The Assistance of Tomosynthesis Image Reformation using MATLAB Image Processing Tool-box for Department of Radiology.
| Autor: | Chih-Sheng Yeh, 葉志盛 |
|---|---|
| Rok vydání: | 2011 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 99 This study was using the tomosynthesis technique and the ACR CT accreditation phantom to verify the relationship of the swing angle of x-ray tube and sectional thickness. In addition, the grayscale linearity, the reconstruction time for different algorithm, the calculated volume accuracy, and the effectiveness of images reconstructed along various directions were evaluated. The tomosynthesis technique in flat panel digital X-Ray system uses similar “twinning principle” theory to tomography to produce serial sectional images of human body. The grayscale linearity of reconstructed images was similar to the results from CT images. The pixel size of reconstructed tomosynthesis images was 0.298 mm*0.298 mm. The reconstruction time for 41 images using filtered back projection was 27 seconds faster than shift and add algorithm. The sectional thickness of each tomosynthesis image was able to reach as thin as 2 mm using 40 degrees swing angle of x-ray tube. In the analysis of the agreement of using fluoroscopy or tomosynthesis to identify the relative location between instrument and specific tissue, the result of Wilcoxon signed-rank test, the p-value, was 0.0016. This implies that using tomosynthesis images in localization of instrument in diagnosis and/or treatment was superior to using conventional superimposed images. The relative distance can be calculated from these images. Therefore, during the fluoroscopy guided diagnosis or treatment, this technique provides additional spatial information to the radiologists and improve the accuracy of localization. With the helpful of these images, the radiologist can have more accurate spatial information during fluoroscopy guided procedures, such as percutaneous gastrostomy, and may shorten the process of operation. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|