Total-Body PET 2024, September 19-21, 2024
Design optimization and performance evaluation of a sparse Walk-Through PET
Maya Abi-Akl, Jens Maebe, Boris Vervenne, Meysam Dadgar, Othmane Bouhali, Stefaan Vandenberghe
​
Abstract
Introduction: With the recent advancement in PET instrumentation, several long
cylindrical Axial Field-Of-View (AFOV) systems (70-195 cm) have been introduced.
However, due to their high acquisition/maintenance cost, such high-end scanners have
limited implementation in standard nuclear medicine departments. A medium-cost long
AFOV system with very eMicient patient throughput and high-resolution monolithic LYSO
detectors with depth-of-interaction (DOI) information has been proposed [1,2]. Patients
walk into the so-called Walk-Through PET (WT-PET) system, position themselves in
standing position and get a scan in less than one minute. The full design consists of two flat
panels of 71x106 cm2 in size placed 50 cm apart for patient positioning. Our first prototype
will have a reduced AFOV of the full system with axially sparse geometry. Each panel
consists of 8 rows of monolithic LYSO detectors of 50x50x16 mm3 with an axial gap of 25
mm (1/2 detector) between the rows in the vertical position, bringing the AFOV to 57.5 cm.
The number of detectors is 2.5x lower in the sparse compared to full design.
Methods: In this work, we perform a Monte Carlo simulation of the reduced sparse design
in GATE with uniform panel motion in the vertical direction to cover the 106 cm range for
torso imaging. Sensitivity is evaluated with a 110 cm line source in 20 cm cylindrical
phantom. We simulate XCAT phantoms for the full and the moving sparse design to check
the eMect of reduced statistics.
Results: The volume sensitivity of the sparse design is 4.8x lower than the full system.
XCAT reconstructions show low count statistics in the brain and leg areas, indicating that
the panels should move non-linearly to scan these areas more slowly. The simulation will
further investigate the optimal motion, and a study will follow to evaluate the spatial
resolution and image quality.
17th International Meeting on Fully Three-Dimensional Image Reconstruction in Radiology and Nuclear Medicine, July 16-21, 2023
Abstract:
A new PET scanner design, the walk-through total-body PET, has recently been proposed. It consists of two flat, vertically placed, 74x106 cm^2 panels made up of monolithic BGO detectors. This new scanner design will offer a lower cost, high patient throughput alternative to existing total-body PET scanners. This study investigates the image reconstruction of the NEMA image quality phantom in two configurations of the scanner: fixed and rotating. In addition, two methods for CT-less attenuation correction in the system are explored: the use of a transmission source and estimation of the attenuation coefficients from the emission data itself. The fixed scanner configuration offers higher sensitivity, but streaking artifacts are observed due to the limited angle problem, which are solved in the rotating configuration. CT-less attenuation correction also proves to be feasible, with generally better image quality observed with the addition of a transmission source.