Investigating the limits of PET/CT imaging at very low true count rates in Ion-Beam Therapy monitoring

4D offline PET-based treatment verification in scanned ion beam therapy: a phantom study

“…The former arise due to inaccuracies in the MC modeling of the underlying nuclear interactions (Böhlen et al 2010), as well as due to inaccuracies by the approximate modeling of the image acquisition and reconstruction process in the MC data post-processing, using only a post-simulation Gaussian filter (see equation ( 1)). On the measurement side, image noise in the order of typically a few percent has to be considered, just as the intrinsic accuracy in activity quantification by the PET/CT scanner (about 2%-5%) (Kurz et al 2013) and, eventually, inaccuracies due to the applied attenuation correction on the basis of a free-breathing 3D CT during 4D gated PET image reconstruction. In the latter aspect, attenuation correction on the basis of a time-averaged 3D CT or phase-correlated 4D CT might improve the results, but is currently not available on the Biograph mCT scanner.…”

“…60 (2015) 6227 considerably more challenging due to the strongly enhanced image noise. Improvements are, however, foreseen by the optimization of the image reconstruction parameters tailored to the encountered low counting statistics (Kurz et al 2013), or by the application of dedicated 4D image reconstruction algorithms making use of the entire recorded LM data, rather than subdividing into separate phases at further reduced counts for reconstruction (Gianoli et al 2014).…”