Assessment of Image Quality In Hypersight CBCT and Conventional CBCT Images Using Planning CT As the Reference through Radiomic Features
Abstract
Purpose
This study aims to illuminate how HyperSight CBCT compares to conventional CBCT and planning CT in terms of radiomic features, applying statistical methods such as the Kruskal-Wallis H Test and the Intra-Class Correlation Coefficient (ICC).
Methods
We conducted a retrospective analysis of 10 cancer patients (six thoracic, two pelvic, and two abdominal treatment sites) who received radiation therapy, each undergoing three imaging techniques: planning CT, HyperSight CBCT, and conventional CBCT, with the first on-treatment scan used for consistency across techniques. The images from both HyperSight and conventional CBCT were registered to the planning CT using a multi-resolution, intensity-based mutual information framework. To ensure uniformity in assessment, all images were standardized by resampling to 1-mm isotropic voxels. Radiomic features were extracted from the Clinical Target Volume (CTV) using the PyRadiomics software. We performed our statistical analyses using SPSS, focusing on evaluating the degree of agreement with planning CT using ICC calculations and on determining the significance of differences among the three imaging techniques.
Results
The findings reveal a statistically significant difference in radiomic features between conventional CBCT and HyperSight CBCT (P-value = 0.03), as well as between conventional CBCT and planning CT (P-value < 0.05). However, no significant differences were observed between HyperSight and planning CT (P-value = 0.63). The ICC calculations indicate poor to moderate agreement between conventional CBCT and planning CT (0.03 < ICC < 0.54) and a moderate to excellent level of agreement between HyperSight and planning CT (0.5 < ICC < 0.95).
Conclusion
The results suggest that the image quality of HyperSight CBCT is comparable to that of planning CT for radiomic feature extraction, supporting its promise for quantitative CBCT imaging. Our findings open the possibility of clinical translation and adoption of CBCT-based radiomics for outcome/toxicity assessment and potential mid-treatment adaptation.