Low-dose radiation with 80-kVp computed tomography to diagnose pulmonary embolism: a feasibility study

Background Mounting collective radiation doses from computed tomography (CT) implies an increased risk of radiation-induced cancer in exposed populations, especially in the young. Purpose To evaluate radiation dose and image quality at 80-kVp CT to diagnose acute pulmonary embolism (PE) compared with a previous study at 100 and 120 kVp with all other scanning parameters unchanged. Material and Methods A custom-made chest phantom with a 12 mg I/mL-syringe was scanned at 80/100/120 kVp to evaluate relative changes in computed tomographic dose index (CTDIvol), attenuation, image noise, and contrast-to-noise ratio (CNR). Fifty patients underwent 80 kVp 16-row detector CT at 100 'Quality reference' mAs. A total of 350 mg I/kg were injected to compensate for increased CNR at 80 kVp, while 300 mg I/kg had been used at 100/120 kVp. CTDIvol, dose-length product (DLP), and estimated effective dose were evaluated including Monte Carlo simulations. Pulmonary artery attenuation and noise were measured and CNR calculated. Two radiologists evaluated subjective image quality using a four-grade scale. Results Switching from 120 to 80 kVp in the phantom study decreased radiation dose by 67% while attenuation and noise increased 1.6 and 2.0 times, respectively, and CNR decreased by 16%. Switching from 120 to 80 kVp in the patient studies decreased estimated effective dose from 4.0 to 1.2 mSv (70% decrease) in median while pulmonary artery attenuation and noise roughly doubled from 332 to 653 HU and from 22 to 49 HU, respectively, resulting in similar CNR (13 vs. 12). At 80 kVp all examinations were regarded as adequate (8%) or excellent (92%). Conclusion Switching from 120 to 80 kVp CT without increased mAs but slightly increased iodine dose may be of special benefit to diagnose PE in younger individuals with preserved renal function where the primary aim is to minimize radiation dose and reaching levels below that of scintigraphy