Like other magnetic resonance (MR) techniques before it [1], MR Fingerprinting (MRF) [2, 3] was developed and applied in the traditional context of a precisely calibrated and uniform radiofrequency excitation field (B₁⁺). In practice, however, the B₁⁺ field distribution is rarely uniform. Therefore, Plug & Play Parallel Transmission (PnP-PTX) [4, 5] was designed, which liberates MRF from these constraints and abandons the longstanding campaign for a uniform B₁⁺ field. In this work we evaluate impact of B₁⁺ field heterogeneities on abdominal MRF experiments at different field strengths and explore the potential of PnP-PTX to enable quantitative body imaging at ultra-high field strengths.

Subject-specific electrodynamic interactions between the incident RF field and patient anatomy preclude the construction of a tractable general-purpose uniform B₁⁺ source [6]. Instead, the circularly polarized (CP) mode of a birdcage coil is often used to approximate a uniform field [7].

We simulated the CP-mode of a 16 rung body coil (40 cm diameter) loaded with the Duke human body model (2×2×2mm³, [8]) at resonant frequencies corresponding to 1.5, 3.0 and 7.0 Tesla [9]. At 7 Tesla, we also extracted the two linear modes corresponding to each of the drive points in the coil. Each of the tissues in the body model were assigned literature T1, T2 and PD values corresponding to 1.5 Tesla [10,11]. Although T1 and T2 values are field strength dependent, in this work we kept the values constant to better visualize the impact of B₁⁺ field heterogeneities on the reconstructed quantitative maps.

The extended phase graph formulism [12] was used to perform synthetic MR experiments in an axial slice through the abdomen using the simulated B₁⁺ fields. The MRF sequence design and reconstruction was performed as described by Ma, et al [2]. At 7 Tesla, the two linear-modes of the body coil were used to create a synthetic PnP-PTX experiment through the same slice. The PnP-PTX sequence design and reconstruction was performed as described by Cloos, et al [5].

Figure 1 shows the B₁⁺ field distribution of the CP-mode an axial slice in the abdomen at 3 different field strengths. As the field strength increased from 1.5 to 3.0 Tesla characteristic B₁⁺ depressions, often observed in large subjects [13], are formed in the anterior and posterior areas of the body. Nevertheless, the fidelity of the adiabatic inversion [14] used at the start of MRF sequence remains excellent (Fig. 2), and the T1-maps reconstructed using MRF remain accurate (Fig. 3, middle row). The T2-maps, on the other hand, already show significant deviations from the ground truth at 1.5 Tesla. The T2 values in the left and right kidney, for example, are radically different. Comparing the B₁⁺ field distributions with the PD maps (Fig. 3, top row), it is clear that the B₁⁺ distribution is directly imposed on the PD map.

At 7 Tesla, the aforementioned RF-depressions grow out to form RF-voids (Fig. 1, right), which cannot be mitigated using an adiabatic pulse (Fig. 2 right). As a result, the reconstructed T1-map shows strong artifacts in those areas. These same imperfections in the inversion pulse also propagate into the T2 and PD map which are heavily distorted. While it is possible to leverage a dedicated B1 calibration scan to correct the distortions produced by the mild B₁⁺ non-uniformities observed at 1.5 and 3.0 Tesla [15], it is impossible to extract a viable signal from areas which are devoid, of RF such those shown here at 7.0 Tesla.

Figure 4 shows the decomposition of the CP-mode into the two linear modes. Although each of these modes is more heterogeneous than the CP-mode, their field patterns have a complementary distribution. Areas that are devoid of RF in one mode, are exposed by the other. Instead of relaying on an adiabatic pulse and relatively uniform field distributions, the PnP-PTX sequence interweaves these modes to extract signal from all areas through the slice making it possible to perform accurate measurements in challenging situations such as the abdomen at 7 Tesla (Fig. 5). In this proof of principle demonstration, we used the two linear modes because these are directly accessible in a birdcage coil. However, other coil-mode configurations may offer additional advantages.

Research systems operating at 7 Tesla predominantly focus on brain and extremity imaging because of the extreme RF non-uniformities encountered in the body. We showed that PnP-PTX has the potential to alleviate these challenges, and thereby opening up a new route towards robust, quantitative, whole-body MRI for ultra-high-field systems.