EA Hewson¹*, A Dipuglia², Y Ge³, R O'Brien¹, S Roderick², L Bell², J Booth^2,4, PJ Keall¹, DT Nguyen^1,5, (1) ACRF Image X Institute, The University of Sydney, Eveleigh, NSW, AU, (2) Northern Sydney Cancer Centre, Royal North Shore Hospital, NSW, AU, (3) Prince of Wales Hospital, Randwick, NSW, AU, (4) School of Physics, University of Sydney, NSW, AU, (5) School of Biomedical Engineering, University of Technology Sydney, Ultimo, NSW, AU

(Wednesday, 7/15/2020) 4:30 PM - 5:30 PM [Eastern Time (GMT-4)]

Room: Track 2

Purpose:
Patients with locally advanced prostate cancer require multiple targets to be treated simultaneously with radiotherapy. Independent motion of targets can result in inaccurate radiation delivery, however, there are currently no methods that can adapt to multiple targets. This study investigated the first experimental implementation of a multi-target MLC tracking method for locally advanced prostate cancer.

Methods:
A multi-target MLC tracking algorithm was developed and integrated on a linac with a real-time 3D motion estimation system utilizing the on-board kV imager. A pelvic phantom was placed on a motion platform with a 3mm posterior and 5mm inferior interfraction prostate shift. Three clinical treatment plans were delivered to the phantom using multi-target MLC tracking and no tracking, with three different prostate motions reproduced. Geometric error was calculated by evaluating the misalignment between the targets and aperture positions. End-to-end latency of the tracking system was measured by tracking sinusoidal motion and calculating the time shift between the target and aperture trajectories on EPID images.

Results:
The root-mean-square geometric error of the multi-target tracking algorithm was 0.5mm parallel and 1.3mm perpendicular to the MLC leaves for the prostate, and 0.7mm parallel and 0.0mm perpendicular to the leaves for the nodes. Geometric error was independent of the magnitude of motion (r<0.1). With the phantom setup to the prostate and treated without tracking, the geometric error was 2.5mm parallel and 1.9mm perpendicular to the leaves for the prostate, and 2.2mm parallel and 5.0mm perpendicular to the leaves for the nodes. The total latency of the motion-adaptation system was 730±20ms.

Conclusion:
This study demonstrated the first experimental implementation of a multi-target MLC tracking method to independently track a moving prostate and stationary nodes. The multi-target tracking method was able to reduce geometric error compared to standard clinical practice to provide superior dose coverage to both targets.

Funding Support, Disclosures, and Conflict of Interest: The authors acknowledge funding from Cancer Council NSW. PJK is an inventor of a licensed and unlicensed patent related to MLC technology.

Keywords

Not Applicable / None Entered.

Taxonomy

TH- External Beam- Photons: Motion management - intrafraction

Contact Email