Impact of transducer position on B 1 + homogeneity and RF safety in transcranial MR‐guided focused ultrasound: An electromagnetic simulation study
Eunwoo Lee, Taewoo Nam, Daniel Hernandez, Kisoo Kim, Kyoung‐Nam KimAbstract
Background
Transcranial magnetic resonance‐guided focused ultrasound (tcMRgFUS) allows precise focal heating under magnetic resonance imaging (MRI) guidance for neurological therapies. However, large hemispherical transducers inherently affect the transmit radiofrequency (RF) field (|B 1 + |) inhomogeneity, producing dark/black‐band artifacts owing to general electromagnetic (EM) field interactions.
Purpose
As the position and orientation of the transducer vary in practice, we assessed the effects of such changes on the |B 1 + | field and specific absorption rate (SAR). We compared relative variations across realistic configurations to identify clinically meaningful trends.
Methods
Three‐dimensional (3D) finite‐difference time‐domain simulations were performed using Sim4Life (v7.2.1, ZMT, Zürich, Switzerland). The simulation model comprised a 16‐leg high‐pass birdcage RF coil matched to a standard clinical 3T whole‐body scanner (bore diameter: 600 mm), a generic hemispherical tcMRgFUS system (transducer and water bolus), and a Duke human model. A spatial voxel mesh was applied with a maximum size of 2×2×2 mm 3 for the RF coil, tcMRgFUS system, and the head region of the human model, yielding approximately 43 to 48 million mesh cells per configuration. Transducer translations (lateral: ± 40 mm, vertical: ± 40 mm, longitudinal: −10 to +20 mm) and rotations (−50° to +10°) were simulated across 85 configurations (78 translations, and 7 rotations). The peak spatial SAR averaged over 10 g ( ps SAR 10g ), head‐averaged SAR ( hd SAR), and |B 1 + |‐field distributions were computed and normalized to 1 W input power to quantitatively assess the impact of the modeled positioning configurations within the investigated parameter space.
Results
Lateral and vertical translations (± x and ± y) that brought the transducer closer to the head intensified EM coupling between the RF coil and the transducer, increasing |B 1 + | inhomogeneity, whereas +z translation similarly strengthened this coupling by concentrating its EM influence over a smaller cranial volume. In contrast, ‐z translation spread dark‐band artifacts more uniformly across the head, paradoxically improving homogeneity despite a lower spatially averaged |B 1 + |‐field. Across all translation configurations, 3D |B 1 + | homogeneity and mean intensity decreased by up to 3.26% and 0.004 µT, respectively, whereas rotational changes produced reductions of up to 8.53% and 0.03 µT. In terms of RF safety, the presence of the transducer increased ps SAR 10g by 15.79% (from 0.0646 to 0.0748 W/kg) and hd SAR by 77.42% (from 0.0186 to 0.033 W/kg). Among positional changes, ps SAR 10g and hd SAR increased by up to 0.04 W/kg at a lateral translation of (40, 0, 0) mm and 0.009 W/kg at a rotation of −50°, respectively. These findings indicate that both the extent of head inclusion within tcMRgFUS and the transducer–head distance should be actively managed to minimize |B 1 + | perturbation and associated SAR increases.
Conclusions
These results demonstrate that transducer position and angle substantially affect |B 1 + | homogeneity and ps SAR 10g , with direct implications for system performance and safety. These simulation‐based findings highlight that the extent of head inclusion within tcMRgFUS and the transducer–head distance are key factors governing EM field perturbation and RF safety. Therefore, accounting for positional and angular variability is essential for the setup and operation of tcMRgFUS systems, and the relative trends identified provide a basis for informing clinical positioning protocols and evaluation frameworks.