Kenneth K. Lau, Presenter: Nothing to Disclose

Eileen C. Ang MBBS, BMedSc, Abstract Co-Author: Nothing to Disclose

Nicholas David Ardley, Abstract Co-Author: Nothing to Disclose

Kevin Buchan, Abstract Co-Author: Employee, Koninklijke Philips NV

Multiple reconstructive techniques including different forms of iterative reconstructions (IR) improve image quality (IQ)/spatial resolution whilst reducing radiation dose in CT. The latest Iterative model reconstruction (IMR) is a knowledge-based algorithm with improved low contrast resolution and produces relatively noise-free images. Improved IQ greatly aids solid organ lesion detection and hence improves patient outcome. The aim of this exhibit is to assess the diagnostic utility of IMR in CT of abdomen.

The CT data sets of 100 patients (mean age of 68) were reconstructed using IMR and iDose IRs. 1. The image noise using SD of attenuation values of liver was significantly improved by 41.3% from iDose to IMR and 39-71% in other organs. 2. The margins and internal architecture of lesions in liver, kidneys and other solid organs were better defined in IMR. There was significant improvement of IQ assessment of liver lesion with Mann-Whitney test. 3.The ureteric visualization and calculus detection, and blood vessel details were enhanced on IMR due to reduction of image noise in the adjacent fat. IMR is superior to conventional iterative reconstruction by producing relatively ‘noiseless’ CT images that enables better lesion detection.

http://abstract.rsna.org/uploads/2014/14019197/14019197_jknz.pdf

Lau, K, Ang, E, Ardley, N, Buchan, K, Advanced Iterative Model Reconstruction in Improving Image Quality of CT Abdomen.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14019197.html