Related Experiment Video
Updated: Jun 28, 2026
Dynamic Lung Tumor Tracking for Stereotactic Ablative Body Radiation Therapy
Published on: June 7, 2015
Iterative active deformational methodology for tumor delineation: Evaluation across radiation treatment stage and
D H Wu1, A D Shaffer, D M Thompson
1Department of Radiological Sciences, Oklahoma University Health Science Center, Oklahoma City, Oklahoma 73104, USA. dee-wu@ouhsc.edu
Purpose:
To introduce, implement, and assess an iterative modification to the active deformational image segmentation method as applied to cervical cancer tumors.
Materials And Methods:
A comparison by Jaccard similarity (JS) between this active deformational method and manual segmentation was performed on tumors of various sizes across preradiation, 3 weeks postradiation, and 6 weeks postradiation using a General Linear Mixed Model across 121 studies from 52 patients with Stage IIB-IV cervical cancers.
Results:
The deformable segmentation method produced promising levels of agreement including JS factors of 0.71+/-0.11 in the preradiation studies. The analysis illustrated a rate of improvement in JS with increasing tumor volume that differed between the preradiation and 6 weeks postradiation stage (P=0.0474). In the large preradiated tumors each additional cm3 of volume was associated with an increase or improvement in JS of 0.0008 (95% confidence interval [CI]: 0.0003, 0.0014). In the smaller postradiation tumors, each additional cm3 of volume was associated with a more robust improvement in JS of 0.0046 (95% CI: 0.0009, 0.0082).
Conclusion:
Agreement was strongly affected by tumor volume, and its performance was most impacted across volume in the later stages of radiation therapy. The deformation-based segmentation method appears to demonstrate utility for delineating cervical cancer tumors, particularly in the earliest stages of radiation treatment, where agreement is greatest.
