0216

FAMASITO - FASTMAP Shim Tool Towards User-Friendly Single-Step B0 Homogenization
Karl Landheer1 and Christoph Juchem1,2

1Biomedical Engineering, Columbia University, New York, NY, United States, 2Radiology, Columbia University, New York, NY, United States

Synopsis

Fast, automatic shimming by mapping along projections (FASTMAP) is an elegant analytical method developed to quantify 3-dimensional first and second order spherical harmonic B0 shapes along six 1-dimensional column projections. The straightforward application of this theoretical concept to B0 shimming, however, neglects crucial aspects of sequence implementation and shim hardware, commonly necessitating multi-step iterative adjustments. Considering experimental imperfections of the employed B0 mapping and shim coil hardware, we demonstrate optimal single-step adjustment of first and second order terms (with potential <3% refinement of linear terms) in the anterior cingulate cortex, one of the most difficult-to-shim areas in the human brain.

Introduction

The best B0 magnetic field homogeneity is of utmost importance to maximize signal-to-noise ratio and spectral dispersion in magnetic resonance spectroscopy (MRS)1. Gruetter et al. presented an efficient analytical method for the determination of first and second order spherical harmonic field terms from selected B0 measurements along six 1D pencil-beam projections2,3 (Figure 1). To date, more than 25 years after the introduction of 'Fast Automatic Shimming Technique by Mapping Along Projections' (FASTMAP), however, the method remains scarcely used in the MR community. Despite the analytical nature of the theoretical concept, in experimental reality, a series of iterations is typically needed for B0 shim fields (and coil currents) to converge at optimal conditions, thereby rendering the procedure lengthy. B0 shimming is a single-step process when the shim system at hand is well-characterized and the appropriate numerical procedures are used. Here, we discuss systematic imperfections inherent to the straightforward implementation of the theoretical FASTMAP concept for B0 shimming and propose corrections for the most significant error sources towards single-step shim adjustments.

Methods

An implementation package is presented that comprises a FASTMAP sequence for a General Electric platform (GE Healthcare, Waukesha, WI, USA), a MATLAB-based (Mathworks, Natick, MA, USA) user-software employing an easy-to-use graphical user interface (GUI, Figure 2A) for calibration and shimming, and scripts for experimental shim current handling. The established FASTMAP Shim Tool (FAMASITO) is made available to the MR community under a term-limited academic license free of charge4. All experiments were performed on a clinical MR750 3T GE MRI system at the New York State Psychiatric Institute (NYSPI) with a standard eight-channel head coil receiver. The in-house FASTMAP sequence consisted of two slice-selective SLR linear-phase RF pulses with an isotropic column width of 5 mm (TE 40 ms, TR 800 ms, bandwidth 15.6/31.3 kHz). FAMASITO shimming for MRS was applied to 30x30x30 cm3 voxels placed in the anterior cingulate cortex of 7 male volunteers, motivated by its relevance in neurology5 and psychiatry6 while being notoriously difficult to shim. An in-house semi-LASER sequence7 was used for MRS and employed fourth-order hyperbolic-secant RF pulses8, VAPOR9 water suppression and DOTCOPS-optimized crusher scheme10 (TE 40 ms, TR 2000 ms). Spectral processing was done with the free MRS software package INSPECTOR11.

Results and Discussion

1) Correction of B0-Induced Geometric Distortions

The use of frequency-encoding for B0 mapping suffers from geometric distortions and leads to systematic errors in coil calibration experiments employing arrayed amplitude schemes. The resultant calibration imperfections, however, hold responsible for the suboptimal adjustment of all future shimming, necessitating additional iterations for refinement. FAMASITO employs a new algorithm for geometric distortion correction based on non-linear spatial regridding to achieve high location accuracy even in the presence of severe B0 background variations (Figure 3).

2) Consideration of Shim Coil Cross-Terms

The conversion of measured field shapes to sets of shim coil currents with FAMASITO furthermore considers cross-terms of the employed shim system.12 The avoidance of secondary field artifacts with the application of shim fields constitutes a second important improvement towards single-step B0 shimming.

3) Automated Calibration

The automated serial acquisition of coil calibration fields with FAMASITO together with its automated routines for processing and calibration analysis are the basis for consistent self- and cross-term characterization of the coil system at hand in absolute units (Hz/cmn) in a consistent and user-friendly fashion (Figure 4).

4) B0 Shimming for 1H MRS of the Human Brain

High quality shimming was consistently obtained in the anterior cingulate cortex with FAMASITO in a single adjustment with residual linear imperfections <3% of the dynamic range (median 0.3%, mean 0.7%) even in this difficult-to-shim region (Figure 2, detailed impact analysis of distortion and cross-term correction not shown). FAMASITO enabled high quality spectra in all volunteers (Figure 5) with measured NAA linewidths of 7.2, 8.2, 8.9, 8.8, 7.0, 6.8, 5.8 Hz, respectively.

5) Limitations

Displacement and bending of 1D beam projections perpendicular to their orientation due to severe background fields can be minimized experimentally with the use of stronger slice selection gradients (which was not possible here due to RF power constraints). They were not considered in this research and are hypothesized to be responsible for the remaining minor adjustment imperfections.

Conclusion

Single-step B0 shimming relies on exact knowledge of 1) the B0 distortion, 2) the employed coil system and 3) use of optimal numerical techniques. A novel software package, FAMASITO, was developed for improved FASTMAP shimming in a user-friendly fashion. The presented one-stop-shop solution alleviates the main hurdles preventing efficient wide-spread use and facilitates high quality MRS even in difficult-to-shim brain regions such as the prefrontal cortex.

Acknowledgements

Special thanks to New York State Psychiatric Institute (NYSPI) and Dr. Feng Liu for their facilities and technical support. This research was supported by the National Multiple Sclerosis Society (NMSS, RG-5319).

