Combining magnetic resonance imaging (MRI) and positron emission tomography (PET) shows promise to be a powerful tool as it provides complementary anatomical and molecular information about diseases (1, 2). However, the availability of integrated PET/MRI (with PET and MRI “locked” together) has been limited due to its high cost. A few research groups have developed PET inserts for simultaneous PET/MRI operation with the insert sharing the same electrical ground as the MRI RF coil; in this case the RF field cannot penetrate the grounded insert, requiring the RF transmitter to be moved inside the PET ring (3-5). Instead, we have developed a PET insert that is electrically floating with respect to the MR system, enabling an external RF field from the MR body coil to penetrate inside. In our previous work, we have shown that the RF transmit field penetrates through the gaps between detectors as well as the ends PET insert with no significant mutual interference (6-8). In this work, we analyzed the B1 maps, MR images and performed EM simulations with various PET configurations to better understand how the RF field enters the PET insert.

To investigate the RF penetrability in various configurations we have performed MR experiments using 3-Tesla MRI (GE Healthcare) and simulation studies using XFdtd (Remcom) and Maxwell (ANSYS).

We experimentally and numerically examined the effect of powering the PET detector modules and the importance of inter-modular gaps and end openings of the PET insert to achieve the goal of RF-penetrability. The PET system comprises a ring of 16 copper-shielded PET detector modules with 1 mm inter-module gaps. Electrical isolation was achieved via analog optical links and floating battery power. The gaps or ends of the PET insert were alternately blocked using the copper shielding tape (Figure 1). For the MR experiment, B1 field maps and gradient echo (GRE) and fast spin echo (FSE) MR images were acquired using the MRI body coil as both transmitter and receiver (requiring the RF field to penetrate into and out of the PET insert). Configuration (a,b,c) in Fig. 1 were auto-prescanned and (d,e) were manual-prescanned to the TG of (c).

EM simulation studies with various module shielding dimensions and inter-module gap size were performed to improve the RF-penetrability. For XFdtd 3D simulations (Figure 2, Left), the PET module shielding length was increased from 20 to 30 cm, inter-modular gaps were widened from 1 to 5 mm, and the shielding height was shortened from 40 to 30 mm. For Maxwell 2D simulations (Figure 2, Right), the inter-modular gap was increased from 1 to 21 mm with increments of 5 mm and the detector height was shortened from 40 to 20 mm with decrements of 5 mm.

The results from MR experiments and electromagnetic simulations are shown in Figure 3.

From the B1 map analysis, surprisingly, higher B1 uniformity was acquired with the PET detector present (auto-pre-scanned) (b) compared to the no-PET case (a), and no significant difference was observed with or without battery power connected to the PET detector (c).

Since the body coil is used as the receiver, the SNR drop in GRE and FSE images are due to RF receive attenuation, but this attenuation can be avoided with a RX-only coil inside the PET.

When either gaps (d) or ends (e) are blocked, poor B1 maps and MR images are acquired due to extremely low receive sensitivity. RF field entering through the ends contributes to the B­1 field magnitude, but poor uniformity (Fig. 3 d). On the other hand, RF field entering through the gaps contribute to the uniformity as long as the ends are open (Fig. 3 c, d). These similar trends were observed in the electromagnetic simulations as well.

For improving the RF-penetrability, shortening the length, height of the PET detector module shielding and/or widening the inter-module gaps can improve the RF-penetrability of the PET insert (Figure 4).

We have shown that the RF field from the MR body coil penetrates through the inter-modular gaps and the ends of an electrically floating PET insert. The RF field transmitted through the ends mostly contributes to the field magnitude while the field entering through the gaps improves the uniformity when the ends are also opened. The electromagnetic simulations show that decreasing the detector height and/or widening the inter-modular gaps enhance the RF-penetrability by ~16%.