plane at 3T (Tim-Trio, Siemens) at baseline and 10 minutes following 0.05mmol/kg Gd-DTPA injection. All images were acquired in mid-diastole with appropriate trigger delay. Imaging parameters included: FOV=350mm×272mm, matrix=144×112, TE/TR=1.2/2.4ms, flip angle=10°, in-plane resolution=2.4mm×2.4mm, GRAPPA ~1.65, temporal resolution=162ms, and receiver bandwidth=990Hz/pix. For validation purposes, myocardial T₁ were compared to reference T₁ measurements using multi-point SR with TurboFLASH readout (~20s-breath-hold): 1 PDw-image, 12 T₁w-images with TD 100to600ms every 100ms, then 800to1800ms every 200ms. A nonlinear Levenberg-Marquardt algorithm was used to fit the normalized multi-point SR data. The proposed T₁-mapping method was also evaluated in a patient with arrhythmia, before and 20min after administrating 0.15mmol/kg Gd-DTPA.

Results

Myocardial T₁ measured using the proposed rapid method were linearly correlated with T₁ measured using the multi-point T₁ method (Fig.1, slope=0.99, bias=29ms, r=0.99, P<10^-5). Pre- and post-contrast T₁-maps obtained in a 52y.o.-volunteer and a 44y.o.-patient with arrhythmia are shown in Fig.2-3, respectively (same T₁-scale).

Figure 1

T₁ measured with the T₁-mapping method vs. multi-point SR T₁ measurements in the LV myocardium.

Figure 2

T₁-maps obtained in a volunteer before and 10 min after 0.05mmol/kg bolus injection.

Figure 3

T₁-maps obtained in a patient with arrhythmias before and 10 min after 0.15mmol/kg slow injection.

Conclusion

The proposed T₁-mapping method is a fast pixel-wise T₁-mapping technique with insensitivity to cardiac motion and arrhythmia. Future work includes evaluation in patients with acute and chronic infarction.