Successful suppression of cone beam artifacts using iterative reconstruction algorithm

Cone beam artifacts are commonly observed in volumetric CT images reconstructed by use of filtered backprojection method such as Feldkamp algorithm when the cone angle is relatively large and bony structures are present, for example in the head imaging task. In this work, we try to answer the question “Is it possible to suppress cone beam artifacts using iterative reconstruction algorithms in circular volumetric CT?” Our answer to this question is “YES”. We support our conclusion by use of clinical head dataset to avoid any possible excuses in simulated data. A set of 320-slice clinical head-data were acquired from Aquilion ONE™ CT scanner (Toshiba American Medical Systems, Tustin, CA, USA) for our test. We intentionally extracted central 256-slice data to synthesize an “easy dataset” and treated the original 320-slice data as a “hard dataset” due to different cone angles these datasets subtended. We demonstrate in this paper that cone beam artifacts in volumetric head CT images reconstructed from both “easy dataset” and “hard dataset” are successfully suppressed using iterative reconstruction algorithm.