Keywords:
Abdomen, Computer applications, Liver, CT-Quantitative, CT, Computer Applications-General, Imaging sequences, Technical aspects, Tissue characterisation, Cancer
Authors:
M. Mottola, A. Bevilacqua; Bologna/IT
DOI:
10.26044/ecr2019/C-0350
Results
Mean eBF and BF values (the latter computed with MS) results equal to [119±16]ml/min/100g and [129±27]ml/min/100g,
respectively.
In fact,
eBF includes the contribution at the outflow,
as expected.
Moreover,
by considering the whole cohort of 46 patients,
the CV coefficient computed on eBF (CV=13%) is almost half than the one on BF with MS (CV=21%),
confirming that perfusion estimates provided eBF are more precise than those of the “classical” BF.
Nonetheless,
correlations between eBF and BF colormaps are good,
with R2≥0.90 for 36 patients and 0.80≤R2<0.90 for the remaining 10 patients.
Fig.
2 shows the colorimetric map of eBF for patients ID16 (a),
63 (b),
and 23 (c).
In addition,
for the same patients Fig.
3 shows the colorimetric maps of BF achieved with MS.
Voxel-based matching of highly and lowly perfused areas is undoubtedly excellent for patient ID16 (R2=0.99),
and very good indeed for patient ID63,
showing R2=0.96.
In addition,
despite the lower correlation index,
R2=0.87,
the perfusion pattern showed in Figs.
1-2 (c) is very similar as well.
In particular,
the latter case emphasizes how eBF estimates spread in narrower ranges if compared to the MS ones.
In fact,
colorbars result compressed and highlights more areas with relatively higher perfusion values.
Finally,
perfusion values and colormaps confirm the overall underestimation of eBF with respect to the classical BF with MS.