Dynamic backlash of a multileaf collimator
PO-1797
Abstract
Dynamic backlash of a multileaf collimator
Authors: Antti Kulmala1, Jenna Tarvonen2, Timo Paasonen2, Anna Rintala2, Lauri Koivula2, Mikko Tenhunen2
1Clinical Research Institute HUCH Ltd., Radiotherapy, Helsinki, Finland; 2Comprehensive Cancer Center, Radiotherapy, Helsinki, Finland
Show Affiliations
Hide Affiliations
Purpose or Objective
With increasing complexity and detailing of radiotherapy treatments, mechanical imperfections, in part caused by multileaf collimator (MLC), are becoming significant for treatment delivery validity. A built-in feature of MLC robotics is backlash, which causes a drive position, defined by a control point, and an actual position of a collimator to differ. Advancing the determination of the backlash may provide more knowledge of the actual leaf positions during treatment delivery and hence enable enhancing of treatment modelling.
To determine the backlash of static collimator leafs as in 'step & shoot' delivery, a robust measurement method has been introduced [1]. The method has also been used to study a variety of MLCs [2]. However, since most of the high-end radiotherapy treatments have a dynamic collimation as in 'sliding window' delivery, there is yet a need to further characterize the backlash.
In this work we present a vendor independent method for determining a backlash of a dynamic multileaf collimation. In addition, the method is applied to two technically different commercial MLCs and to a variety of collimator leaf speeds.
Material and Methods
In this study the backlash is described as a difference in a radiation fluence width when a collimator leaf is pushed in to a radiation beam and when the collimator leaf is pulled out from the beam. Dynamic test fields were designed to highlight the width difference.
The method was used to define backlash for three 'Varian Standard 120M'/TrueBeam (Varian Medical Systems Inc., Palo Alto) and three 'Agility'/VersaHD (Elekta Instrument AB, Stockholm) MLCs. The definitions were repeated with collimator leaf speeds ranging from 0.25 cm/sec to the maximum leaf speed. The actual fluence widths of closing and opening collimation were determined with megavolt imager and ionization chambers.
Results
An average dynamic backlash for studied collimator leafs was: 0.4 mm +/- 0.1 mm for 'Varian Standard 120M' and 0.8 mm +/- 0.2 mm for Agility MLC. The backlash was found to be almost linearly dependent of the leaf speed with both studied MLC models. The maximum backlash was 2.3 mm recorded with Agility MLC with the maximum leaf speed.
Conclusion
We have demonstrated that the proposed method can be used to determine the backlash of a dynamic multileaf collimation. The dynamic backlash has been found to be dependent of the MLC model and the collimator leaf speed.
References
[1] A. Kulmala, A. Rintala, L. Koivula, V. Petäjä and M. Tenhunen, A robust method for determining backlash of a multileaf collimator, ESTRO Annual Meeting, 2022, PO-1510
[2] A. Kulmala, T. Paasonen, L. Koivula, A. Rintala and M. Tenhunen, Backlash of a Multileaf Collimator, AAPM 64th Annual Meeting & Exhibition, 2022, PO-GePV-T-375