Abstract
PURPOSE: The drift signal due to CT couch sagging introduces intrinsic noise into respiratory signals, impairing the accuracy of tumor motion management. The purpose of this work is to present an enhanced calibration method to reduce calibration uncertainty and mitigate baseline drift caused by table sagging for the wall-mounted Respiratory Gating for Scanner (RGSC) camera. METHODS: A weight of approximately 70 kg, simulating a patient's weight, was distributed on the CT table. The external ceiling laser light was adjusted laterally by ±10 cm to align three reflector blocks sequentially with the laser at predefined positions, ensuring accurate placement of the blocks at their corresponding positions. The blocks were then moved to the internal laser plane using the CT console. Subsequently, calibration measurements were performed at nine points at the combination of three lateral positions (the CT isocenter and ±10 cm laterally from the isocenter) and three longitudinal positions (the CT isocenter and ±15 cm longitudinally from the isocenter), by occluding the other two blocks and moving the couch longitudinally. For comparison, the Varian calibration method was also implemented. RESULTS: The block positioning uncertainty was reduced from the millimeter level to the sub-millimeter level. For a typical 40 cm scan length of DIBH, the residual baseline drift was significantly (p-value<0.001) mitigated from 2.84 ± 0.22 mm to 0.64 ± 0.06 mm. CONCLUSION: The proposed calibration method provides a robust solution to minimize block positioning uncertainty and reduce baseline drift caused by CT couch sagging, enhancing the repeatability and accuracy of the wall-mounted camera calibration. Its versatility for wall- and ceiling-mounted cameras further expands its potential clinical utility.