Encrypted login | home

Program Information

Tumor Motion Tracking with a Regional Deformable Registration Model for Four Dimensional Radiation Treatment of Lung Cancer

no image available
M Chao

M Chao1*, Y Lo1 , Y Yuan1 , R Sheu1 , K Rosenzweig1 , (1) The Mount Sinai Medical Center, New York, NY

Presentations

SU-D-18A-1 Sunday 2:05PM - 3:00PM Room: 18A

Purpose: To develop a tumor motion model from four-dimensional computed tomography (4DCT) of thoracic patients and demonstrate its impact on 4D radiation therapy simulation.

Methods: A regional deformable image registration algorithm was introduced to extract tumor motion out of patient's breathing cycle. The gross target volume (GTV) was manually delineated on a selected phase of 4DCT and a subregion with 10mm margin supplemented to the GTV was created on the Eclipse treatment planning system (Varian Medical Systems, Palo Alto, CA). Together with 4DCT the structures were exported into an inhouse research platform. A free form B-Spline deformable registration was carried out to map the subregion to other respiratory phases. The displacement vector fields were employed to propagate GTV contours with which the center of mass (CoM) of the GTV was computed for each breathing phase of 4DCT. The resultant GTV motion and its volumetric shape are utilized to facilitate 4D treatment planning. Five lung cancer patients undergoing stereotactic body radiation therapy were enrolled and their 4DCT sets were included in the study.

Results: Application of the algorithm to five thoracic patients indicates that clinically satisfactory outcomes were achievable with a spatial accuracy better than 2mm for GTV contour propagation between adjacent phases, and 3mm between opposite phases. The GTV CoM was found to be in the range of 2.0mm through 2.5cm, depending upon the tumor location. Compared to the traditional whole image based registration, the computation of the regional model was found to be an order of magnitude more efficient.

Conclusion: A regional deformable registration model was implemented to extract tumor motion. It will have widespread application in 4D radiation treatment planning in the future to maximally utilize the available spatial-tempo information.


Contact Email: