To analyze the design criteria required to integrate an MRI and a medical linac to provide real-time image-tracking guidance during radiotherapy delivery.
Materials and Methods
A novel method to mount a linac on a bi-planar MRI system is introduced. This means of coupling allows radiotherapy treatment of patients from all orientations without mechanical interference between the MRI and linac, and removes most of the magnetic interference between the two devices. Finite element methods are used to model the Linac MRI system, to calculate the magnetic fields in its vicinity, and to further optimize magnetic shielding to confine the magnetic fringe fields within the accelerating waveguide to less than 0.5 Gauss. Measurements of radiofrequency (RF) noise from the linac suggest that such a noise results from a resonance phenomena. RF interference from linac into MRI system can be removed by modifications to the electrical delay lines between PFN and klystron. Monte Carlo dose calculations are performed including magnetic field effects for various tissue scenarios: e.g, tissue only, various air-tissue geometries, and for various beam sizes under different static magnetic fields. MR image quality and acquisition speed are investigated to develop practical methods for real-time tracking during the treatment delivery.
Results
A permanent-magnet design will be used, where the magnet and linac rotate together such that the radiation beam is always unobstructed, and the linac remains stationary relative to the magnetic field thus minimizing image distortions. The fringe fields around a 0.2 T MRI have been modeled and the magnet has been designed in such a way as to minimize the magnetic interference with the linac. Simple passive shielding can be practically incorporated to achieve the magnetic shielding goal which results in minimal magnetic field inhomogeneities that can be easily controlled through conventional “shimming”. MRI using 0.2 T magnetic fields offers substantially better soft tissue image contrast (lungs, prostate, GBM, etc) than any existing imaging devices used for image-guided radiotherapy. Fast image acquisition sequences will be incorporated which will permit real-time image guidance in 3D in addition to pre-treatment imaging. Results of our Monte Carlo simulations for different air-tissue and beam geometries indicate that the optimal magnetic field strength of 0.2 T will have minimal effect on the dosimetry.
Conclusions
The novel method to couple a linac to an MRI can allow rotational radiotherapy with 3D image guidance in real-time. Magnetic field and RF interference can be eliminated. The effects of a 0.2 T magnetic field on the radiotherapy dose distribution are minimal, and can easily be handled with conventional dose calculation methods.
FULL TEXT