Synopsis

Fat-suppressed balanced steady-state free precession (bSSFP) is a rapid imaging sequence often used in cardiovascular MRI. Recently a fast interrupted steady-state (FISS) sequence was proposed which periodically interrupts the steady-state of the bSSFP. The resulting frequency response modulation can be leveraged for suppression of the off-resonant fat signal without the need of fat preparation pulses. FISS was demonstrated for 2D radial acquisitions, however challenges to apply this approach to 3D Cartesian have been reported. Here we propose to extend FISS to 3D Cartesian imaging and investigate its behavior by extended phase graph simulations and in-vivo leg, abdominal measurements.

Introduction

Methods

The pulse sequence diagram of FISS² is depicted in Fig.1. A FISS module is composed of a [+α/2,(-α,readout,+α)_n,-α/2] train where α denotes the RF excitation and n is the number of bSSFP readouts surrounded by the ±α/2 ramp-up/down pulses which cause the steady-state interruption. bSSFP readouts are repeated with a repetition time of TRb. In order to extend FISS to 3D Cartesian imaging here we propose to also consider: a) RF phase cycling, b) gradient spoiling and c) dummy preparation pulses. Between FISS modules the RF phase is incremented according to

$$\varphi_j = \varphi_{j-1}+j\cdot \varphi_0$$

yielding the phase of the j-th FISS module with a golden angle $$$\varphi_0=111.3^\circ$$$ initialization. Together with phase-encoding gradient spoiling, i.e. unbalanced gradients, this ensures sufficient spoiling of residual transverse magnetization before the next excitation. Dummy preparation pulses at the beginning of the sequence bring the magnetization to bSSFP steady-state before it is interrupted. Otherwise, it can occur that off-resonant transient magnetization is excited and even amplified instead of suppressed. This effect is in the Cartesian case more pronounced in the final image, because of the different phase-encoding steps per readout.

FISS signal evolution and frequency behavior were simulated using extended phase graphs⁴ (EPG) for fat (250/70 ms), liver (500/40 ms), myocardium (1000/45 ms) and blood (1600/250 ms) with respective T1/T2 values.

In-vivo data was acquired on a 1.5T MR scanner (Aera, Siemens Healthineers) with imaging parameters as defined in Tab. 1. Images were obtained in 5 healthy subjects (2 female, age = 29±3 years) for upper thigh and abdomen.

Results and Discussion

An exemplary comparison of FISS versus bSSFP is depicted in Fig. 1b) where the spectral frequency response indicates the fat suppression capability. At around 220 Hz (3.5ppm at 1.5T) a broad stop band provides the desired signal suppression for different tissues. The frequency response pattern changes for varying FISS interruption rates which is mainly determined by the timing parameters TRb and n. For fast acquisitions a minimal TRb is desired with maximal bandwidth, hence, the timing is primarily determined by the imaging resolution and therefore FISS behavior changes with imaging resolution and the number of bSSFP readouts n. These dependencies are investigated theoretically via EPG simulations in Fig. 2 and substantiated by measurements (Fig.3 and 4). The reliable operating range of FISS is determined by n and imaging resolution as depicted in Fig. 2 and tends to smaller values n for higher resolutions.

Images for different length n of FISS in comparison to bSSFP are shown for upper thigh in Fig. 3 and for abdomen in Fig. 4. Fat suppression is excellent and consistent over the whole volume and amongst subjects with varying body mass index with FISS while absent in bSSFP (without fat-saturation pulses). Contrast between muscle and vessel is improved (see MIP). Small vessel structures can be observed in FISS and are lost in bSSFP. FISS fat suppression is susceptible to B0 inhomogeneities as observed in the abdominal imaging (Fig.4), but is still able to provide good fat suppression.

Conclusion

Acknowledgements

References