Detection of pulmonary emboli with CT angiography at reduced radiation exposure and contrast material volume: comparison of 80 kVp and 120 kVp protocols in a matched cohort.

OBJECTIVE

The detection rate of pulmonary emboli (PE) with computed tomography angiography (CTA) using either a standard or a low-dose protocol, combining reduced radiation exposure and iodine delivery rate, was retrospectively analyzed in a matched cohort of 120 patients.

MATERIALS AND METHODS

The study was performed according to the regulations of the institutional review board. Four groups of 30 patients each, with a body weight of less than 100 kg and receiving pulmonary CTA were matched by age (range, 21-87 years), gender (female/male, 48/72), weight (range, 41-99 kg), and cross sectional area of the chest (range, 468-885 cm2). Sixty patients had PE and 60 patients had no PE at CTA. The CT tube voltage was either 80 kVp (group A, with PE and group B, with no PE) or 120 kVp (group C, with PE and group D, with no PE). Volume and flow rate of injected contrast medium was lower with the 80 kVp protocol (75 mL at 3 mL/s) compared with the 120 kVp protocol (100 mL at 4 mL/s). Contrast-to-noise ratio (CNR) for the pulmonary trunk was calculated. Two independent readers analyzed all CTAs in a randomized order for the localization of emboli, diagnostic confidence, and image quality. The reference standard for the presence of emboli involved consensus reading and assessment of available clinical data and findings with additional imaging modalities. CNR, subjective image quality, diagnostic confidence, sensitivity, and specificity for emboli at both tube voltages were compared.

RESULTS

All patients with PE were correctly identified with both protocols, corresponding to a sensitivity of 100% at the patient level. For the localizations with emboli, both the sensitivity (83.7% at 80 kVp and 83.6% at 120 kVp; P = 0.921) and the specificity (97.2% at 80 kVp and 97.8% at 120 kVp; P = 0.463) were not significantly different at the 2 tube voltages. The diagnostic confidence was not different at all ramification levels (P = 0.216-1.0). CNR did not differ between the groups (P = 0.202). The overall subjective image quality was higher at 120 kVp compared with 80 kVp (P = 0.017).

CONCLUSION

Detection rate and diagnostic confidence for the presence of pulmonary emboli with low-dose pulmonary CTA using 80 kVp and reduced iodine delivery rate may be equal to that at 120 kVp in patients weighing less than 100 kg.