CT angiography of pulmonary arteries to detect pulmonary embolism: improvement of vascular enhancement with low kilovoltage settings.

PURPOSE: To retrospectively compare a low kilovoltage scanning protocol with a reduced radiation dose with a standard high kilovoltage, moderate-dose protocol for the depiction of central and peripheral pulmonary arteries at single-detector spiral computed tomography (CT).

MATERIALS AND METHODS: This retrospective study had institutional review board approval; informed consent was waived. A 100-kVp protocol (volume CT dose index [CTDI(vol)], 3.4 mGy) was compared with a standard 140-kVp protocol (CTDI(vol), 10.4 mGy) in two groups that were each composed of 35 consecutive patients who were suspected of having pulmonary embolism (PE) and scanned with otherwise identical acquisition parameters and contrast material injection protocols. Mean main pulmonary artery enhancement and maximum enhancement in peripheral pulmonary arteries were compared. In a blinded evaluation, the percentages of segmental and subsegmental arteries that were considered analyzable for assessment of PE were determined. Overall image quality and delineation of various anatomic areas were subjectively assessed. Comparison of percentages of analyzable segmental and subsegmental arteries and subjective grading of image quality between the two different protocols were performed with the Mann-Whitney U test.

RESULTS: There were 38 male and 24 female patients (mean age, 61 years; range, 17-86 years) in the final evaluation. There was a significantly higher average CT number in the main pulmonary artery (379 HU +/- 95) for the 100-kVp protocol than for the 140-kVp protocol (268 HU +/- 63, P < .001, two-sided t test). Maximum CT numbers in peripheral pulmonary arteries at the level of the aortic arch and lung bases, respectively, were 290 HU +/- 91 and 279 HU +/- 100 for 100 kVp and 185 HU +/- 65 and 144 HU +/- 63 for 140 kVp (P < .001). Mean percentage of subsegmental arteries considered analyzable per patient was higher for 100 kVp than for 140 kVp (segmental arteries, 92% vs 88%, P = .13; subsegmental arteries, 71% vs 55%, P < .001). Subjective grading of overall image quality and of the delineation of structures in the lungs, mediastinum, and upper abdomen did not significantly differ between protocols.

CONCLUSION: At reduced radiation exposure, low kilovoltage scanning increases the percentage of central and peripheral pulmonary arteries that can be evaluated with CT angiography without a substantial decrease in image quality.