Gregory Vorona
Iterative reconstruction algorithms for CT, which have been made possible through recent advances in hardware capabilities and algorithm design, allow for significant radiation dose reductions with preservation of image quality. CT image reconstruction has traditionally utilized filtered back-projection, which operates on several limiting fundamental assumptions about scanner geometry. Iterative reconstruction techniques are more computationally intense, and involve a series of updates (iterations) that drive the CT image toward a more accurate representation of the object that was scanned.
I became interested in learning about the application of these techniques through a project that I was involved with as a resident at the Children’s Hospital of Pittsburgh, under the direction of their department chairman, Dr. Ashok Panigrahy. The Children’s Hospital of Pittsburgh was one of the first pediatric hospitals to use GE’s Adaptive Statistical Iterative Reconstruction (ASIR) algorithm, which is a partial iterative reconstruction algorithm that takes into account the statistical fluctuation of noise during image formation. Through our work we were able to demonstrate that approximately 40% dose reduction in pediatric abdominal CT using 40% ASIR, and approximately 20% dose reduction in pediatric head CT using 20% ASIR, is feasible without significantly impacting image quality. The Children’s Hospital of Pittsburgh