Related Experiment Video
Updated: Apr 17, 2026
05:18
Author Spotlight: Improving Radiation Therapy Access with Radiation Planning Assistant
Published on: October 6, 2023
2.1K
Automated contouring error detection based on supervised geometric attribute distribution models for radiation
Hsin-Chen Chen1, Jun Tan1, Steven Dolly1
1Department of Radiation Oncology, Washington University, St. Louis, Missouri 63110.
Medical Physics
|February 6, 2015
Summary
This study introduces a novel strategy using geometric attribute distribution (GAD) models to automatically detect errors in radiation therapy contouring. The method significantly improves accuracy and efficiency in identifying organ-at-risk contouring mistakes.
Area of Science:
- Medical Physics
- Radiotherapy
- Image Analysis
Background:
- Accurate tumor and organ-at-risk (OAR) contouring is critical in radiation therapy.
- Manual and automated contouring methods suffer from inter- and intraobserver variability and limitations in imaging.
- Current verification processes are laborious, time-consuming, and not entirely error-free.
Purpose of the Study:
- To develop a general strategy for automatic detection of radiation therapy OAR contouring errors.
- To facilitate and improve the clinical workflow in radiation therapy treatment planning.
- To introduce novel geometric attribute distribution (GAD) models for error detection.
Main Methods:
- Established GAD models characterizing interstructural (centroid, volume) and intrastructural (shape) variations.
- Developed an iterative weighted GAD model-fitting algorithm for error detection.
- Utilized Receiver Operating Characteristic (ROC) analysis for parameter optimization on 44 head-and-neck cases.
Main Results:
- The strategy achieved high sensitivity and specificity in detecting centroid/volume related errors (avg. sensitivity 0.954/0.906, avg. specificity 0.901/0.909).
- Shape-related contouring errors were detected with avg. sensitivity of 0.816 and avg. specificity of 0.94.
- Demonstrated feasibility for low false detection rates and provided a 3D visualization tool for error detection.
Conclusions:
- The proposed strategy reliably identifies contouring errors using inter- and intrastructural constraints from approved contours.
- It holds significant potential for enhancing radiation therapy workflow efficiency and accuracy.
- Further improvements are planned by incorporating more training data and geometric constraints.