References

  1. Juchem, C. and R.A. de Graaf, B0 Magnetic Field Homogeneity and Shimming for In Vivo MR Spectroscopy. Anal Biochem 529:17-29 (2017).
  2. Gruetter, R., Automatic, localized in vivo adjustment of all first- and second-order shim coils. Magn Reson Med 29:804-11 (1993).
  3. Gruetter, R. and C. Boesch, Fast, noniterative shimming of spatially localized signals. In vivo analysis of the magnetic field along axes. J Magn Reson 96:323-334 (1992).
  4. Landheer, K. and C. Juchem. FAMASITO - FASTMAP Shim Tool for Efficient B0 Shimming. Columbia Tech Ventures (CTV) License CU18208 (2018). Available from: innovation.columbia.edu/technologies/CU18208_famasito.
  5. Prinsen, H., R.A. de Graaf, G.F. Mason, D. Pelletier, and C. Juchem, Reproducibility measurement of glutathione, GABA, and glutamate: Towards in vivo neurochemical profiling of multiple sclerosis with MR spectroscopy at 7T. J Magn Reson Imaging 45:187-198 (2017).
  6. Milak, M.S., C.J. Proper, S.T. Mulhern, A.L. Parter, L.S. Kegeles, R.T. Ogden, X. Mao, C.I. Rodriguez, M.A. Oquendo, R.F. Suckow, T.B. Cooper, J.G. Keilp, D.C. Shungu, and J.J. Mann, A pilot in vivo proton magnetic resonance spectroscopy study of amino acid neurotransmitter response to ketamine treatment of major depressive disorder. Mol Psychiatry 21:320-7 (2016).
  7. Scheenen, T.W., D.W. Klomp, J.P. Wijnen, and A. Heerschap, Short echo time 1H-MRSI of the human brain at 3T with minimal chemical shift displacement errors using adiabatic refocusing pulses. Magn Reson Med 59:1-6 (2008).
  8. Tannus, A. and M. Garwood, Adiabatic pulses. NMR Biomed 10:423-34 (1997).
  9. Tkac, I., Z. Starcuk, I.Y. Choi, and R. Gruetter, In vivo 1H NMR spectroscopy of rat brain at 1 ms echo time. Magn Reson Med 41:649-56 (1999).
  10. Landheer, K. and C. Juchem, Dephasing optimization through coherence order pathway selection (DOTCOPS) for improved crusher schemes in MR spectroscopy. Magn Reson Med (2018), [Epub ahead of print].
  11. Juchem, C. INSPECTOR - Magnetic Resonance Spectroscopy Software. License CU17130 (2016). Available from: innovation.columbia.edu/technologies/cu17130_inspector.
  12. Juchem, C., T.W. Nixon, P. Diduch, D.L. Rothman, P. Starewicz, and R.A. de Graaf, Dynamic shimming of the human brain at 7 Tesla. Concepts Magn Reson 37B:116-128 (2010).

Figures

Figure 1: B0 field distributions are measured with the FASTMAP along six selected 1-dimensional column projections (left: blue columns, right: blue solid curves) and decomposed into first and second order polynomial shapes, thereby allowing to extract the B0 behavior across a region-of-interest (vertical green bars). The solution of the resultant set of linear equations is then converted to first and second order B0 shim requirements (red dashed) and to predict the achievable B0 homogeneity (gray solid).

Figure 2: Example FAMASITO shimming in the anterior cingulate cortex (ACC). A) Graphical user-interface (GUI) for data visualization, processing and calculation of optimized shim currents, B) Standard output summarizing the acquisition pair for each of the six column projections (blue: TE 40 ms, red: TE 40+3 ms, vertical green: voxel boundaries), C) Standard output describing the measured B0 field conditions (blue), the determined optimal parabolic representation (red) and the predicted B0 behavior with shimming (gray), D) B0 mapping (blue) confirms homogenous conditions under the influence of the applied shim fields as horizontal line locally within the optimized region (green).

Figure 3: Impact of spatial correction procedure (right vs. left) on accuracy of 1D magnitude images (top) and field projections (bottom) of selected coil calibration experiments (blue-to-red: equidistantly spaced calibration shim settings). A) Linear shims lead to systematic image deformation and B0 quantification errors that are compensated with the established algorithm (black arrows). B) The Z2 coil introduces systematic frequency offsets that are misinterpreted as field gradients if not corrected. C) Minor but noticeable imaging errors were induced by other coils. D) Even coils with negligible image distortions alter the B0 behavior along projections that should not be affected (cross-terms).

Figure 4: FAMASITO employs automated B0 calibration series and analyses of all shim coils considering the intended B0 shapes of individual coils (self-terms) as well as shape imperfections that can be described by other coils of the coil system (cross-terms). A) Example X2-Y2 calibration analysis. B) The normalized calibration matrix is dominated by self-terms (on the diagonal) and second-to-first order cross-terms (5x3 area in lower left corner). If the shim system's cross-terms are known, they can be immediately considered in the first conversion of B0 field imperfections to optimized coil currents, thereby rendering the subsequent compensation of cross-terms by additional iterations obsolete.12

Figure 5: Axial anatomical for subject 1 with MRS voxel overlaid (top left) and seven spectra obtained from seven different healthy volunteers in the anterior cingulate cortex. Volunteers 1-4 had only a single adjustment of FASTMAP applied, whereas volunteers 5-7 benefited from a single refinement of the linear shims at <3% of the dynamic range. All seven subjects exhibited strong B0 field distortions prior to shimming comparable to the FAMASITO example in Figure 2.

Proc. Intl. Soc. Mag. Reson. Med. 27 (2019)
0216