Publicly available framework for simulating and experimentally validating clinical PET systems

Abstract

Background: Monte Carlo (MC) simulations are a powerful tool to model medical imaging systems. However, before simulations can be considered the ground truth, they have to be validated with experiments. Purpose: To provide a pipeline that models a clinical positron emission tomography (PET)/CT system using MC simulations after extensively validating the results against experimental measurements. Methods: A clinical four-ring PET imaging system was modeled using Geant4 application for tomographic emission (v. 9.0). To validate the simulations, PET images were acquired of a cylindrical phantom, point source, and image quality phantom with the modeled system and the simulations of the experimental procedures. For the purpose of validating the quantification capabilities and image quality provided by the simulation pipeline, the simulations were compared against the measurements in terms of their count rates and sensitivity as well as their image uniformity, resolution, recovery coefficients (RCs), coefficients of variation, contrast, and background variability. Results: When compared to the measured data, the number of true detections in the MC simulations was within 5%. The scatter fraction was found to be 30.0% ± 2.2% and 28.8% ± 1.7% in the measured and simulated scans, respectively. Analyzing the measured and simulated sinograms, the sensitivities were found to be 8.2 and 7.8 cps/kBq, respectively. The fraction of random coincidences were 19% in the measured data and 25% in the simulation. When calculating the image uniformity within the axial slices, the measured image exhibited a uniformity of 0.015 ± 0.005, whereas the simulated image had a uniformity of 0.029 ± 0.011. In the axial direction, the uniformity was measured to be 0.024 ± 0.006 and 0.040 ± 0.015 for the measured and simulated data, respectively. Comparing the image resolution, an average percentage difference of 2.9% was found between the measurements and simulations. The RCs calculated in both the measured and simulated images were found to be within the EARL ranges, except for that of the simulation of the smallest sphere. The coefficients of variation for the measured and simulated images were found to be 12% and 13%, respectively. Lastly, the background variability was consistent

Respiratory signal estimation for cardiac perfusion SPECT using deep learning

“…In PET, DL based corrections for respiratory motion have primarily focused on image registrations between gated PET images, and were mostly based on simulated phantom studies. [33][34][35] Here, we present a DL approach for respiratory motion correction of cardiac perfusion SPECT by accurately estimating a respiratory signal from 100ms sub-projections reframed from list-mode acquisition. This proposed method was developed and evaluated using a large clinical study with a total of 900 patients.…”

“…For SPECT, DL has been proposed for SPECT attenuation correction, 26,27 image denoising, 28,29 image reconstruction, 30 lesion detection, 31,32 etc. In PET, DL based corrections for respiratory motion have primarily focused on image registrations between gated PET images, and were mostly based on simulated phantom studies 33–35 . Here, we present a DL approach for respiratory motion correction of cardiac perfusion SPECT by accurately estimating a respiratory signal from 100‐ms sub‐projections reframed from list‐mode acquisition.…”