Thomas Allmendinger, Abstract Co-Author: Employee, Siemens AG

Rainer Raupach PhD, Abstract Co-Author: Employee, Siemens AG

Herbert Bruder, Abstract Co-Author: Employee, Siemens AG

Bernhard Schmidt PhD, Abstract Co-Author: Employee, Siemens AG

Thomas G. Flohr PhD, Presenter: Employee, Siemens AG

Evaluation of a novel temporal domain extended 4D iterative cardiac reconstruction algorithm applied to clinical coronary CT angiography (CCTA) data . The reconstruction is compared to an established iterative reconstruction algorithm (SAFIRE, Siemens Healthcare, Germany) and weighted filtered back projection.

The main limitation of cardiac image reconstruction especially in obese patients is the underlying goal of maximizing the temporal resolution, which subsequently does not allow noise reduction by data accumulation, e.g. by a slow rotation time or similar techniques. An iterative 4D reconstruction algorithm enables data accumulation in the temporal direction by analyzing possible cardiac motion. The 4D algorithm is an extension of the established SAFIRE iterative reconstruction algorithm. The 4D algorithm is based on an initial 4D volume reconstructed at 5 adjacent cardiac phases with a tempoiral spacing of half a reconstruction angle. In an iterative optimization a raw data supported image noise model is utilized to analyze local image contrast. In comparison to the 3D approach where each voxel is evaluated relative to its 26 neighboring voxels, the 4D volume extends this to 80 neighboring voxels. This increase in statistics yields a more robust noise modeling and therefore in the subsequent subtraction process a improvement in the contrast to noise ration in each iteration. We used 15 clinical CCTA data sets acquired with a second-generation dual-source CT (Siemens Healthcare, Germany) to evaluate the 4D algorithm. The data sets were reconstructed with a filtered back projection algorithm, the SAFIRE algorithm and the extended 4D algorithm. We evaluated image noise and contrast to-noise ratio as well as image quality with respect to spatial resolution and the presence of motion artifacts on a subjective rating five point scale.  

On average the noise was reduces by 45 % in the standard SAFIRE case and by 65 % in the extended 4D cases compared to filtered backprojection. Image quality and artifact level was comparable in all three cases.

Inital experience with an 4D iterative algorithm indicates an greatly improved contrast to noise ratio compared to standard 3D iterative techniques.

An improved contrast to noise ratio in cardiac images could be useful for CCTA diagnosis especially in obese patients where image noise is often the limiting factor.

Allmendinger, T, Raupach, R, Bruder, H, Schmidt, B, Flohr, T, Analysis of Image Quality and Contrast in Computed Tomography Coronary Angiography Using a 4D Iterative Reconstruction Algorithm.  Radiological Society of North America 2012 Scientific Assembly and Annual Meeting, November 25 - November 30, 2012 ,Chicago IL. http://archive.rsna.org/2012/12028325.html