Estimation of rectal dose using daily megavoltage cone-beam computed tomography and deformable image registration

Yuichi Akino, Yasuo Yoshioka, Shoichi Fukuda, Shintaroh Maruoka, Yutaka Takahashi, Masashi Yagi, Hirokazu Mizuno, Fumiaki Isohashi, Kazuhiko Ogawa

Research output: Contribution to journalArticlepeer-review

17 Scopus citations


Purpose The actual dose delivered to critical organs will differ from the simulated dose because of interfractional organ motion and deformation. Here, we developed a method to estimate the rectal dose in prostate intensity modulated radiation therapy with consideration to interfractional organ motion using daily megavoltage cone-beam computed tomography (MVCBCT). Methods and Materials Under exemption status from our institutional review board, we retrospectively reviewed 231 series of MVCBCT of 8 patients with prostate cancer. On both planning CT (pCT) and MVCBCT images, the rectal contours were delineated and the CT value within the contours was replaced by the mean CT value within the pelvis, with the addition of 100 Hounsfield units. MVCBCT images were rigidly registered to pCT and then nonrigidly registered using B-Spline deformable image registration (DIR) with Velocity AI software. The concordance between the rectal contours on MVCBCT and pCT was evaluated using the Dice similarity coefficient (DSC). The dose distributions normalized for 1 fraction were also deformed and summed to estimate the actual total dose. Results The DSC of all treatment fractions of 8 patients was improved from 0.75±0.04 (mean ±SD) to 0.90 ±0.02 by DIR. Six patients showed a decrease of the generalized equivalent uniform dose (gEUD) from total dose compared with treatment plans. Although the rectal volume of each treatment fraction did not show any correlation with the change in gEUD (R2=0.18±0.13), the displacement of the center of gravity of rectal contours in the anterior-posterior (AP) direction showed an intermediate relationship (R 2=0.61±0.16). Conclusion We developed a method for evaluation of rectal dose using DIR and MVCBCT images and showed the necessity of DIR for the evaluation of total dose. Displacement of the rectum in the AP direction showed a greater effect on the change in rectal dose compared with the rectal volume.

Original languageEnglish (US)
Pages (from-to)602-608
Number of pages7
JournalInternational Journal of Radiation Oncology Biology Physics
Issue number3
StatePublished - Nov 1 2013


Dive into the research topics of 'Estimation of rectal dose using daily megavoltage cone-beam computed tomography and deformable image registration'. Together they form a unique fingerprint.

Cite this