Radiation Dose Reduction by Using CT with Iterative Model Reconstruction in Patients with Pulmonary Invasive Fungal...

Purpose To compare the diagnostic quality of reduced radiation dose computed tomography (CT) with iterative model reconstruction (IMR) versus that of conventional low-dose CT in patients with pulmonary invasive fungal infection. Materials and Methods This prospective observational study included 48 patients (mean age ± standard deviation, 39.9 years ± 11.3) known to have or suspected of having pulmonary invasive fungal infection between October 2016 and July 2017. Patients underwent CT with IMR (at 80 kV with 20 mA) immediately after low-dose CT (at 80 kV with automatic exposure control). Images were reconstructed by using a hybrid iterative reconstruction (HIR) algorithm and IMR. Two radiologists independently assessed subjective image quality, noise, and visibility of normal and abnormal findings by using a five-point scale. Objective measurements, including image noise, contrast-to-noise ratio (CNR), and corresponding figure of merit (FOM), were compared by using repeated-measures analysis of variance with Bonferroni post hoc tests for multiple comparisons. Results The mean effective dose was 0.3 mSv ± 0.3 for CT with IMR and 0.7 mSv ± 0.2 for low-dose CT (P < .01). When the image noise and CNR were normalized to the effective dose, CT images obtained with IMR had significantly higher FOM than did other image series (P < .0001). Subjectively, visibility of CT features of invasive fungal infection on CT scans reconstructed with IMR was rated as noninferior to that on low-dose CT scans reconstructed with HIR, except for the halo sign. Conclusion CT with IMR had approximately 60% dose reduction compared with conventional low-dose CT, with reduced noise and improved depiction of abnormal findings, in patients with pulmonary invasive fungal infection.