Optimization Online


Reconstruction of CT Images from Parsimonious Angular Measurements via Compressed Sensing

Necdet Serhat Aybat(nsa2106***at***columbia.edu)
Amit Chakraborty(amit.chakraborty***at***siemens.com)

Abstract: Computed Tomography is one of the most popular diagnostic tools available to medical professionals. However, its diagnostic power comes at a cost to the patient- significant radiation exposure. The amount of radiation exposure is a function of the number of angular measurements necessary to successfully reconstruct the imaged volume. Compressed sensing on the other hand is a technique that allows one to reconstruct signals from a very limited set of samples provided that the target signal is compressible in a transform domain. Combining the two gives clinicians the benefits of CT while at the same time limiting the risk posed by excessive radiation. In this work we formulate the computed tomography reconstruction problem within a compressed sensing framework using partial pseudo-polar Fourier transform. Simulated results indicate that the number of angular projections and hence the associated radiation can be cut by a factor of 4 or more without noticeable loss of image quality.

Keywords: Compressed sensing, Computed tomography, Pseudo-Polar coordinates, Linogram, Total variation norm

Category 1: Applications -- Science and Engineering (Biomedical Applications )

Category 2: Convex and Nonsmooth Optimization

Citation: Siemens Corporate Research, September 2008

Download: [PDF]

Entry Submitted: 06/25/2009
Entry Accepted: 06/25/2009
Entry Last Modified: 06/25/2009

Modify/Update this entry

  Visitors Authors More about us Links
  Subscribe, Unsubscribe
Digest Archive
Search, Browse the Repository


Coordinator's Board
Classification Scheme
Give us feedback
Optimization Journals, Sites, Societies
Mathematical Programming Society