Quantitative analysis of PET image quality in PET/MR brain examinations

Ziel/Aim:

According to [1] and [2] the Philips Ingenuity PET/MR (PETMR) has the same NEMA sensitivity as the Siemens ECAT HR+ standalone PET (PET). However, the NEMA sensitivity is measured without taking into account any attenuating materials such as the patient bed or MRI head coil which are present in every PET/MR brain examination. The aim of this work is to measure and compare the effective PET sensitivity and signal to noise ratios of reconstructed images (SNR) of PETMR for brain examinations under clinical conditions.
Methodik/Methods:
A solid 68-Ge phantom (33 MBq) was measured in PETMR and PET on the same day. The total acquisition time was 5min. The phantom was positioned in the MRI head coil in the PETMR and on a dedicated carbon head support in PET as used in clinical examinations. The ratio of prompt coincidences (PR) from the listmode files was calculated. Images were reconstructed employing parameters used in clinical routine. The SNR was calculated as mean over a 3D ROI with a diameter of 10cm divided by standard deviation.
Ergebnisse/Results:
PR was 0.52 (PETMR / PET). SNR was (8.6, 5.2, 3.7, 1.6) (PETMR) and (6.2, 3.9, 2.9,1.6) (PET) for frame durations of 300s, 100s, 50s and 10s, respectively.
Schlussfolgerungen/Conclusions:
Although the effective sensitivity of PETMR for brain examinations is lower than the one of PET, the resulting SNR in the reconstructed images is superior except for short frame durations. The influence on clinical quantification is currently investigated and will be reported.
Literatur/References:
[1] Zaidi et al. Design and performance evaluation of a whole-body Ingenuity TF PET-MRI system. Phys Med Biol 2011
[2] Herzog et al. NEMA NU2-2001 guided performance evaluation of four Siemens ECAT PET scanners, IEEE TMI
2004