Ph. Pretorius et al., The influence of attenuation and scatter compensation on the apparent distribution of Tc-99m sestamibi in cardiac slices, J NUCL CARD, 8(3), 2001, pp. 356-364
Background. Our objective was to study the differences in relative count di
stributions in the left ventricular walls with attenuation compensation (AC
) versus AC and triple-energy-window scatter compensation (SC), compared wi
th standard filtered backprojection (FBP),
Methods ann Results, Two hundred patients identified as having normal cardi
ac perfusion with FBP after undergoing either pharmacologically or physiolo
gically induced stress were included in this study Projection data were rec
onstructed with FBP, 10 iterations of ordered-subset expectation-maximizati
on (OSEM) with AC, and OSEM with AC+SC, A comparison was made of average pe
rcentage of maximum counts within each of 9 regions of CEqual (Marconi Medi
cal Systems, Inc, Cleveland, Ohio) polar maps tie, the apex, 3 midventricul
ar regions, and 4 basal regions), Compared with OSEM(AC), a slight decrease
at the apex exists when SC is included, The elevated inferior-to-anterior
count ratio in the midventricular and basal regions noted with OSEM(AC) dec
reased to close to 1.0 with OSEM(AC+SC). The anterior-to-lateral ratio for
both regions was closest to 1.0 for OSEM(AC+SC), In the midventricular regi
on, the lateral-to-septal ratio decreased further below 1.0 with OSEM(AC+SC
) than it did with OSEM(AC). This was the only basal ratio not to improve t
o close to 1.0 with OSEM(AC+SC). In a subset of patients identified at the
time of clinical reading as having a possible attenuation-caused decrease i
n the inferior region, AC elevated the inferior-to-anterior ratio to above
1.0 for the midventricular region. AC+SC resulted in a ratio of near 1.0 fo
r this region. In another subset of patients identified as having anterior
attenuation artifacts, compensation methods (either AC or AC+SC) failed to
show an improvement compared with FBP.
Conclusions. AC and SC improve the uniformity of the polar map, especially
by bringing the inferior-to-anterior ratio closer to 1,0, Further investiga
tion is necessary to determine the cause of the increased midventricular se
ptal polar map count. In addition, the subset of patients identified as hav
ing breast-like attenuation artifacts causing a decreased polar map count i
n the anterior mall (relative to the inferior wall) also needs further atte
ntion.