Fig. 4

The differences between the visual assessment scores among the PET images of torso oncologic patients with different reconstruction algorithms. These were shown for OSEM + TOF (3 iterations, 28 subsets, 4.0-mm filter), Q.Clear (β = 100, 200, 300, 400, 500, 800, 1100 and 1400). Values are presented as medians (IQRs). * and ^# represent statistical difference compared to OSEM + TOF for physician 1 and physician 2, respectively (Bonferroni corrected P < 0.05)