Integrating a low-field open MR scanner with a static proton therapy research beamline: proof of concept

Purpose
On-line image guidance using magnetic resonance (MR) imaging is expected to improve targeting precision in proton therapy (PT). However, to date no hybrid MR-PT system exists. This is partly due to unknown mutual electromagnetic interactions between the MR scanner and the PT beamline, which may compromise the quality of the proton beam or the MR image. The aim of our study was to integrate an MR scanner with a static PT research beamline and to test the feasibility of simultaneous irradiation and imaging.

Materials & methods
An open MR scanner (MrJ2200, Paramed) featuring a 0.22 T vertical permanent magnetic (B0) field was RF-shielded by a compact Faraday cage and placed at the beam exit of a horizontal static PT research beamline (IBA Proton Therapy). To account for Lorentz force-induced beam deflection in B0, the scanner had to be laterally displaced by 2 cm from the beam’s central axis, such that a proper co-alignment of the beam and the scanner’s field-of-view (FOV) could be confirmed by radiochromic film (Gafchromic EBT3, Ashland) measurements. The beam was collimated to Ø10 mm before entering the Faraday cage through a cylindrical beam guide (Fig. 1).

MR test: With the beamline magnets off, the knee, wrist and hip of a volunteer were scanned with STIR gradient echo (GE), T1-weighted GE, and T1-weighted spin echo (SE) imaging, respectively. T1-weighted SE images of a mixed-meat sausage were acquired using a knee coil without beam, with beamline magnets on and while being irradiated at 215 MeV and 5 nA for 5 minutes.

Beam test: With Faraday cage removed, beam profiles were acquired with and without MR scanner for 72, 125 and 219 MeV beams using a pixelated scintillation detector (Lynx, IBA Dosimetry) positioned 110 cm downstream of the MR scanner's isocentre. These measurements were repeated with the MR scanner in place during acquisition of three different SE and GE images (max. gradient strength up to 5.7 mT/m).

Results
MR test: MR imaging in the nearby presence of a PT beamline was feasible and showed sufficient quality for anatomical imaging of human musculoskeletal structures. MR images of the meat sausage acquired with operating beamline showed a uniform translation of <2 mm in frequency encoding direction, but no geometrical distortions.

Beam test: The scintillation detector showed a horizontal beam deflection of 22, 16 and 11 cm for 72, 125 and 219 MeV, respectively, and a vertical beam deflection <0.6 mm. Horizontal deflection was taken into account to install an in-cage beam stopper, while vertical deflection could be neglected. The beam profiles showed no influence of the gradient fields applied during image acquisition.

Conclusions
For the first time, the integration of an MR scanner and a static PT research beamline has been realized. By taking into account the scanner’s B0 field induced beam deflection, simultaneous proton beam irradiation and MR imaging of an object placed in the scanner’s FOV was feasible with acceptable beam and image quality. The beamline-induced MR image shift is subject of an ongoing performance evaluation study.