Related Experiment Video
Updated: Nov 22, 2025
08:17
Dynamic Lung Tumor Tracking for Stereotactic Ablative Body Radiation Therapy
Published on: June 7, 2015
16.0K
Simultaneous Truth and Performance Level Estimation Method for Evaluation of Target Contouring in Radiosurgery
Helena SandstrÖm1, Iuliana Toma-Dasu2,3, Caroline Chung4
1Medical Radiation Physics, Department of Physics, Stockholm University, Stockholm, Sweden; helena.sandstrom@ki.se.
Anticancer Research
|January 9, 2021
Summary
The Simultaneous Truth and Performance Level Estimation (STAPLE) algorithm accurately estimates radiosurgery target volumes using multiple contours. Its robustness in determining true target volume depends on the number and variability of input contours.
Area of Science:
- Radiosurgery and Radiation Oncology
- Medical Imaging and Image Analysis
- Computational Biology and Bioinformatics
Background:
- Target delineation variability in Gamma Knife radiosurgery mirrors that in linac-based radiotherapy.
- Standardization is lacking in radiosurgery target contouring, leading to inconsistent results.
Purpose of the Study:
- To quantify contouring variability for five radiosurgery targets using the STAPLE algorithm.
- To estimate the true target volume based on multiple delineations.
- To assess the STAPLE method's robustness concerning the number of input contours.
Main Methods:
- A multicenter analysis involved 12 experienced Gamma Knife planners delineating five cases.
- The STAPLE algorithm was used to estimate true target volumes.
- Contour sets were randomly selected to evaluate STAPLE's robustness with varying numbers of inputs.
Main Results:
- High similarity (90-100%) between STAPLE-estimated true volume and 50%-agreement volume was achieved with all 12 contours.
- Lower similarity (10-70%) was observed when using fewer contours.
Conclusions:
- The STAPLE method is valuable for estimating true target volumes from multiple contours when a sufficient number of inputs are available.
- STAPLE provides sensitivity and specificity for each contour relative to the true volume.
- The method's robustness relies on the quantity and variability (shape, size, position) of the input contours.
