We're using [15O]H2O- PET to study effects of deep anesthesia on rCBF.
According to a quantitative ROI analysis (n=16) regional flow awake was
roughly 65 (ml/100g/min) in gray matter and 25 in white matter. During
anesthesia flow values were reduced to 29 and 14, respectively. Thus, the
g/w flow ratio reduced from 2.6 to 1.8 (30% reduction).
In SPM the results from subtraction analysis show large white matter
"activation", partly masked out by Analysis treshold (0.8). We used mean
voxel value (w. fullmean/8 mask) for global flow calculation, and
proportional scaling for global normalisation.
Could this white matter activation reflect a problem in proportional
scaling, caused by the change in g/w matter flow ratio? If I have understood
correctly, the voxel volume used in global flow calculation (wholemean/8
mask for each image) is not identical to the one that actually participates
in statistical analysis (mask.img for all images). In normal activation
studies this is propably not an issue as relative whm flow remains constant.
In our case we detect significant changes in areas that are included in the
first but not the latter. Could this result in a see-saw effect that might
compromise all statistics?
1) If not, the white matter "activation" could be "hidden" from actual
analysis using explicit gray matter mask. I propably sound naive, but is
this advisable?
2) If true, one could use gray matter mask on images before taking them to
SPM, but such manipulation seems a bit extreme?
I couldn't find anything quite similar from the e-mail archive. I would
highly appreciate any suggestions.
Yours,
Kaike
Kaike Kaisti
________________________________________
=#= =#= Work: =#= =#= !
TUCH/ Dept of Anesthesiology and Intensive Care
P.O.box 52 ; 20521 TURKU ; Finland
tel+358-2-2611611 ; fax+358-2-2613960
www.utu.fi/med/pet/anesthes.htm
www.utu.fi/med/pet/index.html .
=#= =#= Home: =#= =#= !
Laikkakuja 14 B ; 20660 Littoinen ; Finland
tel+358-2-2365446 ; GSM:040-7686871
________________________________________
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|