Loading...

Performing an Efficient Architecture for Compressive Sensing Algorithms in CT Application

Abbasi, Hassan | 2015

765 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 47701 (05)
  4. University: Sharif University of Technology
  5. Department: Electrical Engineering
  6. Advisor(s): Shabany, Mahdi; Kavehvash, Zahra
  7. Abstract:
  8. Computerized tomography (CT) is a powerful tool among existing bioimaging techniques for capturing bio-images. In fact, CT imaging systems have attracted great attention in the last decades, because of their fast and high-quality reconstruction, low complexity and low-cost hardware solutions. In a CT scan procedure, linear sensors receive x-ray radiations, passed through the patient’s body and The quality of the reconstructed image is essentially influenced by the number of captured line projections. Nevertheless, gathering the adequate amount of data requires the patient to being exposed to X-ray radiations for a long time. However, the intensification of x-ray radiations could lead to ionization of body cells and in turn raise the risk of cancer. Thus, one of the most important challenges in using the CT technique for biomedical imaging is to reduce the required samples without degrading the image quality. Traditionally, image reconstruction requires a number of samples, which are dictated purely by Nyquist limits. Due to the Nyquist constraint, an image capturing procedure with fewer samples than the Nyquist rate leads to a performance degradation. So high dosage of X-ray radiotion is the main challenge of these algorithms. The recently proposed Compressed Sensing(CS) theory, introduces new methods to rectonstruct signals and images with few and incomplete amount of data set. In this thesis, A novel CT imaging structure based on CS is proposed. The main goal is to mitigate the CT imaging time and thus x-ray radiation dosage without compromising the image quality. The utilized compressive sensing approach is based on partial Fourier sampling. Thanks to the intrinsic relation between captured radon samples in a CT imaging process and the radial Fourier samples, partial Fourier sampling could be implemented systematically. This systematic compressive sampling helps in better control of required conditions such as incoherence and sparsity to guarantee adequate image quality in comparison to previous CS based CT imaging structures
  9. Keywords:
  10. Compressive Sensing ; Medical Imaging ; Image Reconstruction ; Partial Fourier Sampling ; Computerized Tomography

 Digital Object List

 Bookmark

...see more