This poster is published under an
open license. Please read the
disclaimer for further details.
Keywords:
Cardiac, Cardiovascular system, Computer applications, CT, CT-Angiography, CT-Quantitative, Computer Applications-3D, Computer Applications-Detection, diagnosis, Image compression, Arteriosclerosis, Image registration
Authors:
M. Károlyi1, B. Szilveszter1, R. Raaijmakers2, W. Giepmans2, C. CELENG1, A. Bartykowszki1, A. Panajotu1, B. Merkely1, P. Maurovich-Horvat1; 1Budapest/HU, 2Best/NL
DOI:
10.1594/ecr2015/C-2348
Results
The qualitative analysis showed that IMR improves image quality and image sharpness as compared to FBR and HIR (p<0.0001 all).
Image noise was significantly lower with HIR as compared to FBR and was further reduced with IMR as compared to HIR (p<0.0001 all).
By quantitative analysis mean attenuation in the aorta did not differ among all three reconstruction techniques (507.9±87.5 vs.
508.1±87.6 vs.
507.1±87.6 HU with FBR,
HIR and IMR techniques,
respectively; p=1.0). Image noise (SD) was highest using FBR and lowest using IMR (42.1±10.7 vs.
28.7±7.3 vs 12.9±2.7,
respectively; p<0.001).
Fig.
2.
Mean attenuation did not differ in the proximal coronaries among FBR,
HIR and IMR (529.3±108.3 vs.
529.1±108.3 vs.
565.9±118.6 HU,
respectively),
while it was significantly higher in the distal coronary segments with IMR as compared to FBR and HIR techniques and showed no difference between FBR and HIR (496.5±91.6 vs.
494.3±93.2 vs.
576.7±113.9 HU,
respectively). Fig.
3.
CNR in the proximal coronary arteries were improved with HIR and further improved with IMR,
as compared to FBR (22.9±8.0 and 52.2±11.4 vs.
15.7±5.3,
respectively; p<0.001 all).
CNR in the distal coronary arteries were also improved with HIR and further improved with IMR,
as compared to FBR technique (18.4±6.7 and 45.5±11.2 vs.
12.5±4.5,
respectively; p<0.001 all).
Fig.
4.