Nick Weir PhD, Presenter: Nothing to Disclose

Michelle Claire Williams MBCHB, BSC, Abstract Co-Author: Nothing to Disclose

Fraser R Millar, Abstract Co-Author: Nothing to Disclose

Edwin J.R. Van Beek MD, PhD, Abstract Co-Author: Research support, Toshiba Corporation Research support, Siemens AG Founder and owner, QCTIC, Inc Speaker, Toshiba Corporation

This phantom study was designed to validate a new iterative reconstruction algorithm in 320 detector-row computed tomography (CT) for potential radiation dose reduction. We evaluated various image quality parameters across a range of exposure factors, to find the optimal level of dose reduction while maintaining image quality.

The Catphan 600 phantom was used with a 30 cm expansion annulus to simulate body scanning conditions. Using a cardiac prospective electrocardiogram-gated protocol (120 kVp, 0.35s rotation, 128 mm collimation) a series of scans were acquired with tube current (mA) reducing in 10% steps from 510 to 150 mA. Selected mA levels were investigated at 100 and 135 kVp. Each image volume was reconstructed using standard Quantum Denoising Software (QDS+) and then using Adaptive Iterative Dose Reduction (AIDR) and QDS+ combined. Noise, uniformity, high contrast resolution and contrast-to-noise ratio (CNR) were evaluated for each volume. Image quality parameters were averaged over 30 images.

Reconstructions using AIDR at 50% mA produced images with nearly equivalent noise levels to those acquired at 100% mA without AIDR at all kVp. At 120 kVp there was no difference in CNR using AIDR at 50% mA compared to 100% mA without AIDR (10.8 vs 10.2, p=0.31). Image uniformity was acceptable in all reconstructions, improving with AIDR (3.4 vs 5.1 HU, p<0.001). At a given tube current, high contrast spatial resolution in terms of MTF10 was 4-8% lower in images reconstructed using AIDR (p<0.001). At low mA, MTF10 increased due to automatic switching to small focal spot such that, at 50% mA, resolution with AIDR exceeded that at 100% mA without AIDR (mean 0.82 vs 0.71 lp/mm, p<0.05).

The AIDR algorithm offers potential for 50% dose reduction in body CT without significant reduction in quantitative image quality. This will enable many body scanning protocols to utilize fine focal spot scanning with a concomitant increase in high contrast spatial resolution.

Phantom validation tests demonstrate that Adaptive Iterative Dose Reduction software for 320 detector-row CT offers potential for 50% dose reduction in body scanning while maintaining image quality.

Weir, N, Williams, M, Millar, F, Van Beek, E, Feasibility of Radiation Dose Reduction in 320 Detector-row Body Computed Tomography Using Iterative Reconstruction: A Phantom Study.  Radiological Society of North America 2011 Scientific Assembly and Annual Meeting, November 26 - December 2, 2011 ,Chicago IL. http://archive.rsna.org/2011/11034493.html