Individual selection of X-ray tube settings in computed tomography coronary angiography: Reliability of an automated software algorithm to maintain constant image quality

• Topogram-based selection of X-ray tube settings guarantees diagnostic image quality in CTCA.
• Topogram-based selection of X-ray tube settings may ensure a low radiation exposure in CTCA.
• However, individual CNRs may vary considerably between subjects.

ObjectivesTo evaluate a software tool that claims to maintain a constant contrast-to-noise ratio (CNR) in high-pitch dual-source computed tomography coronary angiography (CTCA) by automatically selecting both X-ray tube voltage and current.MethodsA total of 302 patients (171 males; age 61 ± 12 years; body weight 82 ± 17 kg, body mass index 27.3 ± 4.6 kg/cm2) underwent CTCA with a topogram-based, automatic selection of both tube voltage and current using dedicated software with quality reference values of 100 kV and 250 mAs/rotation (i.e., standard values for an average adult weighing 75 kg) and an injected iodine load of 222 mg/kg.ResultsThe average radiation dose was estimated to be 1.02 ± 0.64 mSv. All data sets had adequate contrast enhancement. Average CNR in the aortic root, left ventricle, and left and right coronary artery was 15.7 ± 4.5, 8.3 ± 2.9, 16.1 ± 4.3 and 15.3 ± 3.9 respectively. Individual CNR values were independent of patients’ body size and radiation dose. However, individual CNR values may vary considerably between subjects as reflected by interquartile ranges of 12.6–18.6, 6.2–9.9, 12.8–18.9 and 12.5–17.9 respectively. Moreover, average CNR values were significantly lower in males than females (15.1 ± 4.1 vs. 16.6 ± 11.7 and 7.9 ± 2.7 vs. 8.9 ± 3.0, 15.5 ± 3.9 vs. 16.9 ± 4.6 and 14.7 ± 3.6 vs. 16.0 ± 4.1 respectively).ConclusionA topogram-based automatic selection of X-ray tube settings in CTCA provides diagnostic image quality independent of patients’ body size. Nevertheless, considerable variation of individual CNR values between patients and significant differences of CNR values between males and females occur which questions the reliability of this approach