SU-E-J-79: Using the Distance of Closest Approach to Improve Compton Camera Image Quality.

Area of Science:

• Medical Physics
• Nuclear Imaging
• Radiation Oncology

Background:

• Prompt gamma radiation from proton therapy enables beam range verification.
• Compton cameras (CC) offer efficient 3D imaging but suffer from reduced resolution due to Doppler broadening and measurement inaccuracies.
• Stochastic Origins Ensemble (SOE) is an algorithm for image reconstruction in CC.

Purpose of the Study:

• To evaluate if filtering CC events based on distance of closest approach (DCA) threshold improves image resolution.
• To assess the impact of DCA filtering on images reconstructed with the SOE algorithm for proton therapy applications.

Main Methods:

• Simulated a 3-stage Compton camera for a 0.511 MeV point source and for prompt gammas during proton therapy.
• Reconstructed images using the SOE algorithm with all events, and with events filtered by DCA < 3 mm and DCA < 1 mm.
• Measured the point-spread-function (PSF) for the point source and compared 2D images for the proton therapy simulation.

Main Results:

• The full width at half maximum (FWHM) of the PSF was reduced by 65% when applying a DCA < 3 mm threshold.
• Image quality of reconstructed proton beam 2D images visibly improved as the DCA threshold was lowered.
• Filtering events with DCA < 3 mm significantly reduced resolution loss caused by detector effects.

Conclusions:

• Filtering Compton camera events using a distance of closest approach (DCA) threshold enhances image resolution.
• Removing events with significant resolution loss before reconstruction yields higher quality images.
• Methods for predicting DCA from scatter data can enable this technique with real detector data.