Lower-Dose Dual-Energy CTPA Preserves Pulmonary Embolism Detection Accuracy

Key Takeaways

• The dual-flow dual-energy approach was associated with a 55% lower radiation dose and about 40% less contrast medium than routine CTPA.
• Pulmonary embolism detection was similar between groups, with preserved pulmonary branch visualization and fewer superior vena cava artifacts.
• Forty-keV virtual monoenergetic reconstructions produced higher pulmonary arterial attenuation and a higher figure of merit, while central-vessel contrast-to-noise ratio remained comparable.

Among patients referred for suspected pulmonary embolism, a dual low-dose dual-energy CTPA protocol reduced the median effective dose to 1.61 mSv from 3.62 mSv in a head-to-head comparison with routine CTPA. Pulmonary embolism diagnostic performance did not differ significantly between protocols, contrast use was lower with the dual-flow approach, and pulmonary artery assessment remained diagnostically acceptable with the lower-dose protocol.

In this prospective single-center study, investigators consecutively enrolled patients referred for suspected pulmonary embolism and randomly assigned 61 to the dual low-dose group and 60 to routine CTPA. Exclusions included age under 18 years, BMI of at least 40 kg/m², severe renal dysfunction, and pregnancy. The tested protocol used a relative 30 HU trigger, 1-second monitoring, a 5-second delay, and a 1:1 contrast-saline mixture, followed by separate contrast and saline injections at 4.0 mL/s; the comparison arm underwent routine CTPA acquisition. Using 40-keV virtual monoenergetic images versus 100-kV routine images, investigators compared quantitative image metrics, while blinded radiologists assessed qualitative image quality and embolism detection.

At 40 keV, main pulmonary trunk attenuation measured 1027.0 ± 287.6 HU in the dual low-dose group and 391.8 ± 109.0 HU with routine imaging, a significant difference at p < 0.001. Pulmonary trunk contrast-to-noise ratio was comparable at 33.6 versus 41.0, whereas figure of merit favored the dual low-dose protocol at 711.4 versus 461.5. Reported p values were 0.115 for contrast-to-noise ratio and 0.006 for figure of merit, alongside about 40% less contrast medium use. Radiation exposure was also lower in the dual low-dose arm, although image noise was higher and signal-to-noise ratio was lower, while pulmonary artery visualization remained preserved.

Pulmonary embolism was identified in 13 cases, or 21.3%, in the dual low-dose group and 19 cases, or 31.7%, in the routine group. Diagnostic performance did not differ significantly between protocols in the reader study. Two suboptimal dual low-dose examinations were salvaged with virtual monoenergetic imaging, and subjective image-analysis agreement ranged from moderate to good between readers. κ values were 0.541 to 0.713 in the dual low-dose group and 0.614 to 0.793 in the routine group. The authors noted the single-center design, limited sample size, vendor-specific dual-source DECT without deep-learning reconstruction, exploratory diagnostic analysis, and lack of downstream outcome assessment.