Keywords:
Cardiac, CT, Contrast agent-intravenous, Metabolic disorders
Authors:
C. I. Mihalache1, J. Mayer2, T. Damy3, A. Luciani2, H. Kobeiter2, A. Rahmouni2, J. F. Deux4; 1Meaux/FR, 2Creteil/FR, 3Paris/FR, 4Cedex/FR
Results
The patients with CA exhibited significant higher wall thickness than control patients (Table 1).
On first pass images,
patients with CA exhibited a significantly lower SNR myoc than control subjects (Table 2).
Myocardial attenuation trended to be lower in patients with CA than in control patients but the difference did not reach statistical significance (P=0.1).
Blood pool attenuation,
SNR blood and CNR blood-myoc were in the same range between groups.
On delayed images (Table 3),
myocardial attenuation and SNR myoc were significantly higher in patients with CA than in control patients.
Blood pool attenuation and SNR blood were in the same range between groups and consequently CNR blood-myoc was significantly lower in patients with CA than in control patients.
The mean value of the RAI was 0.12 ± 0.25 in patients with CA.
It was positive in all but 2 patients with CA and reflected the increase of myocardial attenuation between first pass and delayed images.
In contrast,
all control patients exhibited a negative value of the RAI because their myocardial attenuation decreased between first and delayed acquisitions.
The mean value of the RAI in control patients was -0.56 ± 0.21.
It was statistically (P<0.05) lower than index of patients with CA.
Images of patients with and without CA are reported in Figures 1 and 2.
The ROC curve and the Youden test identified an optimal threshold of -0.32 for relative attenuation index (AUC: 1; 95%CI [1–1]; p=0.0001) as the best threshold for predicting cardiac involvement on MDCT.
Sensitivity and specificity of the -0.32 HU threshold value to predict cardiac involvement were 100% and 100%,
respectively.
Weak but significant negative correlations were noticed between myocardial attenuation measured on first pass images and anterior wall thickness (r=-0.46; p=0.03),
inferior wall thickness (r=-0.45; p=0.04),
interventricular wall thickness (r=-0.43; p=0.05) and lateral wall thickness (r=-0.66; p=0.002) (Figure 3).
No significant correlation was detected between myocardial attenuation measured on delayed enhanced images and wall thickness.