Abstract Archives of the RSNA, 2013
SSJ23-05
Artifact-suppressed, Low-dose C-arm CBCT Imaging of Low-contrast Cerebral Lesions
Scientific Formal (Paper) Presentations
Presented on December 3, 2013
Presented as part of SSJ23: Physics (Non-Conventional CT Imaging)
Xiao Han MSc, Presenter: Nothing to Disclose
Satoru Oishi PhD, Abstract Co-Author: Employee, Toshiba Coporation
Tetsu Satow MD, Abstract Co-Author: Nothing to Disclose
Hiromichi Yokoyama RT, Abstract Co-Author: Nothing to Disclose
Masanobu Yamada RT, Abstract Co-Author: Nothing to Disclose
Michael D. Silver PhD, Abstract Co-Author: Employee, Toshiba Corporation
Yu-Bing Chang, Abstract Co-Author: Nothing to Disclose
Emil Y. Sidky PhD, Abstract Co-Author: Nothing to Disclose
Xiaochuan Pan PhD, Abstract Co-Author: Research Grant, Koninklijke Philips Electronics NV
Research Grant, Toshiba Corporation
Consultant, UtopiaCompression Corporation
Three-dimensional images of a patient brain can be obtained by use of a C-arm-based CBCT system for clinical evaluation of cerebral lesions of low-contrast such as Intracranial Hemorrhage (IH) to surrounding soft tissues. Current CBCT systems employ FDK-based algorithms for yielding brain images, which require data acquired at a large number of projection-views and thereby incur a high level of radiation dose. In addition, FDK-based reconstructions may be susceptible to noise and shading artifacts, which can mimic or obscure low-contrast lesions. In this work, we develop an optimization-based algorithm for reconstructing C-arm CBCT brain images, with specific objectives of suppressing artifacts and significantly lowering radiation dose.
A clinical C-arm CBCT system was used for collecting brain data of patients at 607 views over 200° in 20 seconds. We refer to the acquired data as the full-view data, from which we formed a half-view data set by removing one projection frame at every other view. An iterative algorithm, referred to as ASD-POCS, was adapted to fully incorporate calibration information characterizing the actual scan geometry, which deviates from a circular trajectory due to gantry wobble. We applied the adapted ASD-POCS algorithm to the half-view data, and compared the reconstructions to the FDK reconstructions from full- and half-view data sets.
The half-view ASD-POCS reconstructions show suppressed artifacts than both full- and half-view FDK reconstructions. The soft-tissue contrast of the half-view ASD-POCS reconstruction is superior to the half-view FDK reconstruction, and is visually comparable to that of the full-view FDK reconstruction.
Our new algorithm is capable of reconstructing from half-view data patient-brain images with reduced artifacts and comparable soft-tissue contrast than the full-view FDK reconstruction.
We have demonstrated a C-arm CBCT imaging technique with improved image quality at considerably lowered imaging dose for clinical evaluation of low-contrast lesions such as IH.
Han, X,
Oishi, S,
Satow, T,
Yokoyama, H,
Yamada, M,
Silver, M,
Chang, Y,
Sidky, E,
Pan, X,
Artifact-suppressed, Low-dose C-arm CBCT Imaging of Low-contrast Cerebral Lesions. Radiological Society of North America 2013 Scientific Assembly and Annual Meeting, December 1 - December 6, 2013 ,Chicago IL.
http://archive.rsna.org/2013/13026028.html