SU-E-J-139: One Institution's Experience with Surface Imaging in Proton Therapy

Authors


Abstract

Purpose:

X-ray system is commonly used for IGRT in proton therapy, however image acquisition not only increases treatment time but also adds imaging dose. We studied a 3D surface camera system (AlignRT) performance for proton therapy.

Methods:

System accuracy was evaluated with rigid phantom under two different camera location configurations. For initial clinical applications, post mastectomy chest wall and partial breast treatments were studied. X-ray alignment was used as our ground truth. Our studies included: 1) comparison of daily patient setup shifts between X-ray alignment and SI calculation; 2) interfractional breast surface position variation when aligning to bony landmark on X-ray; 3) absolute positioning using planning CT DICOM data; 4) shifts for multi-isocenter treatment plan; 5) couch isocentric rotation accuracy.

Results:

Camera locations affected the system performance. After camera relocation, the accuracy of the system for the rigid phantom was within 1 mm (fixed couch), and 1.5 mm (isocentric rotation). For intrafractional patient positioning, X-ray and AlignRT shifts were highly correlated (r=0.99), with the largest difference (mean ± SD) in the longitudinal direction (2.14 ± 1.02 mm). For interfractional breast surface variation and absolute positioning, there were still larger disagreements between the two modalities due to different focus on anatomical landmarks, and 95% of the data lie within 5mm with some outliers at 7 mm–9 mm. For multi-isocenter shifts, the difference was 1 ± 0.56 mm over an 11 cm shift in longitudinal direction. For couch rotation study, the differences was 1.36 ± 1.0 mm in vertical direction, 3.04 ± 2.11 mm in longitudinal direction, and 2.10 ± 1.66 mm in lateral direction, with all rotation differences < 1.5 degree.

Conclusion:

Surface imaging is promising for intrafractional treatment application in proton therapy to reduce X-ray frequency. However the interfractional discrepancy between the X-ray and SI requires future validation.

Ancillary