Program Information
Improvement of Tumor Localization in Cone Beam CT Using An Optical Tracking System
B Barraclough1*, J Li2 , C Park2 , C Liu2 , S Lebron1 , G Yan2 , (1) J. Crayton Pruitt Family Department of Biomedical Engineering, University of Florida, Gainesville, FL, (2) Department of Radiation Oncology, University of Florida, Gainesville, FL
Presentations
SA-B-BRD-6 (Saturday, March 5, 2016) 10:30 AM - 12:30 PM Room: Grand Ballroom D
Purpose: In cone-beam CT (CBCT) guided external radiotherapy, baseline drifts in respiration-induced tumor motions negatively affect CBCT image quality and tumor localization accuracy. The purpose of this work is to use an in-house optical tracking system (OTS) to detect baseline drifts and develop correction strategies.
Methods: The OTS consists of a pair of Polaris CCD cameras. It tracks an infra-red reflective marker affixed on patients at 15 Hz throughout the treatment. State vectors in 2D augmented state space were formed with adjacent OTS signals. The vectors within a moving window were fitted to an elliptical shape prior, from which the centers of motion and baseline drifts were monitored. Individual CBCT projections acquired prior to baseline drifts were shifted accordingly in the 2D projection space to compensate for the drifts. The FDK algorithm was used to reconstruct the 3D CBCT volume data. A 4D motion phantom with a built-in spherical tumor (3 cm diameter) was used to validate the algorithm. A regular sinusoidal pattern (2 cm peak-to-peak amplitude, 5 sec period) was first simulated; then a 0.5 cm baseline drift was introduced at different stages of the CBCT scan (1/3, 2/3 way through and near end). The reconstructed tumor position was compared among the four breathing patterns.
Results :Without correction, the reconstructed tumor was respectively 0.21, 0.33 and 0.42 cm away from the reference position. The later the drifts occurred in the scan, the larger the deviation in the tumor position. After correction, the deviation reduced to 0.08, 0.04 and 0.04 cm, respectively.
Conclusion: Baseline drift in tumor motion during CBCT scanning negatively impacts tumor localization accuracy. Aided by an OTS, the proposed correction strategy can significantly improve the accuracy. Future work will examine its efficacy with different breathing patterns and drifts as well as its clinical impact in 4D CBCT.
Contact Email: