Multi-frequency reconstruction for frequency-modulated stack-of-stars bSSFP
Anne Slawig1, Tobias Wech1, Valentin Ratz1, Johannes Tran-Gia1,2, Henning Neubauer1, Thorsten Bley1, and Herbert Köstler1

1Departement for Diagnostic and Interventional Radiology, University of Würzburg, Würzburg, Germany, 2Department of Nuclear Medicine, Würzburg, Germany

Synopsis

Banding artefacts in images acquired by bSSFP are a big challenge in fast MRI as they can considerably reduce image quality and deteriorate the diagnostic value. As the steady state tolerates small shifts in frequency, it is possible to acquire a frequency-modulated bSSFP. Unfortunately a simple gridding reconstruction of such a measurement suffers from signal loss. Our study proposes a multi-frequency reconstruction and demonstrates its capability of reconstructing banding-free 3D images while retaining the high signal levels of standard bSSFP.

Purpose

Balanced steady state free precession (bSSFP) sequences are frequently used in fast MR imaging due to short scan times, high signal yield and excellent contrast. Unfortunately, these advantages are often outweighed by serious banding artefacts, which can considerably reduce image quality. In clinical routine, this is typically resolved by acquiring multiple images using different offset-frequencies1. In 2002, Foxall et al. demonstrated that the steady state of bSSFP does tolerate a slow change in frequency2. Such an acquisition with dynamic frequency-sweep proved suitable to create banding-free images even in the presence of high field inhomogeneities3. The purpose of this study was to present an optimized reconstruction algorithm for a frequency-modulated stack-of-stars bSSFP acquisition and to apply it to in vivo imaging of the inner ear.

Materials and Methods

The bSSFP measurements of a healthy volunteer were performed on a 3T MR system (Siemens MAGNETOM Prisma) using a 32-channel head coil. The acquisition was performed with and without frequency-modulation, where the latter featured a shift in the offset frequency for each projection, covering one cycle in the off-resonance profile off bSSFP. All datasets were acquired using a stack of stars trajectory with 64 partitions, each comprising 680 projections. Other imaging parameters were: TE 2.6ms TR 5.2ms, flip angle = 50°, resolution 0.6x0.6x0.6 mm3, total acquisition time: 5min 9s . All images were reconstructed using gridding, a 3D Fourier transform and a square-root of the sum-of-squares combination of all coils. Additionally, all acquisitions featuring the shift in frequency were reconstructed for 30 different off-resonance frequencies. The procedure first compensates for the linear drift in signal phase caused by the acquisition pattern and leads to a binary distribution of phase with one 180° phase shift per cycle of the off-resonance profile. The algorithm then eliminates this phase jump for different assumed frequencies, and therefore positions of the jump, causing all measured lines to combine constructively. Reconstructing all assumed off-resonances generates an image stack, whose maximum intensity projection provided the final result.

Results

Measurements show severe banding artefacts for standard bSSFP imaging. All frequency-modulated acquisitions removed bandings successfully, but suffered from a loss in signal intensity with respect to the standard reconstruction. The proposed multi-frequency reconstruction allows retaining the high signal levels of standard bSSFP while still removing banding artefacts.

Discussion

In our study we were able to demonstrate that a multi-frequency reconstruction is capable of reconstructing banding-free 3D images of the inner ear. Additionally it provides an increased signal level compared to a simple gridding reconstruction. Main advantages of bSSFP, namely acquisition speed and image contrast, could be preserved by the multi-frequency reconstruction of frequency-modulated stack-of-stars bSSFP.

Acknowledgements

No acknowledgement found.

References

1. Bangerter et.al., Analysis of multiple-acquisition SSFP. MRM. 2004;51(5): 1038-47

2. Foxall D.L. Frequency-Modulated Steady-State Free Precession Imaging. MRM. 2002;48:502-508

3. Benkert et.al. Fast isotropic banding-free bSSFP imaging using 3D dynamically phase-cycled radial bSSFP (3D DYPR-SSFP). Z Med Phys. 2015

4. Man LC1, Pauly JM, Macovski A. Multifrequency interpolation for fast off-resonance correction. MRM. 1997;37(5):785-92.

Figures

Off-resonance profile of frequency modulated bSSFP, a) amplitude, b) phase c) phase after adding linear term, d) phase for one step in the multi-frequency reconstruction showing the optimal case where the assumed off-resonance fits the true off-resonance of this pixel

left: standard bSSFP acquisition showing banding artefacts (red arrows), right: multifrequency-reconstruction, inlay: reformatted view of the cochlea



Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
0517