Author: Giorgi Archuadze, William P. Donahue, David Kanchaveli, Shih-Chi Lin, Sebastian Meyer, Grace Tang 👨🔬
Affiliation: Department of Medical Physics, Memorial Sloan Kettering Cancer Center 🌍
Purpose: Breathing traces captured at simulations are often used for deep-inspiration breath-hold (DIBH) treatments. However, systematic drift can occur in the traces that is unrelated to the patients’ breath-hold efficacy. This study quantifies the impact of couch sag on RGSC drift, examining their dependence on CT vendor/model and patient weight.
Methods: Couch sag and RGSC drift were measured on three CT/PETCTs: Philips Brilliance Big Bore (BBB) CT, Siemens SOMATOM go.Open Pro (SOM) CTs, and Siemens Biograph PET/CT. All were equipped with Varian RGSC couch-mounted camera, except for SOM, where the wall-mounted camera was installed and calibrated with the weighted calibration procedure. Couch sag was measured with a ruler using a self-leveling laser as a reference, with a 75kg phantom. Data was collected at multiple couch positions, with the initial position and longitudinal motion range emulating lung/abdomen DIBH scans. Similarly, RGSC drift was measured at the same locations, with a reflective marker box placed on the same phantom. For SOM, couch sag and RGSC drift were also measured with 50kg and 115kg. In addition, all measurements were repeated on SOM with a couch-mounted camera.
Results: No couch sag was observed on the PET/CT, while up to 8mm and 3.5mm were measured on BBB and SOM, respectively. Sag varied with weight from 2.5mm to 6.5mm on SOM. No RGSC drift was observed on BBB and PET/CT, but SOM exhibited a non-linear drift (with couch positions) of up to 2 – 3.5mm for 75kg -115kg. With the couch-mounted camera on SOM, drift reached 3mm while closely tracking the mechanical sag pattern.
Conclusion: Siemens SOMATOM go.Open Pro was the only CT that exhibited RGSC drift and could not be accounted for with the vendor-recommended alternative/weighted calibration. The magnitude of intrinsic/systematic drift introduces erroneous signals to lung/GI DIBH patient simulations with a narrow gating window.