Radiation Induced Current Effects On MR Images From An Integrated Linac-MR System
B Burke1*, K Wachowicz1,2, B Fallone1,2, S Rathee1,2, (1) University of Alberta, EDMONTON, AB, (2) Cross Cancer Institute, Edmonton, AlbertaTH-E-BRA-9 Thursday 1:00:00 PM - 2:50:00 PM Room: Ballroom A
Purpose: During the real-time MR image acquisition of an integrated linac/MRI system, the MRI's RF coils is exposed to pulsed radiation resulting in radiation induced currents (RIC). This work will: (a) visualize the RIC signal as artefacts in k-space and determine its effect on MRI's signal-to-noise ratio (SNR), (b) examine the effects of linac repetition rate (MU/minute) and MRI imaging sequence parameter 'TR' on the RIC artefact, (c) use post processing methods to remove the unwanted RIC signal from the MR images.
Methods: A small test phantom was imaged with no radiation first. Phantom imaging was then repeated concurrently with linac producing radiation at various repetition rates in two scenarios: (1) the radiation beam was incident on the RF coil unobstructed and (2) a lead block attenuated the radiation beam before reaching the RF coil. Scenario (1) was repeated by obtaining images for several values of TR. Finally, a post-processing algorithm was applied to the corrupted MR k-space data to remove the RIC artefact.
Results: The RIC artefact presented as near vertical lines in k-space data for integer ratios of TR to linac pulsing period (rate = 180 Hz). For non-integer ratios, the artefact lost its regular pattern and became random in appearance. The RIC artefact disappeared from the k-space data when the linac's radiation was blocked. The image SNR decreased with increasing linac repetition rate. The post-processing method was successful in restoring a significant fraction of the lost image SNR.
Conclusions: Signal spikes observed in the k-space data are confirmed to result from RIC. The SNR reduction in MRI images, due to the RIC, is directly related to the linac repetition rate. The artefact's appearance depends on the relationship between linac pulse repetition rate and image sequence timing. Our post-processing algorithm allows the recovery of the lost image SNR.