PET | Evaluation of MR Based Respiratory Motion Correction Technique i…

Yong Ho Do, Hong Jae Lee, Jin Eui Kim

Department of Nuclear Medicine,  Seoul  National University Hospital, Seoul,  Korea

[Purpose]  Respiratory  motion during  PET/MR  acquisition may   result in  image   blurring and  error   in measurement for  volume   and  quantification of  lesion. The aim  of  this study was to evaluate changes of  quantitative accuracy, tumor volume   and  image   quality by  applying MR  based respiratory motion correction technique (MBRMCT) using integrated PET/MR scanner.

[Materials   &  Method]　 Data  of  8  patients  (aged  60.1  ±  9.1  y)   underwent  18F-FDG  liver   PET/MR (Biograph  mMR   3.0T,  Siemens)  study   were    collected.  PET   listmode  data   for   7   minutes  was

simultaneously acquired with  maximum average  gate  (MAG),   minimum time gate  (MTG)  and  none gate  (NG)   T1  weighted  MR   images.  Gated  PET  reconstruction  was  performed  using  mu-maps generated from  MAG  and  MTG by  setting 35%  of efficiency window. Maximum SUV  (SUVmax), peak SUV  (SUVpeak), tumor  volume   and  full  width at  half  maximum (FWHM)  in  the z-axis  direction of MAG,  MTG and  NG PET images were  evaluated.

[Results]　 Compared  to  NG,   mean   SUVmax   and   SUVpeak    were   increased  in   MAG   13.62%(p=0.002),

7.42%(p=0.002),   MTG  14.50%(p=0.001),   7.97%(p=0.001)   and   mean   tumor  volume   and   FWHM  were decreased    in     MAG     13.47%(p=0.004),     17.48%(p=0.056),     MTG    13.98%(p=0.003),     17.66%(p=0.05) respectively. Mean  SUVmax  and  SUVpeak  of  MTG were  increased by  0.77%(p=0.28),  0.51%(p=0.398)  and mean  tumor volume  and  FWHM were  decreased by  0.58%(p=0.523),  0.23%(p=0.487)  compared to MAG. There was no  statistically  significant difference between  MTG and  MAG  which   increase total scan time for  7 and  1.5 minutes.

[Conclusion]  SUV,  accuracy of  tumor volume   and  image   quality were   improved in  both of  MAG  and MTG  by  applying  MBRMCT without  installing  additional hardware  in  liver   PET/MR  study.  More accurate  information  can   be   provided  with  the  increase  of   1.5   minutes  scan  time  if   applying MBRMCT to various abdominal PET/MR studies affected by  respiratory motion.

[Key  Words]  Respiratory motion correction, PET/MR,  SUV,  Liver  cancer