Click here to


Are you sure ?

Yes, do it No, cancel

An Investigation of Image Registration Techniques for Dose Summation of External Beam Radiation Therapy and Prostate Seed Implants

S To1,2*, D Guest2 , A Riegel1,2 , (1) Department of Physics and Astronomy, Hofstra University, Hempstead, NY 11549, (2) Department of Radiation Medicine, Northwell Health, Lake Success, NY 11042


(Sunday, 7/14/2019)  

Room: ePoster Forums

Purpose: Combining external beam radiation therapy (EBRT) and prostate seed implants (PSI) has yielded improved local control for intermediate to high risk prostate cancer at the cost of increased genitourinary toxicity. Accurate dosimetry is needed to determine tolerances for organs-at-risk near the boosted prostate volume. This study compares the dosimetric impact of registration techniques on summing 3-D EBRT and PSI dose distributions.

Methods: Post-operative CT, EBRT CT simulation, and corresponding dose distributions were retrospectively acquired for 25 patients. Three CT/CT registration techniques were utilized: (1) Image-based rigid registration; (2) Rigid registration based on implanted EBRT localization fiducials; (3) B-spline deformable image registration focused on the prostate and nearby critical organs. Physical dose distributions were converted to biological effective dose (BED) utilizing the linear-quadratic model and summed according to each CT/CT registration. Relevant dosimetric parameters were extracted for prostate, urethra, rectum, and bladder and compared statistically using analysis of variance and Tukey honestly significant difference test (α<0.05).

Results: Among 21 dosimetric values extracted from all regions-of-interest, only 4 in the Rectum (D15: p<0.05, D25: p<0.05) and Prostate (Dmean: p<0.01, D90: p<0.01) were significantly different. Tukey’s test showed image-based rigid and deformable were significantly different for the rectum while beacon-based rigid was significantly higher than other registration methods for the prostate. The average standard deviation of image-based rigid was significantly higher compared to beacon-based rigid and deformable. In some cases, deformable registration produced artifacts, most likely due to the presence of seeds in post-operative imagine and lack of seeds in EBRT CT simulation.

Conclusion: Rigid registration based on implanted EBRT localization fiducials may be sufficient for spatial registration and accurate dose summation of combined EBRT/PSI therapy for prostate cancer. Future work includes an expansion to a larger sample of patients and correlation of combined dosimetry with toxicity.


Prostate Therapy, Registration, Brachytherapy


TH- Brachytherapy: Registration

Contact Email