Johnson Yuen
St. George Hospital, Cancer Care Centre
Medical Physicist

Andrew Howie
Medical Physicist
St. George Hospital, Cancer Care Centre

Ryan Brown
Medical Physicist
St. George Hospital, Cancer Care Centre

Dean Cutajar
Research Physicist
Centre for Medical Radiation Physics, University of Wollongong

Andrej Bece
Radiation Oncologist
St. George Hospital, Cancer Care Centre

Joseph Bucci
Radiation Oncologist
St George Cancer Care Centre

Background and Purpose

Deformable image registration(DIR) allows for the summation of doses from multiple radiation treatment deliveries to accounts for previous treatment –assessed to be frequently missed and hard to detect by TG100[1]. High dose rate (HDR) brachytherapy treatments involving multiple fractions were assessed using the Brachytherapy Treatment Planning System (BTPS) Oncentra Brachy and DIR software (MIM). Recommendations for optimisation of this process have been developed, which provide a framework to account for multiple treatments in brachytherapy, extending to treatments that include external beam radiotherapy plans.


Both rigid and deformable registration were performed for HDR brachytherapy procedures involving multiple fractions, with the newer fraction set as the primary dataset. Deformable registrations were subject to user optimisation. Registration QA analysis with the Dice similarity coefficient and mean distance to agreement for both registrations were compared, indicating spatial error. Dose volume histogram and dose constraints for the urethra, rectum, and target were evaluated.


The assessment of HDR brachytherapy treatments involving multiple fractions was successful. Commissioning results from BTPS to the DIR software revealed inherent sampling errors in the DVH metrics. Deformable registration QA reduced spatial error compared to rigid registration and DIR error can be dependent on user optimisation[2]. The effect of DIR errors on DVH metrics were assessed. Limitations of this technique were identified which include factors such as variations in contouring, changes in organ, and radiobiology considerations.


Dose accumulation with DIR can offer superior results to rigid registration in terms of spatial errors which translate to improved accuracy of dose metrics in DVHs. Clinical use of dose accumulation of multiple treatments was demonstrated in the department with scripted reports based on patient-specific QA that include estimated errors as well as a list of residual limitations. These investigations form the framework for arbitrary accounting for treatments to and from external beam plans with dose.


1. Huq, M.S., et al., The report of Task Group 100 of the AAPM: Application of risk analysis methods to radiation therapy quality management. Medical Physics, 2016.

2. Johnson, P.B., et al., Evaluation of the tool “Reg Refine” for user-guided deformable image registration. Journal of Applied Clinical Medical Physics, 2016. 17(3).

← Back to all abstracts