Low dose computed tomography of the lung for detection and grading of interstitial lung disease: A systematic simulation study

Abstract

Purpose: HRCT is the preferred imaging technique to evaluate Interstitial-Lung-Disease. Optimal Low-Dose-Computed-Tomography protocol for monitoring ILD with lowest radiation dose and optimal diagnostic accuracy and image quality unknown. Methods: 28 Patients underwent HRCT. Image reconstructions with varying combinations of tube current (50mA, 20mA, 15 mA, 10mA) and image-thickness/increment (1/1mm, 2/2mm, 3/2.4mm, 5/4mm) were simulated from raw data. 448 CTs evaluated by 2 readers on image quality and ILD-specific features (ground glass opacification (ggo), honeycombing (hc), reticulation (ret)). Results: Reduced dose settings with 20 mA did not show any significant difference to standard dose settings for all parameters in reader 1, while results were significantly altered in reader 2. Slice thickness did not significantly influence rating of typical ILD features like ggo, hc, ret or total disease extent. The correct differentiation between UIP and NSIP could be made on all dose settings and with all slice thickness. It was even found, that an increased slice thickness can compensate for the noise associated image quality degradation. Overall, for ggo detection a combination of 20 mA and 3 or 5 mm slice thickness was not different to the original evaluation. Conclusions: Assessment of ILD specific CT features down to 20 mA and a slice thickness of 3 or 5 mm is feasible.

Keywords

• Interstitial Lung Disease
• Computed Tomography
• Low-dose
• Simulation study
• Diagnostic performance