01-06-2014 | Cardiac
Evaluation of monoenergetic late iodine enhancement dual-energy computed tomography for imaging of chronic myocardial infarction
Published in: European Radiology | Issue 6/2014
Login to get accessAbstract
Objectives
To evaluate image quality and diagnostic accuracy of selective monoenergetic reconstructions of late iodine enhancement (LIE) dual-energy computed tomography (DECT) for imaging of chronic myocardial infarction (CMI).
Methods
Twenty patients with a history of coronary bypass surgery underwent cardiac LIE-DECT and late gadolinium enhancement (LGE) magnetic resonance imaging (MRI). LIE-DECT images were reconstructed as selective monoenergetic spectral images with photon energies of 40, 60, 80, and 100 keV and the standard linear blending setting (M_0.6). Images were assessed for late enhancement, transmural extent, signal characteristics and subjective image quality.
Results
Seventy-nine myocardial segments (23 %) showed LGE. LIE-DECT detected 76 lesions. Images obtained at 80 keV and M_0.6 showed a high signal-to-noise ratio (15.9; 15.1), contrast-to-noise ratio (4.2; 4.0) and sensitivity (94.9 %; 92.4 %) while specificity was identical (99.6 %). Differences between these series were not statistically significant. Transmural extent of LIE was overestimated in both series (80 keV: 40 %; M_0.6: 35 %) in comparison to MRI. However, observers preferred 80 keV in 13/20 cases (65 %, κ = 0.634) over M_0.6 (4/20 cases) regarding subjective image quality.
Conclusions
Post-processing of LIE-DECT data with selective monoenergetic reconstructions at 80 keV significantly improves subjective image quality while objective image quality shows no significant difference compared to standard linear blending.
Key Points
• Late enhancement dual-energy CT allows for detection of chronic myocardial infarction.
• Monoenergetic reconstructions at 80 keV significantly improve subjective image quality.
• 80 keV and standard linear blending reconstructions show no significant differences.
• Extent of CMI detected with LIE-DECT is overestimated compared with MRI.