Skip to main content
Fig. 7 | EJNMMI Physics

Fig. 7

From: MLAA-based attenuation correction of flexible hardware components in hybrid PET/MR imaging

Fig. 7

Results for HP experiment: activity. a Coronal views of attenuation maps and corresponding activity distributions. Neglecting headphone attenuation (uncorrected) results in an activity underestimation compared to the CT-based reconstruction (CTAC) for regions embraced by the headphones. Correction for headphone attenuation using xMLAA compensates for the activity underestimation. The activity difference images in the bottom row give the relative difference to the CT-based reconstruction. b Average activity values evaluated within the phantom for transversal planes with varying z-position. Plotted are the transversal planes with plane numbers 35 to 105, as indicated in (a). In the uncorrected case, the average activity is underestimated by up to 13.4% compared to the reference scan. Average activity obtained by xMLAA is slightly higher than in the reference scan with a maximum overestimation of 1.7%

Back to article page