Heidelberg 2022 – wissenschaftliches Programm

Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

ST: Fachverband Strahlen- und Medizinphysik

ST 9: Detectors and Applications II

ST 9.2: Vortrag

Donnerstag, 24. März 2022, 14:15–14:30, ST-H4

Technical aspects of simulation-based scatter correction in total-body Positron Emission Tomography (PET) imaging using the uEXPLORER PET scanner — •Reimund Bayerlein, Edwin K. Leung, Zhaoheng Xie, Eric Berg, Benjamin A. Spencer, Negar Omidvari, Qian Wang, Lorenzo Nardo, Simon R. Cherry, and Ramsey D. Badawi — University of California Davis

Positron Emission Tomography (PET) is a powerful tool for molecular imaging and has brought enhancements to biological research with widespread oncological and clinical adoption. The limited sensitivity of conventional PET scanners with a short axial field of view (AFOV) has been overcome by the uEXPLORER total-body PET scanner with a total AFOV of 194 cm. With a 15-68-fold increase in sensitivity and a spatial resolution of 3.0 mm the uEXPLORER can provide improved image quality, or reduced scan duration, or reduced radioactivity in the subject, or late time point imaging, or some combination of these. This state-of-the-art PET scanner constitutes a paradigm shift in nuclear medicine with the ability to address open questions in medicine and biology. However, the large number of detectors and the widened acceptance angle dramatically increase the data sizes, setting higher demands on the image reconstruction algorithms. Specifically, quantitative techniques for the correction of scattered events become more complex and computationally expensive. Due to the high number of lines of response (92 x 10⁹) scatter correction by direct computation using the Klein-Nishina formula is challenging in total-body PET and Monte-Carlo (MC) methods are preferred. In the spirit of a technical note, this contribution will describe the procedure of scatter correction in total-body PET using MC simulations embedded in a framework using a list-mode ordered subset expectation maximization image reconstruction. The method was developed and validated using phantom studies conducted at the EXPLORER Molecular Imaging Center at UC Davis. In the presentation, mathematical, physical, and computational aspects will be highlighted.