Click here to


Are you sure ?

Yes, do it No, cancel

Validation of a Commercial Continuous Hounsfield Unit (HU) Pelvis Synthetic CT Platform for MR-Only Prostate Radiotherapy Treatment Planning

V Yu*, M Hunt, L Cervino, A Damato, N Tyagi, Memorial Sloan Kettering Cancer Center, New York, NY


(Sunday, 7/12/2020)   [Eastern Time (GMT-4)]

Room: AAPM ePoster Library

Purpose: (MR-for-Calculating-ATtenuation)Prostate, a commercial T1-weighted mDixonFFE-based synthetic CT solution with bulk density assignment, is routinely utilized in our clinical workflow for MR-only prostate radiotherapy. In this study, we evaluate the HU assignment and dosimetric accuracy of MRCAT-GeneralPelvis, an extended version with continuous-HU assignment, for clinical use.

Methods: and MRCAT-GeneralPelvis images were acquired on a Philips 3T Ingenia MR-RT simulator for eight patients receiving MR-only prostate radiotherapy to compare dosimetric differences between the two versions. HU was compared for 3 patients that had associated CT images due to prior brachytherapy or re-simulation. The HU of bony and soft tissue landmarks were evaluated and compared among CT, MRCAT-Prostate, and MRCAT-GeneralPelvis. Dosimetric comparison was performed by recalculating clinical VMAT plans on the MRCAT-GeneralPelvis synthetic-CT and associated CT after rigid registration. Plans varied between hypofractionated and standard fractionation for prostate or prostate with lymph nodes. Dose metrics for the target volumes and surrounding organs-at-risk (OAR) were compared. Digitally reconstructed radiographs (DRR) were generated and qualitatively compared among the CT and the two synthetic-CT methods.

Results: HU agreement to CT was shown for MRCAT-GeneralPelvis. The average absolute HU difference to CT was 95.3 for MRCAT-Prostate, and 34.8 for MRCAT-GeneralPelvis. For all plans evaluated, the average percent dose difference between the two synthetic CTs for the PTV D95, maximum and mean were <0.8%. OAR maximum and mean dose differences were within 1.3%. For plans with CT, the PTV metrics of both synthetic-CTs were within 1.5% from the CT, with MRCAT-GeneralPelvis trending closer to the associated metrics on CT. Improved visualization of fiducial markers and rectal spacer was also observed. Realistic DRRs were generated from MRCAT-GeneralPelvis with excellent agreement with those from CT.

Conclusion: continuous-HU MRCAT-GeneralPelvis provides more accurate synthetic CTs and DRRs, and is dosimetrically comparable to the currently utilized MRCAT-Prostate with bulk density assignments.

Funding Support, Disclosures, and Conflict of Interest: This work is supported by a masters research agreement with Philips Healthcare


Not Applicable / None Entered.


IM/TH- MRI in Radiation Therapy: MRI for treatment planning

Contact Email