Abstract

Purpose:To investigate prostate imaging onboard radiation therapy machines using a novel robotic, 49‐pinhole Single Photon Emission Computed Tomography (SPECT) system.Methods:Computer‐simulation studies were performed for region‐of‐interest (ROI) imaging using a 49‐pinhole SPECT collimator and for broad cross‐section imaging using a parallel‐hole SPECT collimator. A male XCAT phantom was computersimulated in supine position with one 12mm‐diameter tumor added in the prostate. A treatment couch was added to the phantom. Four‐minute detector trajectories for imaging a 7cm‐diameter‐sphere ROI encompassing the tumor were investigated with different parameters, including pinhole focal length, pinhole diameter and trajectory starting angle. Pseudo‐random Poisson noise was included in the simulated projection data, and SPECT images were reconstructed by OSEM with 4 subsets and up to 10 iterations. Images were evaluated by visual inspection, profiles, and Root‐Mean‐ Square‐Error (RMSE).Results:The tumor was well visualized above background by the 49‐pinhole SPECT system with different pinhole parameters while it was not visible with parallel‐hole SPECT imaging. Minimum RMSEs were 0.30 for 49‐pinhole imaging and 0.41 for parallelhole imaging. For parallel‐hole imaging, the detector trajectory from rightto‐ left yielded slightly lower RMSEs than that from posterior to anterior. For 49‐pinhole imaging, near‐minimum RMSEs were maintained over a broader range of OSEM iterations with a 5mm pinhole diameter and 21cm focal length versus a 2mm diameter pinhole and 18cm focal length. The detector with 21cm pinhole focal length had the shortest rotation radius averaged over the trajectory.Conclusion:On‐board functional and molecular prostate imaging may be feasible in 4‐minute scan times by robotic SPECT. A 49‐pinhole SPECT system could improve such imaging as compared to broadcross‐section parallel‐hole collimated SPECT imaging. Multi‐pinhole imaging can be improved by considering pinhole focal length, pinhole diameter, and trajectory starting angle.The project is supported by the NIH grant 5R21‐CA156390.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call