Introduction

High resolution 3D cardiac magnetic resonance (CMR) scar imaging is usually acquired following a bolus of gadolinium based contrast agent and based on the assumption that contrast concentration in blood pool and myocardial tissue does not change significantly during the CMR acquisition. This limits the duration over which imaging can be acquired and impacts the resolution of imaging that can be achieved. The use of an infusion to achieve contrast equilibrium has been previously reported in the calculation of extracellular volume (ECV) maps(1). We hypothesised that continuous contrast infusion could be used during late gadolinium enhancement (LGE) imaging to establish a stable myocardial T₁, thereby allowing extended acquisition for high-resolution 3D fibrosis imaging without significant change in inversion times (TI) for blood and myocardium.

Methods

All imaging was performed at 1.5T (MAGNETOM Aera, Siemens Healthcare, Germany). Eight pigs with chronic myocardial infarction were studied. Ten minutes after 0.1mmol/kg gadobutrol (Gd) bolus, TI scout was acquired in a single short axis slice. Clinical standard 2D-LGE scar sequences were acquired in 8 short axis slices using an inversion recovery (IR) sequence with phase-sensitive IR (PSIR) reconstruction (SSFP; linear k-space reordering; TE/TR/α: 1.21ms/3ms/45°; slice thickness: 6mm; in plane resolution: 1.4x1.4mm²). Fifteen minutes following Gd bolus, 0.0011mmol/kg/min Gd infusion was commenced. TI-scout was repeated at regular intervals. 30 minutes following Gd bolus, isotropic navigator-gated ECG-triggered 3D IR sequence was acquired (SSFP; coronal orientation; linear k-space reordering; TE/TR/α: 1.58ms/3.6ms/90°; gating window = 7mm; resolution 1.2x1.2x1.2mm³). Scar imaging was reviewed by 2 experienced observers who graded it on a 4 point scale according to image sharpness, image contrast, quality of myocardial nulling, complexity of scar pattern and overall scan quality. Signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) were estimated using mean signal within blood, remote (healthy) myocardium and scar and standard deviation of signal within lung.

Results

Contrast steady state (CSS, <5% variation in blood/myocardium TI) was achieved in 7 pigs and maintained for up to 150 minutes (figure 1). Mean acquisition for CSS 3D-LGE imaging was 51 minutes. CSS 3D-LGE imaging was acquired using a total Gd dose of <0.2mmol/kg. SNR_Scar, SNR_Blood and CNR_Scar-remote was higher in CSS 3D-LGE imaging than the 2D-LGE imaging (p<0.001) (figure 3). SNR_Remote was lower in the CSS 3D-LGE than the 2D-LGE imaging (p<0.05), indicating more effective nulling of the myocardium. There was a significant median increase in sharpness (4.0 vs 2.5, p=0.026), contrast (4.0 vs 3.0, p=0.024), appreciation of scar complexity (4.0 vs 2.5, p=0.038) and overall scan quality (4.0 vs 2.5, p=0.041) in the CSS 3D-LGE imaging compared to the post Gd bolus 2D imaging (figure 3).

Discussion

We have demonstrated that CSS imaging facilitates extended acquisition times without requiring contrast doses above those used in routine clinical practice(2). Despite the prolonged acquisition, SNR in scar, healthy myocardium and blood pool were significantly above that observed in clinical standard 2D LGE imaging, acquired in under 5 minutes. Higher SNR and CNR between scar and healthy tissue will facilitate accurate identification of scar. Higher signal in the blood pool reduced the contrast between blood pool and scar which may make accurate identification of the endocardium challenging in areas of transmural scar. Black blood scar imaging techniques may be helpful to overcome this issue. The maintenance of CSS for up to 150 minutes demonstrates that this technique may be used in experimental conditions, where total Gd dose is not restricted, to directly compare imaging sequences by allowing extended imaging of the same tissue under consistent contrast conditions.

Conclusion

CSS allows extended acquisition to facilitate high-resolution 3D LGE imaging with improved SNR, CNR and overall image quality. CSS-LGE imaging may be valuable when detailed myocardial substrate assessment is required, for example when planning cardiac electrophysiology procedures. In addition, CSS-LGE imaging provides an experimental model in which post-contrast sequences may be compared with consistent contrast distributions.

Acknowledgements

No acknowledgement found.