Keywords:
CNS, Interventional vascular, Emergency, CT, Thrombolysis, Recanalisation, Complications, Haemorrhage, Embolism / Thrombosis, Ischaemia / Infarction
Authors:
S. VAN HEDENT, N. Grosse Hokamp, R. Beck, R. Kessner, P. R. Ros, D. W. Jordan; Cleveland, OH/US
DOI:
10.1594/ecr2018/C-3280
Results
There was a significant correlation between attenuation measurements on the conventional and VNC images,
for all samples containing blood (R2 > 0.97,
p=0.00) (Fig.
2).
For the samples containing blood,
correlation was highest for pure blood dilutions (R2 = 1.00) and lowest for samples containing 33% blood (R2 = 0.97).
Correlation was not significant for samples with pure iodine (R2 = 0.77,
p = 0.08).
Relative retention of VNC attenuation was significantly different between all categories (p< 0.05),
shown in table 1 and figure 3.
Results for all categories showed no overlap,
except for the 50%-50% and 33%-67% blood-iodine mixtures,
with minimum and maximum values of 58.06% – 83.44% and 44.35% - 70.98% respectively.
Applying a ≥98% and ≤20% cut-off value for VNC attenuation retention,
showed 100% accuracy for classification as pure blood and pure iodine respectively.
Consequently,
intermediate results (20% - 98%) accurately classified dilutions as blood-iodine mixtures.