Real-time cardiovascular magnetic resonance at high temporal resolution: radial FLASH with nonlinear inverse reconstruction

Abstract

Background: Functional assessments of the heart by dynamic cardiovascular magnetic resonance (CMR) commonly rely on (i) electrocardiographic (ECG) gating yielding pseudo real-time cine representations, (ii) balanced gradient-echo sequences referred to as steady-state free precession (SSFP), and (iii) breath holding or respiratory gating. Problems may therefore be due to the need for a robust ECG signal, the occurrence of arrhythmia and beat to beat variations, technical instabilities (e.g., SSFP "banding" artefacts), and limited patient compliance and comfort. Here we describe a new approach providing true real-time CMR with image acquisition times as short as 20 to 30 ms or rates of 30 to 50 frames per second.

Methods: The approach relies on a previously developed real-time MR method, which combines a strongly undersampled radial FLASH CMR sequence with image reconstruction by regularized nonlinear inversion. While iterative reconstructions are currently performed offline due to limited computer speed, online monitoring during scanning is accomplished using gridding reconstructions with a sliding window at the same frame rate but with lower image quality.

Results: Scans of healthy young subjects were performed at 3 T without ECG gating and during free breathing. The resulting images yield T1 contrast (depending on flip angle) with an opposed-phase or in-phase condition for water and fat signals (depending on echo time). They completely avoid (i) susceptibility-induced artefacts due to the very short echo times, (ii) radiofrequency power limitations due to excitations with flip angles of 10 degrees or less, and (iii) the risk of peripheral nerve stimulation due to the use of normal gradient switching modes. For a section thickness of 8 mm, real-time images offer a spatial resolution and total acquisition time of 1.5 mm at 30 ms and 2.0 mm at 22 ms, respectively.

Conclusions: Though awaiting thorough clinical evaluation, this work describes a robust and flexible acquisition and reconstruction technique for real-time CMR at the ultimate limit of this technology.

Figures

Figure 1

Multislice localizer images using radial FLASH. The serial T1-weighted images (spoiled FLASH, TR/TE = 2.2/1.4 ms, flip angle 10°) at 1.5 mm resolution, 8 mm section thickness (312 mm coronal field-of-view, free breathing, no ECG or respiratory gating), and 1000 ms acquisition time (455 spokes, 208 data samples, gridding reconstruction) demonstrate the motion robustness of radial CMR sequences and serve to plan functional assessments of the heart.

Figure 2

T1 contrast in real-time radial FLASH CMR. T1-weighted images (spoiled FLASH, TR/TE = 2.0/1.3 ms, 15 spokes, acquisition time 30 ms) with increasing T1 contrast as a function of flip angle (4°, 8°, 12°) for a short-axis view (2.0 mm resolution, 8 mm section thickness) during (a) diastole and (b) systole. The images were selected from respective real-time CMR movies reconstructed by nonlinear inversion.

Figure 3

Water and fat signals in real-time radial FLASH CMR. Opposed-phase (spoiled FLASH, TR/TE = 2.0/1.3 ms, flip angle 8°, 15 spokes, acquisition time 30 ms) and in-phase images (TR/TE = 3.2/2.5 ms, acquisition time 48 ms) for (top) a short-axis and (bottom) a 4-chamber view (2.0 mm resolution, 8 mm section thickness). The images were selected from respective real-time CMR movies reconstructed by nonlinear inversion.

Figure 4

Simultaneous opposed-phase and in-phase real-time radial FLASH CMR. T1-weighted images (spoiled multi-echo FLASH, TR = 3.1 ms, flip angle 8°, 11 spokes) of a short-axis view with (a) an opposed-phase (TE = 1.3 ms) and (b) in-phase condition (TE = 2.4 ms) for overlapping water and fat signals (2.0 mm resolution, 8 mm section thickness, 34 ms total acquisition time). The images cover a post-systolic expansion phase by every second frame (68 ms apart). They were selected from respective real-time CMR movies reconstructed by nonlinear inversion and zoomed by a factor of 1.5.

Figure 5

Spatial vs temporal resolution in real-time radial FLASH CMR. T1-weighted images (spoiled FLASH, flip angle 8°) of a short-axis view during systole at 2.0 mm (TR/TE = 2.0/1.3 ms), 1.5 mm (TR/TE = 2.2/1.4 ms), and 1.0 mm resolution (TR/TE = 2.8/1.8 ms) using data sets with only 19, 15, and 11 spokes, respectively. Individual acquisition times are given by the # spokes × TR and range from 22 ms (upper right: 11 spokes and TR = 2.0 ms) to 52 ms (lower left: 19 spokes and TR = 2.8 ms). The images were selected from respective real-time CMR movies reconstructed by nonlinear inversion and zoomed by a factor of 1.5.

Figure 6

Spatial vs temporal resolution in real-time radial FLASH CMR. Same as Fig. 4 but for a diastolic phase.

Figure 7

Real-time radial FLASH CMR at 1.5 mm spatial resolution and 33 ms temporal resolution. Consecutive T1-weighted images (spoiled FLASH, TR/TE = 2.2/1.4 ms, flip angle 8°, 15 spokes) of a short-axis view at 1.5 mm resolution, 8 mm section thickness, and 33 ms acquisition time. The 20 images cover a 660 ms period of a single cardiac cycle from systole (upper left) to diastole (lower right). They were selected from a respective real-time CMR movie reconstructed by nonlinear inversion (see Additional File 1) and zoomed by a factor of 1.5.

Figure 8

Interleaved dual-slice real-time radial FLASH CMR. T1-weighted images (spoiled multi-slice FLASH, TR = 4.0 ms for two sections, TE = 2.0 ms, flip angle 10°, 11 spokes, 44 ms total acquisition time) of two simultaneously acquired short-axis views (a) in the apical segment and (b) 20 mm above in the mid-cavity segment (2.0 mm resolution, 8 mm section thickness). The images cover a post-systolic expansion phase by every second frame (88 ms apart). They were selected from respective real-time CMR movies reconstructed by nonlinear inversion and zoomed by a factor of 1.5.

Figure 9

Real-time radial FLASH CMR at 2.0 mm spatial resolution and 22 ms temporal resolution. Consecutive T1-weighted images (spoiled FLASH, TR/TE = 2.0/1.3 ms, flip angle 8°, 11 spokes) of a 2-chamber view at 2.0 mm resolution, 8 mm section thickness, and 22 ms acquisition time. The images depict two 176 ms periods (8 images) of a single cardiac cycle that are 682 ms (31 images) apart: (a) opening of the mitral valve and resulting blood stream into the left ventricle, (b) turbulent flow leading to a vortex at the apex of the left ventricle. They were selected from a respective real-time CMR movie reconstructed by nonlinear inversion (see Additional File 2) and zoomed by a factor of 1.5.