Researcher(s)
- Christopher Cephas, Biological Sciences, University of Delaware
Faculty Mentor(s)
- Hank Chen, ChristianaCare, University of Delaware
- Mojtaba Moazzezi, ChristianaCare, University of Delaware
Abstract
Adaptive radiation therapy is a dynamic cancer treatment protocol that updates treatment plans in anticipation of tumor and critical organ changes. Advanced image processing can generate synthetic CT images from cone-beam CT (CBCT) datasets, enabling dose calculation, dose tracking, and assessment of anatomical changes. This study evaluates an efficient adaptive workflow utilizing a synthetic (virtual) CT generated from CBCT for treatment planning, with the goal of reducing the need for repeat CT (re-CT) simulation in appropriately selected patients. Included patients underwent a re-CT during radiotherapy and had a CBCT acquired on the same day with corresponding anatomical changes. A virtual CT was generated from each CBCT and used for adaptive planning. Three representative cases were evaluated: an esophageal cancer patient with interval pleural effusion, a left chest wall patient with substantial tumor regression, and a right breast patient with treatment-related breast edema. Dose distributions and clinical goals from plans generated on the virtual CT were compared with those from the corresponding re-CT. The proposed workflow demonstrated feasibility across clinically relevant adaptive scenarios, with good agreement in cases with high-quality CBCT images while also illustrating the impact of CBCT image quality, patient anatomy, positioning, and imaging artifacts on virtual CT accuracy. These findings support the potential of virtual CT–based adaptive planning in appropriately selected patients.