Publication / RCMCI
Published Date: 2018/07/09
Published By: Dr Mohammad Reza Ay
Published At: Journal of Instrumentation
Attachment Files:
Published URL:

Sheikhzadeh, P., Sabet, H., Ghadiri, H., Geramifar, P., Ghafarian, P., Ay, M.R.


Introduction: In recent years, different geometries of brain-dedicated PET scanners have been developed due to the fact that there are great emerging demands for high-quality and low-cost brain-dedicated scanners. To meet the demand, we present a novel conceptual design based on a partial cylindrical detector geometry and evaluate the performance of the designed scanner. Methods and materials: The proposed design was modeled with a ring diameter of 245 mm and axial field of view (FOV) of 277 mm, thus covering the top and sides of patient's head. Detection efficiency of our new optimized design, which we refer to as HeadphonePET, was calculated against conventional cylindrical brain PET and whole-body positron scanners. Also, in order to measure the system spatial resolution and evaluate image quality, different simulated array sources and their related projections were extracted. Finally the obtained projections were reconstructed using analytical and iterative algorithms. Results: The simulation results indicate that the normalized mean efficiency of the HeadphonePET, with 30% less used crystals, can provide 1.2% higher sensitivity with respect to conventional cylindrical brain PET. However, it was demonstrated that the HeadphonePET, due to its partial coverage nature, cannot provide better spatial resolution when using analytical reconstruction algorithms. We obtained much-improved results when we used iterative reconstruction algorithms (1.61 mm FMHM for the HeadphonePET, versus 1.16 mm FWHM for the same full-ring scanner). Conclusion: The HeadphonePET as a novel conceptual design, with about 70% number of detectors with respect to a conventional full-ring cylindrical scanner, can achieve higher detection efficiency and comparable image quality (on the level of resolution of point sources). This proposed low-cost, compact and open view design, inducing no patient fear, possesses a big potential for brain imaging for different patient positions. © 2018 IOP Publishing Ltd and Sissa Medialab.