PET NEMA IQ phantom dataset for quantification study

PET imaging has the potential to produce quantitative images of tracer uptake using SUV index. Although SUV is widely available and convenient to use in clinical routine, it is significantly influenced by reconstruction parameters and VOI definition. Moreover, resolution recovery, as an advanced algorithm, dramatically affects PET quantification.

Our data consisted of raw and analyzed data. The presented quantitative dataset includes four Winzip folders of DCM phantom images and Excel spreadsheets for four SBRs. Six quantitative metrics, including RCmax, RC50%, RCpeak, SUVmax, SUVmean, and SUVpeak, and three volumetric indices of MTV, VRC, and TLG in different reconstruction settings were represented in these Excel files. Furthermore, the contrast to noise ratio (CNR) and background variability were evaluated as image quality parameters. All metrics were calculated in six different sphere sizes of NEMA IQ phantom.