PET | Correlation between barium contrast agents and the artifact in PE…

Byung  jin  Kim,  Jae  Ryong  Woo,  So  Won  Oh,  Yu Keong  Kim.

NUCLEAR  MEDICINE,  SMG-SNU  BORAMAE  MEDICAL  CENTER,   Seoul,   Korea

Purpose  :We   have   been   found   various  artifacts  disturbing  interpretation of   PET/CT  images.  The   main   factors  that  may   cause  the  artifacts  in PET/CT  images  were  contrast  media,    patient  motion,  prosthesis,   etc. The   purpose of  this study was   to recognize the  correlation between the density  of  barium   contrast  agents  and  artifacts  in  PET/CT  images  with the phantom test.  Methods :  Small  inserts  with various barium   sulfate concentration  were  inserted   into  cylindrical  phantom.  12   small   inserts had  a  cylindrical shape(diameter  18  mm,  volume   43  ml)  and  have   filled up   with  barium    concentrations   ranging  from    1.5    to  200%  W/V.   We acquired  PT/CT  images  for   cylindrical  phantom  that  filled   with  F-18 solution(   F-18   80±1  MBq,   water  6700  ml).  Also,   CT   images  were acquired and  then PET   data   were  acquired after  moving   the  phantom to simulate the  motion artifact.  In  other  words, we  acquired  CT  images  of the cylindrical phantom that  was  placed   in  PET  scanner by  aligning

its center axis to the axis of  the gantry and  then acquired the PET   data with  a  spacing  2  mm   ranging  from   0  to  10   mm.   The   PET   data   were reconstructed with OSEM algorithm. We measured the coefficient of variation(CV)   around    the  small    insert   to  get   to  know    whether   the barium   concentrations  have   any  effect  toward the  variation of  the signal to noise   in  the  PET/CT image.   Results  :  If  the CV  is over   14.2%,  the false  positive  regions  appeared  around    the  other  barium    small  inserts except  1.5%   W/V  in  CT   based  attenuation  correction  of   PET   images. The   barium   contrast 40  %  W/V  exhibited a  strong opaque

with high  hounsfield unit (2951±92)  and  showed the  ring  artifact in  the vicinity   of    itself.   Additionally,  barium    concentration   caused   by    the artifacts was   lower   by  10%   W/V  when   the  mismatched spacing between the PET   image   and  the CT  image   was  over   4  mm.  However, the artifact by  barium   contrast  agents  not  appeared in  the  nonattenuation corrected PET   images.   Conclusions  :   High   density  barium    contrast  lead   to  the false  positive  image   in  PET/CT.  If  there  is  the  motion at   the  portion that exists  the high  density  barium, the  effect  of  the artifacts  could   be further increased.