Related Experiment Video
Updated: Mar 12, 2026
07:34
Utilizing vmTracking to Improve the Accuracy of Multi-Animal Pose Estimation in Rodent Social Behavior Studies
Published on: November 7, 2025
411
Automatic detection of MLC relative position errors for VMAT using the EPID-based picket fence test
Damianos Christophides1, Alex Davies, Mark Fleckney
1Radiotherapy Physics, Level 1 Bexley Wing, St. James's Institute of Oncology, Beckett Street, Leeds LS9 7TF, UK. University of Leeds, Leeds Institute of Cancer and Pathology, Leeds, UK.
Physics in Medicine and Biology
|November 8, 2016
Summary
Automating multi-leaf collimator (MLC) quality assurance using electronic portal imaging picket fence tests detects errors ≥0.5 mm. This method provides early warnings, improving treatment accuracy and reducing clinical downtime.
Area of Science:
- Medical Physics
- Radiation Oncology
- Radiotherapy Quality Assurance
Background:
- Multi-leaf collimators (MLCs) are crucial for precise radiation delivery in advanced techniques like IMRT and VMAT.
- Current MLC quality assurance relies on qualitative assessments, which may not be sensitive enough for early error detection.
Purpose of the Study:
- To automate the detection of MLC relative position errors (≥0.5 mm) using electronic portal imaging device (EPID)-based picket fence tests.
- To compare the automated detection results with current qualitative assessment methods.
Main Methods:
- Weekly picket fence tests were performed on three Varian linacs over 14-20 months.
- An algorithm was developed to calculate MLC errors from picket fence images.
- Baseline error distributions were established and compared statistically with intentional errors.
Main Results:
- The automated method successfully distinguished between baseline and intentional MLC errors, showing no overlap in statistical distributions.
- The automated detection of MLC errors agreed with fault reports and interventions for MLC repair.
- The method provided early warning of MLC errors, correlating with clinical downtime.
Conclusions:
- The developed method enables fully automated MLC quality assurance tailored to individual linac performance.
- Automated detection offers early detection of MLC errors, enhancing treatment safety and reducing unplanned downtime.
