Room: AAPM ePoster Library
To build a QA program for 4D CT acquisition in accordance with CPQR guidelines for a CT Simulator equipped with Varian’s RPM system.
An in-house platform was used to test geometric accuracy of RPM marker block in the lateral, cranio-caudal and anterior-posterior directions. To determine quality metrics for 4DCTs, a Quasar’s respiratory motion phantom with a cedar insert containing a ‘tumour’ and a Catphan phantom placed on an in-house moving platform were scanned either static or set to move (±1cm with 4s period and ±4mm with 3.5s period respectively). 4DCTs were acquired with a thorax clinical protocol in cine mode, with 5.6s cine duration and 0.3s cine time between images. Reconstruction were performed with Advantage software (GE) resulting in 10 phase-binned CTs, a MIP and AVE CT. CTDI measurements were performed with a Fluke TNT-12000DoseMate system with 4D scan settings as above.
4DCT scan analysis included manual and 4D auto-contouring tools within Eclipse (Varian, v.13) and in-house software for Catphan image quality analysis. All 4D results were compared among 3 GE CT systems equipped with RPM, as well as against static, 3D helical acquisitions.
Geometric accuracy of ROI motion was accurately represented for a hypodense ROI within the cedar insert. Spatial resolution, as indicated by lp/cm was preserved in the CT0 (max inhale) and CT50 (max exhale) compared to a static scan. The CT:HU curve was unchanged in 4D scans as compared to static scans. The uniformity of HU within a specific region did not adhere to CPQR guidelines (Mean HU ±10 HU bin variability, ±10% from helical). There was no significant inter-scanner variability. CTDI values were consistent with scanner-reported values and within expected limits.
An annual 4DCT program was successfully implemented, adhering to CPQR guidelines. Further interrogation of HU noise in 4D scans is warranted.