Computationally Efficient Combination of Multi-channel Phase Data From Multi-echo Acquisitions (ASPIRE)

5 years ago

Computationally Efficient Combination of Multi-channel Phase Data From Multi-echo Acquisitions (ASPIRE)

Siegfried Trattnig, Korbinian Eckstein, Karin Poljanc, Beata Bachrata, Wolfgang Bogner, Barbara Dymerska, Simon Daniel Robinson

Purpose To develop a simple method for combining multi-echo phase information from a number of coils in an array that requires no volume coil or additional scans and yields signal-to-noise ratio-optimal images that reflect only ΔB0-related phase. Theory and Methods Two SNR optimal coil combination methods were developed which retrieve the ΔB0-related phase by determining the coil-dependent phase offsets. The first variant, MCPC-3D-S, requires the unwrapping of one phase image; the second variant, ASPIRE, allows unwrapping to be avoided if two echoes j and k satisfy the echo time relation m⋅TEk=(m+1)⋅TEj, where m is an integer, making this a particularly fast and robust approach. Both developed methods constitute improvements over a prior method, MCPC-3D, in terms of robustness and computational expense. Results In the brain at 7 T, phase matching and contrast-to-noise ratio were higher with MCPC-3D-S and ASPIRE than with phase difference reconstruction, and similar to the reference coil–dependent Roemer combination. Unlike the Roemer and virtual reference coil methods, the proposed approaches also eliminated all non– ΔB0-related phase. Conclusion MCPC-3D-S is an improvement over prior multi-echo methods, which is useful if the ASPIRE echo time condition cannot be fulfilled. ASPIRE is a particularly fast and robust approach that runs on the scanner's reconstructor in a small fraction of the acquisition time. Magn Reson Med, 2017. © 2017 International Society for Magnetic Resonance in Medicine.

Publisher URL: http://onlinelibrary.wiley.com/resolve/doi

DOI: 10.1002/mrm.26963