Author: Ryan Andosca, Rojine T. Ariani, Peter Boyle, Minji Victoria Kim, Michael Vincent Lauria, Daniel A. Low, Claudia R. Miller, Drew Moghanaki, Louise Naumann, Dylan P. O'Connell, Ricky R Savjani 👨🔬
Affiliation: Department of Radiation Oncology, University of California, Los Angeles, University of California, Los Angeles, UCLA Radiation Oncology 🌍
Purpose: To demonstrate that 5DCT can provide an accurate internal tumor volume (ITV) while 4DCT cannot.
Methods: The 5DCT imaging protocol uses a motion model and 25 deformably registered free-breathing CTs to generate patient images at arbitrary breathing phases with measured accuracy and precision. It’s then used to model tumor motion over user-specified breathing amplitudes from which an ITV can be created for radiation therapy treatment planning (TP). Under IRB approval, we acquired sequential 4D and 5D scans from which ITVs were generated to demonstrate the overall benefit of 5DCT against commercial 4DCT. The built-in 5DCT workflow accuracy analyses justified its use as the standard against which 4DCT was compared. Patients were scanned for both protocols using a Siemens SOMATOM Drive dual-source CT scanner. The two ITVs were defined by an expert on a treatment gated free-breathing scan using both the typical 4DCT breathing phase images and the 5DCT images. To compare the contours, we computed dice similarity coefficients (DSC), volume percent differences (VPD), and Hausdorff distance (HD). We also quantified cases in which the 4DCT protocol failed and was unusable in TP.
Results: Of 4 patients, 2 had irregular breathing causing 4DCT failure. Of the remaining patient cases, patient 2 had a DSC of 0.90, VPD of 17.79%, and HD of 2.31 mm between the 4D and 5D imaging. Patient 3 had a DSC of 0.94, VPD of –0.35%, and an HD of 2.01 mm. In all analyzed cases, the 5D protocol was used for clinical ITV generation in TP.
Conclusion: 5DCT's ability to quantify its precision and accuracy allowed us to use it as a standard against which the 4DCT protocol was compared. Current quantifiable discrepancies suggest 5D may provide more accurate ITV visualization. Further data collection is ongoing to more thoroughly compare the two imaging modalities.