Thesis flyer
What is CT?
CT stands for computed tomography. A clinical CT scanner is shown in Fig. A. Figure B shows the working principle, by which x-ray beams pass through the body to be detected as projections. A set of projections across various angles forms a tomography (or a sinogram). Figure C shows a very simplistic image reconstruction problem where the measured projection data: {7,2,5,4}, and the unknowns { x1, x2, x3, x4} obey the following:
x1 + x2 = 5
x1 + x3 = 7
… and so on
This reconstruction problem can be resolved by simply solving a set of simultaneous equations. Unfortunately, the most common scenarios involve 106 unknowns and a greater number of measurements! More complicated reconstruction algorithms have to be used (the most common being filtered back projection or FBP).
What is micro-CT?
Clinical CT has a resolution of ~1 mm. Micro-CT has a resolution of ~1 µm (i.e. a magnification factor of 1000).
What limits the imaging capability of micro-CT?
Scan time is extremely long (may be > 6 hours per scan). Also a mathematical property demands that, for exact reconstruction, the x-ray beam must cover the entire sample. This restriction severely limits the allowable sample size?
What is compressed sensing (CS)?
CS is a new mathematical theory that allows significant improvements in CT reconstruction by the use of certain iterative reconstruction algorithms like SART-TV (simultaneous algebraic reconstruction technique, with total variation minimization).
What is the contribution of this thesis?
This thesis aims to show that:
Compressed sensing (CS) based reconstruction methods can enhance the imaging capability of micro-CT scanners._