Encrypted login | home

Program Information

Using Multiple Cameras to Improve Image Reconstruction of Prompt Gamma Emission During Proton Beam Radiotherapy

no image available
H Chen

H Chen1*, E Draeger1 , D Mackin2 , S Peterson2 , S Beddar2 , J Polf1 , (1) Department of Radiation Oncology, University of Maryland School of Medicine, Baltimore, MD, USA, (2) Department of Radiation Physics, University of Texas MD Anderson Cancer Center, Houston, TX, USA, (3) Department of Physics, University of Cape Town, Rondebosch 7700, South Africa


SU-F-601-8 (Sunday, July 30, 2017) 2:05 PM - 3:00 PM Room: 601

Purpose: Iterative reconstruction algorithms can reconstruct the prompt gamma (PG) emission using photons measured by a single Compton camera (CC) during proton therapy. Because these algorithms often weight more along the trajectory of the intersection of two cones, the reconstructed emission distribution is stretched out into an elliptical shape in the direction perpendicular to the camera surface. This reduces the accuracy of CC imaging in clinical applications, such as range verification, especially for beams oblique to the camera surface. We hypothesize that if two or more cameras are placed at different view/gantry angles, this artifact can be reduced.

Methods: Monte Carlo simulations were carried out to test the efficacy of combining multiple camera views. A clinical 150 MeV proton pencil beam irradiating a water phantom was simulated and all characteristic prompt gammas and positron annihilation gammas were tracked. Cameras were placed at 0 degrees (above the phantom), 30, 45, 60 and 90 degrees, respectively. Different camera combination scenarios were explored to study the effects on the reconstructed PG image. Specifically, for each case, 3D PG images were reconstructed and the slice at the Bragg peak was fitted to a 2D Gaussian, the major- and minor-axes are examined and their ratio calculated.

Results: As the angular span of the cameras increases from 0 to 90 degrees, the major-axis in the PG image, representing the unwanted elliptical spread toward the camera, decreases while the minor-axis remains almost unchanged. The axes ratio decreases accordingly. With two orthogonal cameras (spaced 90 degrees), the two axes are nearly identical meaning the artifact is eliminated. Including more cameras in between the orthogonal cameras does not further improve the reconstruction.

Conclusion: Using multiple cameras reduces the spread perpendicular to the camera surface in the reconstructed images. The best reconstruction results from using two cameras that are perpendicular.

Funding Support, Disclosures, and Conflict of Interest: This research is supported by NIH grant R01CA187416.

Contact Email: