Diagnostic accuracy of dual-source computed tomography in the diagnosis of coronary artery disease

Purpose: The aim of this study was to evaluate the diagnostic accuracy of dual-source computed tomography (DSCT) with reference to invasive coronary angiography in the diagnosis of coronary artery disease (CAD) on a per-patient as well as on a per-segment basis. Materials and Methods: Thirty-five patients with known or suspected CAD underwent both DSCT (Somatom Definition, Siemens Medical Solutions) and quantitative x-ray coronary angiography (QCA). Parameters of CT acquisition were gantry rotation time 0.330 seconds (ie, temporal resolution 83 milliseconds), tube vol 120 kV, tube current 560 mA with ECG-triggered tube current modulation and full current at 70% of the cardiac cycle for heart rates below 70 beats per minute or full current between 30% and 80% for higher and arrhythmic heart rates. The pitch was also adapted to the heart rate, ranging from 0.2 to 0.43. Volume and flow rate of contrast material (Ultravist 370, Schering AG) were adapted to the patient's body weight. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of DSCT in the detection or exclusion of significant CAD (ie, stenoses > 50%) were evaluated on a per-patient and per-segment basis. Results: All 35 CT angiograms were of diagnostic image quality. QCA demonstrated significant CAD in 48% (n = 17) and nonsignificant disease or normal coronary angiograms in 52% (n = 18) of the patients. Sensitivity, specificity, PPV, and NPV of DSCT on a per-patient basis were 100%, 89%, 89%, and 100%, respectively. On a per-segment basis, 473 of 481 coronary artery segments were assessable (98%). QCA demonstrated stenoses > 50% in 32 segments (7%), and no disease or nonsignificant disease in 433 segments (93%). For the detection of stenoses > 50% on a per-segment basis, DSCT showed a sensitivity, specificity, PPV, and NPV of 88%, 98%, 78%, and 99%, respectively. Conclusions: The comparison of coronary DSCT with QCA shows a very robust image quality and a high diagnostic accuracy in a patient-based as well as a per-segment analysis. Maximal sensitivity and NPV in the per-patient analysis show the strength of the technique in ruling out significant CAD.