| 講演抄録/キーワード |
| 講演名 |
2009-01-19 13:30
[ポスター講演]Phase Unwrapping Method for Phase-Contrast X-Ray CT Using AR-Model ○Takaaki Horii・Qingkai Huo・Tadanori Fukami・Tetsuya Yuasa(Yamagata Univ.)・Jin Wu・Thet-Thet- Lwin・Tohoru Takeda(Univ. of Tsukuba)・Kazuyuki Hyodo(High Energy Accelerator Research Organization)・Akio Yoneyama(Hitachi)・Takao Akatsuka(Yamagata Univ.) MI2008-89 |
| 抄録 |
(和) |
Phase-contrast CT tomographically delineates biological soft tissues, which cannot be clearly imaged using absorption contrast CT, at high contrast and high spatial resolution. Since the phase-contrast CT extracts phase-shift information from interferograms acquired using the x-ray interferometer, phase unwrapping is required in order to generate projections for CT reconstruction. In this research, we proposed a novel phase-unwrap method based on 2-D AR (Auto Regressive) model as follows: First, AR parameters modeling sinogram are estimated using pixels in partial regions unwrapped accurately with a conventional unwrap method. Next, based on the AR model, we detect sites of phase jump in the wrapped sinogram, and then to eliminate the phase jump by adding 2n (n is an integer) so that the residuals are as close to zero, i.e., an average of the residual-distribution, as possible. We investigated the efficacy of the algorithm by simulation data based on actual data acquired at KEK. |
| (英) |
Phase-contrast CT tomographically delineates biological soft tissues, which cannot be clearly imaged using absorption contrast CT, at high contrast and high spatial resolution. Since the phase-contrast CT extracts phase-shift information from interferograms acquired using the x-ray interferometer, phase unwrapping is required in order to generate projections for CT reconstruction. In this research, we proposed a novel phase-unwrap method based on 2-D AR (Auto Regressive) model as follows: First, AR parameters modeling sinogram are estimated using pixels in partial regions unwrapped accurately with a conventional unwrap method. Next, based on the AR model, we detect sites of phase jump in the wrapped sinogram, and then to eliminate the phase jump by adding 2n (n is an integer) so that the residuals are as close to zero, i.e., an average of the residual-distribution, as possible. We investigated the efficacy of the algorithm by simulation data based on actual data acquired at KEK. |
| キーワード |
(和) |
位相コントラスト / X線CT / 位相アンラップ / ARモデル / / / / |
| (英) |
Phase contrast / X-ray CT / Phase Unwrap / AR-model / / / / |
| 文献情報 |
信学技報, vol. 108, no. 385, MI2008-89, pp. 139-144, 2009年1月. |
| 資料番号 |
MI2008-89 |
| 発行日 |
2009-01-12 (MI) |
| ISSN |
Print edition: ISSN 0913-5685 Online edition: ISSN 2432-6380 |
著作権に ついて |
技術研究報告に掲載された論文の著作権は電子情報通信学会に帰属します.(許諾番号:10GA0019/12GB0052/13GB0056/17GB0034/18GB0034) |
| PDFダウンロード |
MI2008-89 |
|