A Three-Dimensional Thermoplastic Prostate Phantom for Evaluation of Deformable Image Registration
C Reber1,2, A Neo1,2, E Schoenhoff1,2, N Kirby2*, K Singhrao2, J Pouliot2, (1) University of California Berkeley, Berkeley, CA,(2) University of California San Francisco, San Francisco, CATU-C-141-10 Tuesday 10:30AM - 12:30PM Room: 141
Purpose: Proof-of-concept of a three-dimensional (3D) pelvic phantom for evaluation of deformable image registration algorithms.
Methods: The phantom represents a cylindrical section of the human pelvic anatomy, modeled after human CT scan data, and is intended for use as an insert in a Modus QUASAR phantom. Polyurethane plastic was used to represent both fat and muscle sections with additives used to achieve appropriate Hounsfield units (HU) for the fat regions. The phantom consists of two halves, manufactured using silicon molds, which are placed together when the phantom is deformed. Keying on the inner surfaces was used to ensure proper alignment of the two halves. A grid of non-radiopaque markers was printed on the surface of each half-cylinder. We created two identical phantom sets, then deformed one. Optical imaging was used to characterize marker position pre- and post-deformation to calculate the 2D deformation. These can then be superimposed on laser scan profiles of the surface to measure the deformation in the third dimension. Identical markers were printed on both phantom halves to evaluate the effectiveness of the keys.
Results: We achieved deformations up to 1.6 mm. The keys remained intact through the deformation process. The mean distance to agreement between the markers on the two halves of the phantoms was 0.28 mm. The deformation field centered around the keys suggests that the keys were effective in keeping the phantom together.
Conclusion: We have demonstrated feasibility of a three-dimensional deformable phantom for evaluation of DIR algorithms. Future tests will be performed to further characterize and optimize performance of the keys. Development of this 3D phantom will enable improved assessment of DIR algorithm accuracy, providing quality assurance and optimization of DIR algorithm usage in the clinical setting.
Funding Support, Disclosures, and Conflict of Interest: This work was supported in part by a UC Proof of Concept Program Commercialization Gap Grant.