Purpose: The CIRS ATOM anthropomorphic phantoms are comprised of tissue-equivalent materials with radiation attenuation properties that correspond to bone, lung, and soft tissue. Thus, these phantoms can be used to accurately measure radiation therapy (RT) doses. However, their use in contemporary RT treatment planning is limited because individual organs are not visually distinguishable on computed tomography (CT) images. In contrast, the University of Florida/National Cancer Institute (UF/NCI) reference phantoms have well-defined organs but are virtual computational phantoms that cannot be used for measurements. The objectives of this study were to  demonstrate the feasibility of integrating organ contours from a UF/NCI computational phantom into a CT scan of a physical CIRS anthropomorphic phantom and  illustrate the clinical utility of this hybrid phantom through end-to-end testing.
Methods: A CT scan of the CIRS adult male phantom was acquired and imported into the RayStation treatment planning system. After contouring the femoral heads, it was co-registered with the UF/NCI adult male phantom. Then UF/NCI phantom organ contours were individually reverse mapped/integrated into the CIRS phantom. The integrated contours and their positioning were evaluated and modified by a radiation oncologist to create a hybrid phantom. This hybrid phantom was used to create a volumetric modulated arc therapy (VMAT) plan for high-risk prostate cancer following standard-of-care dosimetric objectives. Thermoluminescent dosimeters (TLDs) were inserted in select sections of the CIRS phantom corresponding to organs of interest. The VMAT plan was delivered to the phantom. Finally, TLDs were analyzed following a standard protocol.
Results: We successfully developed a hybrid phantom and created and delivered a VMAT plan that met RT-planning objectives. Good agreement between measured and calculated doses was observed, within 3 cGy on average.
Conclusion: The hybrid phantom approach can be used to create realistic contemporary RT-plans, which can be validated with physical phantom measurements.