1028: Use of On-Board Imaging to Evaluate Residual Errors for Target Localization in SBRT: A Feedback Analysis

Abstract

Accurate target localization for stereotactic body radiotherapy (SBRT) requires identification of 3D anatomical information. On-board imaging provides both 2D radiographs and cone beam CT (CBCT) for target localization and is being introduced for clinical applications. However, the effective use of this system is still under development. This study is aimed to develop a new scheme of using on-board imaging as a feedback system to evaluate residual localization errors for SBRT due to both inter-fraction variation between simulation and treatment and intra-fraction patient movement associated longer treatment time. The process for on-board target localization for SBRT includes: 1) align patient within immobilization device; 2) obtain orthogonal 2D kV/MV radiographs to compare to reference DRRs and document shifts accordingly; 3) obtain 3D CBCT for comparison to the planning CT; 4) shift patient to align target between CBCT and planning CT; 5) obtain orthogonal 2D kV/MV radiographs to document shifts; 6) treatment; 7) obtain orthogonal 2D kV/MV radiographs to assess intra-fraction motion. The shifts as identified by steps 2 and 3 indicate the improvement of target localization accuracy from inter-fraction deviation by incorporating 3D anatomical information. Steps 5 and 7 reflect the degree of intra-fraction variation which provides feedback information about the treatment accuracy. We analyzed localization data for 17 treatments from 10 patients treated with SBRT during last 6 months. The localization accuracy as identified in step 3 is illustrated in the figure, where CBCT target is overlaid by the planning target contour. Its accuracy is visually judged within 1 mm. The figure also shows the residual errors of target localization based on CBCT after 2D radiographic corrections for each treatment. The intra-fraction movement assessed via 2D radiographs prior to and after treatment ranged from 0 to 4 mm with an average of 1 mm, indicating further improvement of patient immobilization may be needed. On the other hand, 3D imaging takes at least 3 times more working efforts compared to 2D imaging. The on-board imaging system provides a powerful tool to assess the overall accuracy of treatment. While CBCT is favorable for the identification of inter-fraction deviation, the intra-fraction errors could be effectively identified using 2D imaging method. The residual positioning errors identified by on-board imaging provide useful feedback information for improving patient treatment accuracy.