Non-selective double inversion recovery pre-pulse for flow-independent black blood myocardial viability imaging

Introduction

MRI late gadolinium enhancement (LGE) using the inversion-recovery (IR) sequence is the current gold standard for assessing myocardial viability. Although it achieves high contrast between infarct and normal myocardium, there is often poor infarct-to-blood contrast. Flow-dependent and diffusion-prepared black-blood LGE techniques have previously been described.[1, 2] Recently a quadruple-inversion recovery pre-pulse was introduced for T₁-independent flow suppression in carotid plaque imaging[3]. We introduced a modification to this pre-pulse aiming to achieve flow-independent signal suppression over a wide user-defined T₁-range and to improve sub-endocardial infarct detection in LGE myocardial viability imaging.

Methods

NS-DIR pre-pulse

A non-selective double-inversion recovery (NS-DIR) sequence with two time delays, TI₁ and TI_2, was implemented on a 3 T Philips Achieva MR-scanner (Philips-Healthcare, Best, NL). TI₁ and TI₂ were optimized in MATLAB simulations by minimizing M _Z ^NS-DIR over several user-defined T₁-ranges for a given heart rate.

Phantom experiments

A T₁-phantom containing 11 T₁-samples (T₁-range = 120 ms-1730 ms) was imaged with the NS-DIR pre-pulse using optimized TI₁ and TI₂ times. The signal-to-noise ratio (SNR) was calculated for each sample.

Patient Study

A 78-year-old man with previous myocardial infarctions was imaged with a 32-channel coil ~15 minutes after injection of 0.12 mmol/kg Gd-DOTA (Gadovist). Firstly a breath-hold 2D IR segmented gradient-echo (TFE) sequence was acquired in standard views. Imaging parameters included: spatial-resolution = 1.54 × 1.75 × 8 mm, TR/TE = 3.8 ms/2 ms, FA = 25°, TFE-factor = 25 and TI = 350 ms(chosen using LookLocker sequence).

Subsequently, identical planes were repeated with the IR replaced by the NS-DIR pre-pulse with imaging parameters maintained. TI₁ = 411 ms and TI₂ = 156 ms were used (optimized to minimize M _Z ^NS-DIR for T₁-range = 300-1400 ms, heart rate = 70 bpm).

Results

Simulations & Phantom experiments

M _Z ^NS-DIR simulations (Fig. 1a) indicate excellent signal suppression over the desired T₁-range for all heart rates with corresponding phantom studies in good agreement (Fig. 1b).

Figure 1

a) Simulated M _z ^NS-DIR curves for TI ₁ and TI ₂ values optimized to minime M _z ^NS-DIR for T ₁ values between 300 and 1400 ms for difference heart rates. Figure 1b) The corresponding SNR values measured in phantom images using the same TI₁ and TI₂ settings and heart-rates are in good agreement with the simulations.

Patient Study

NS-DIR images demonstrate excellent signal suppression of blood and normal myocardium (Fig. 2a) while conventional IR-TFE images (Fig. 2b) display similar infarct and blood signal. Whilst both techniques demonstrate transmural anterior and inferior wall infarcts, the NS-DIR image depicts an apical, non-transmural sub-endocardial defect, which is difficult to distinguish from blood in the IR image.

Figure 2

A 78-year-old man with previous myocardial infarctions was imaged using a) the NS-DIR pre-pulse and b) the standard IR sequence. Arrows indicate transmural infarcts in the anterior and inferior walls and a non-transmural apical infarct which is better visualized with the NS-DIR pre-pulse.

Conclusion

We have developed a new flow-independent LGE sequence for improved contrast visualization. Simulations and phantom studies demonstrate excellent tissue suppression over a wide T₁-range. Preliminary patient data suggests improved visualization of small sub-endocardial defects. Further studies are warranted to investigate the clinical usefulness of this novel approach.