Author: Mubasheer Chombakkadath, Tara E. Tyson, Iris Z. Wang 👨🔬
Affiliation: Roswell Park Comprehensive Cancer Center, University at Buffalo (SUNY) 🌍
Purpose: Deformable image registration (DIR) is critical in adaptive radiation therapy (ART). Existing DIR phantoms either simulate tumor shape or volume changes but lack comprehensive motion simulation. This project aims to design a phantom capable of evaluating DIR performance under respiratory motion scenarios, with the inclusion of Gafchromic film to assess ART dosimetry accuracy.
Methods: Using the commercially available QUASAR™ Phantom (pRESP, Modus QA, London, Ontario) as a base, we designed a cylindrical insert, 3D-printed using a Pulse E-111 printer with 1.75-mm PLA filament, to simulate tumor and organs-at-risk (OARs) with customizable position and shape variations. The insert dimensions (diameter: 3.142”, length: 7.086”) enable a 4/9th to 5/9th cross-sectional division for tumor position alternation. Removable cores within the cylindrical insert allow flexible simulation of tumor shrinkage and shape changes. To test the coherence of the phantom design, its fit within the QUASAR™ Phantom, and DIR functionality, we initially used locally available materials with different densities for the tumor and the OAR. The prototype was scanned on a CT simulator with 1.25-mm slice thickness for two different cylindrical cores in two orientations.
Results: The phantom design was successfully implemented, with precise assembly and secure fitting into the Quasar™ phantom. Visual DIR assessments using RADformation’s DIR software (NY, USA) on stationary CT images confirmed its functionality. Future evaluations will incorporate a better tumor substitute material, different cylindrical cores for tumor shape and volume changes. It will be implemented in the clinic for evaluating DIR algorithms and ART workflow. Quantitative evaluation metrics will include Dice Similarity Index (DSI) for DIR, and 2D film dosimetry Gamma-index analysis.
Conclusion: This innovative phantom design bridges the gap in current DIR testing by incorporating respiratory motion simulation. It provides a platform for assessing DIR performance under realistic clinical scenarios, potentially enhancing ART program quality.