Related Experiment Video
Updated: Jun 23, 2026
Dynamic Lung Tumor Tracking for Stereotactic Ablative Body Radiation Therapy
Published on: June 7, 2015
Tumor localization using cone-beam CT reduces setup margins in conventionally fractionated radiotherapy for lung
Anamaria R Yeung1, Jonathan G Li, Wenyin Shi
1Department of Radiation Oncology, University of Florida College of Medicine, Gainesville, FL, USA.
Purpose:
To determine whether setup margins can be reduced using cone-beam computed tomography (CBCT) to localize tumor in conventionally fractionated radiotherapy for lung tumors.
Methods And Materials:
A total of 22 lung cancer patients were treated with curative intent with conventionally fractionated radiotherapy using daily image guidance with CBCT. Of these, 13 lung cancer patients had sufficient CBCT scans for analysis (389 CBCT scans). The patients underwent treatment simulation in the BodyFix immobilization system using four-dimensional CT to account for respiratory motion. Daily alignment was first done according to skin tattoos, followed by CBCT. All 389 CBCT scans were retrospectively registered to the planning CT scans using automated soft-tissue and bony registration; the resulting couch shifts in three dimensions were recorded.
Results:
The daily alignment to skin tattoos with no image guidance resulted in systematic (Sigma) and random (sigma) errors of 3.2-5.6 mm and 2.0-3.5 mm, respectively. The margin required to account for the setup error introduced by aligning to skin tattoos with no image guidance was approximately 1-1.6 cm. The difference in the couch shifts obtained from the bone and soft-tissue registration resulted in systematic (Sigma) and random (sigma) errors of 1.5-4.1 mm and 1.8-5.3 mm, respectively. The margin required to account for the setup error introduced using bony anatomy as a surrogate for the target, instead of localizing the target itself, was 0.5-1.4 cm.
Conclusion:
Using daily CBCT soft-tissue registration to localize the tumor in conventionally fractionated radiotherapy reduced the required setup margin by up to approximately 1.5 cm compared with both no image guidance and image guidance using bony anatomy as a surrogate for the target.
