CT Flashcards
Recall the limitations of projection imagery
No depth info
Nom uniform background prevents visibility of details
Tissue characterisation not possible
Describe CT in a nutshell
Possible to reconstruct 2D distribution of a quantity of its a line integral
Rotate 360° line integrals to provide 3D images
Provides info about attenuation coeffs (tissue characterisation)
Done using Fourier transforms
What assumptions are there in CT
Parallel laminar beam
Monochromatic beam
Laminar detector
Negligible scatter
Beam and detector rotated around the sample
1D image collected at every angular position
Define a sinogram
Random transforms of the initial image
Each line is the projection obtained at a certain angle
What is backprojection
Info from projections ‘projected back’ onto the image plate and summed
Issues with backprojection
Centre always brighter (many projections crossing it)
What is Fourier slice theorem
The FT of a projection is evenly sampled along a line rotated by the same angle
Recall the three CT filters and their use
Ramp— high frequency (noise, small details)
Gen-Hamming—low frequencies (less noisy, lose small details)
Shepp-Logen—compromise
Define the spiral pitch
The distance covered by the patient bed while the gantry covers a full rotation
Recall the equation for CT number
CT = (mu_tissue - mu_water) / mu_water * 1000
Define: artifacts
A feature of the image that doesn’t result from a feature of the object
List system related artefacts
Beam hardening
Detector calibration/misfunctioning
Penrose volume effect
Dead pixels
Different pixel gain
Noise spikes
List patient related artefacts
Partial volume effect
Strong attenuation
Motion
How do you reduce beam hardening
Pre harden with Cu filter to get rid of low energy
‘Bow-tie’ filter to compensate for different thicknesses of patient
Calibration corrections
Solutions for attenuation artefacts
Modulated current
Scout image and adjust current
