Program Information
Low-Fluence Proton Radiography and CT for In-Room Patient Alignment and Range Verification
T.E. Plautz1*, V.A. Bashkirov2 , R.P. Johnson3 , C. Ordonez4 , H. F.-W. Sadrozinski3 , R.W. Schulte1,2 , (1) University of California, San Francisco, San Francisco, CA, (2) Loma Linda University, Loma Linda, CA, (3) University of California, Santa Cruz, Santa Cruz, CA, (4) Northern Illinois University, Dekalb, Illinois
Presentations
WE-G-605-7 (Wednesday, August 2, 2017) 4:30 PM - 6:00 PM Room: 605
Purpose: To describe a novel algorithm for low-fluence particle radiography and CT and to evaluate its potential use for in-room patient alignment and range verification at the time of treatment.
Methods: An algorithm based on the most-likely path (MLP) concept was developed. It estimates the column voxel sums of water equivalent thickness (WET) of a discretized object using tracking and water equivalent path length (WEPL) data of individual particles. The algorithm was optimized and tested by using proton data provided by a prototype particle computed tomography (pCT) scanner. Proton data (200 MeV) of a homogeneous polystyrene step phantom and a pediatric head phantom were acquired, MLPs were reconstructed and WEPL voxel histograms of all protons intersecting a given voxel were summed along individual columns, renormalized, and their mode was calculated using a robust mode estimation method. The individual modes of voxel columns formed a pixelated radiograph for a given projection. For CT reconstruction, projection images were combined for 3D CT-reconstruction using a standard filtered back projection algorithm.
Results: The algorithm produced high-quality radiographs even for very low fluences of 50 protons/ mm²-pixel (~5 µGy). Testing of fast FBP-based reconstruction using projections obtained with this algorithm is in progress using quality assurance phantoms of the Catphan 600 series. The algorithm’s performance will be compared to previously established proton CT reconstruction algorithms. Evaluation of patient registration and proton range verification using the proton images with a modified pediatric head phantom outfitted with radiochromic film is underway. The residual alignment and residual dose errors will be evaluated using γ-index analysis.
Conclusion: A new algorithm has been developed that promises improved performance and efficiency of particle radiography and CT at very low particle fluences. This algorithm is suitable for fast, low-dose patient alignment and range verification for proton therapy at the time of treatment.
Contact Email: