Accurate HU Generation for MR-LINAC Using MV-CBCT
Y Hu1,2, C Williams1,2, R Berbeco1,2, M Myronakis*, (1) Brigham & Women's Hospital, Harvard Medical School, Jamaica Plain, MA, (2) Dana Farber Cancer Institute, Boston, MA
Presentations
SU-J-202-2 (Sunday, 7/10/2022) 4:00 PM - 5:00 PM [Eastern Time (GMT-4)]
Room 202
Purpose: Incorporation of an MV imager in an MR-LINAC may have several applications, including beam QA and direct measurement of electron density maps. In this study, we demonstrate proof-of-concept of electronic portal imaging device (EPID) integration in a 0.35T MR-LINAC for 3D megavoltage cone-beam CT (MV-CBCT) imaging acquisition.
Methods: MR-LINAC and EPID digital models were constructed based on geometrical and operational details of the MR-LINAC currently installed in our institute. The shims used to stabilize the magnetic field were included in the MR-LINAC model. The MV detector location was outside the magnetic field during operation. A CATPHAN phantom section with material inserts was modeled and MV-CBCT image acquisition with 6MV FFF beam compatible with the MR-LINAC was simulated. The GATE Monte Carlo tool was used for the simulation. The FDK algorithm with Hanning filter was used for the reconstruction. Scatter images were extracted from simulation and used during reconstruction to correct for scatter. Hounsfield Unit (HU) accuracy and Contrast-to-Noise Ratio (CNR) were calculated.
Results: A subtraction algorithm successfully removed the shims from the projection images. Complete MV-CBCT was reconstructed from 144 projections covering 360 degrees. In the Catphan phantom, the inserts with highest CNR were Air and 50% Bone with 9.7 and 5.34 CNR respectively. The HU difference between the Air insert at top and bottom of the CATPHAN was 0.43%. The HU value for each material insert was consistent with the expected range of values measured with 120kV-CBCT and 6MV FFF conventional LINAC.
Conclusion: Preliminary results demonstrated consistent HU values in the reconstructed CATPHAN image. Implementation of MV-CBCT imaging in the MR-LINAC can be used for accurate electron density maps that will be better registered to the simultaneous MR images for more robust treatment planning.
Funding Support, Disclosures, and Conflict of Interest: The study is funded by ViewRay
Keywords
Taxonomy
IM/TH- MRI in Radiation Therapy: Development (new technology and techniques)
Contact Email