Exhibit Hall | Forum 6
Purpose: Normal expiration and inspiration breathhold scans are recommended by a robotic radiosurgery vendor to achieve submillimeter end-to-end tracking accuracy, but in some institutions 4DCT is preferred. Therefore, known fiducials were examined in rigid motion phantoms to evaluate apparent target deformation. Results are compared to clinical practice.
Methods: Two standard vendor-supplied ball-cube motion phantoms were scanned in three settings: static, moving at 12 breaths per minute (bpm) and at 20 bpm to cover the span of normal adult restful breathrate. Autodetected scanner parameters were used for the 20 bpm scan, while the table pitch was manually decreased to 0.04 for the 12 bpm scan. For all datasets, 4 fiducials in each of the 10 phases were fused to the rigid scan, and rigid-body-errors (RBE) were calculated for all 40 measured fiducial positions.
Results: 95% of RBE were submillimeter for the slow-breath test (38/40). Only 15% of RBE were submillimeter at the high end of normal adult breathing rate (6/40); half of the fiducials were blurred beyond their known length, and the rigid phantom appeared to deform. In contrast to 4D acquisition, in breathhold scans it is possible to fuse inhale fiducials to exhale fiducials to ensure minimal deformation throughout the breathing cycle. Visualization of the breathing waveform during the CT scan is helpful to ensure the breathholds are near the end of normal excursion. If any discrepancies are observed, the scans may be repeated before the patient leaves, avoiding treatment delays.
Conclusion: It is advisable to follow manufacturer recommendations to use breathhold scans, unless it could be quantitatively ensured that RBE of known fiducials in 4DCT remain submillimeter for a variety of target motions. Several methods are available to visualize breathing waveforms during CT scans to ensure representative breathholds are acquired.
Target Localization, Patient Movement, Radiosurgery