Related Experiment Video
Updated: Sep 2, 2025
08:17
Dynamic Lung Tumor Tracking for Stereotactic Ablative Body Radiation Therapy
Published on: June 7, 2015
15.8K
Erring Characteristics of Deformable Image Registration-Based Auto-Propagation for Internal Target Volume in
Benjamin J Rich1, Benjamin O Spieler1, Yidong Yang2
1Department of Radiation Oncology, University of Miami, Miami, FL, United States.
Frontiers in Oncology
|August 8, 2022
Summary
Deformable image registration (DIR) auto-propagation for internal gross tumor volume (IGTV) generation in lung cancer radiotherapy shows errors dependent on tumor location and deformation. This highlights the need for personalized internal target volume design in radiotherapy for locally advanced non-small cell lung cancer (LA-NSCLC).
Area of Science:
- Medical physics
- Radiation oncology
- Radiotherapy planning
Background:
- Respiratory motion in locally advanced non-small cell lung cancer (LA-NSCLC) complicates radiotherapy (RT) targeting.
- Internal gross tumor volume (IGTV) is often generated using deformable image registration (DIR)-based auto-propagation from single-phase 4DCT scans.
- Assessing the accuracy of DIR-based auto-propagation for IGTV generation is crucial for improving RT targeting.
Purpose of the Study:
- To evaluate the accuracy of DIR-based auto-propagation for generating internal gross tumor volumes (IGTVs) in LA-NSCLC patients.
- To identify characteristics associated with errors in DIR-based IGTV auto-propagation.
- To enhance radiotherapy targeting for LA-NSCLC by understanding IGTV generation inaccuracies.
Main Methods:
- Acquired 4DCT scans from 19 LA-NSCLC patients with 10 respiratory phases (RPs).
- Generated ground-truth IGTVs (GT-IGTVs) via manual segmentation and union of GTVs across all phases.
- Performed IGTV auto-propagation using two DIR algorithms on manually contoured GTVs from each RP, assessing differences (DICE, MSSD, ASSD, PVD) relative to GT-IGTVs, tumor location, RP, and deformation index (DI).
Main Results:
- No significant differences in accuracy metrics (DICE, MSSD, ASSD, PVD) were found across the 10 RPs for either DIR algorithm.
- MSSD and ASSD varied significantly with tumor location (central-peripheral, superior-inferior), and PVD varied with central-peripheral location.
- Accuracy metrics (DICE, MSSD, ASSD) significantly correlated with the deformation index (DI), and 79% of RT plans based on auto-propagated IGTVs failed prescription constraints.
Conclusions:
- Errors in DIR-based IGTV propagation for LA-NSCLC vary and depend on DI and anatomical tumor location.
- These findings underscore the necessity for personalized considerations in designing RT internal target volumes.
- The study highlights potential dosimetric discrepancies when relying solely on auto-propagated IGTVs for radiotherapy planning.
