CT Flashcards
X-ray attenuation data through a single slice of a patient through a range of different angles
Tomo = slice
Array of detectors that then digitise the image
Each point or PIXEL in this image, has an attenuation value = Hounsfield Unit
Standardised to water = 0
Balancing image quality and radiation dose
CT Machine: XRAY Production
Filament within the cathode - NEGATIVELY charged electrode
Tube current = electrons being accelerated across to the POSITIVE ANODE
Tube potential = determines NUMBER and ENERGY of the electrons (kVp)
Focusing cup - focuses -e’s onto focal spot
UNBIASED focusing cap, where the current to the filament and the current to the cathode are linked
At focal spot
99% - THERMAL interactions
1% - XRs
Deals with heat by rotating
Tungsten - high melting point
Cooling
Anode angle = SMALLER effective focal spont
Line focus principle equation
XRs generating by Bremsstrahlung radiation
Incident -e experiences ELECTROSTATIC force of the positive nucleus
Loss of energy = release of an XR
Effect is inversely proportional from the distance to nucleus
This forms a Bremsstrahlung spectrum
INVERSE LINEAR RELATIONSHIP WITH kVp
XRs generating by Characteristic radiation based on difference in binding energies
K alpha from one shell above
K beta from two shells above
Unfiltered XR spectrum
Filtered XR spectrum - lose more LOWER energy, max and k edges remain the same
XR Detectors - Can be direct or indirect
INDIRECT
Incident X Ray converted to light by a scintillator layer e.g. Gadolinium Oxysulfide
Photodiode layer converts light into electrical signal
ALL PROPORTIONAL
XRs -> Light -> Photodiode Current
ADC
DIRECT
XRs into electrical current
Collimator - narrows down the incident beam e.g. Lead REDUCING SCATTER
Grid Ratio = H / W
H = HEIGHT of grid
W = WIDTH BETWEEN them
Higher grid ratio = more scatter attenuated
t = width OF THE GRID
GRID FREQ = 1 / (W + t)
Filtration
PROBABILITY OF PE’S
Non-uniform to filter to ensure more uniform irradiation
HVL = thickness of tissue to attenuate the XR beam by 1/2
Beam shaping filter ACCOUNTS FOR DIFFERENCE IN THICKNESS
So differences in detection is purely based on ATTENUATION of the tissue
Bow tie filter
SLIP rings conduct electricity, provide power to the X ray tube.
CT Generations
1st Gen -
PENCIL BEAM = Highly collimated
Translate Rotate
Single Detector
Translates across the patient filling one line in the band
It’s going to take by one degree and then translate again across the patient
“Translate and rotate”
2nd Gen
Narrow Fan Beam (Collimated in Z direction only)
Translate-Rotate
SINGLE ROW detector array with upto 30
Beam is no longer parallel, now fanning, so bits left out
Overlap now required as you translate/rotate
3rd Gen - Most commonly used
Fan beam
Rotate-Rotate Geometry because of slip rings (both source and detectors rotate together)
MULTIROW CURVED detector array
3rd Gen - UNIQUELY SUFFERS RING ARTEFACT due a SINGLE faulty or miscalibrated detector
4th Gen - deals with ring artefact
Fan Beam
Rotate-Station (Source rotes around a static ring of detectors)
Multi-row 360 degree array
No longer susceptible to ring artefact as one faulty detector can be averaged out
CT Beam Geometry
Fan Angle - 15-30 degrees in 3rd gen
Cone angle - angle formed by beam width
Beam width matches row of detectors in z direction
Slice thickness = width of the detector
Can be summated to count as a THICKER slicer
Isocentre - point that is the centre of the rotational axis - NOT necessarily half way away from the source to detector
Maximum FOV - where you can most accurately acquire attenuation data
Magnification
The magnification factor at the Isocenter takes
source to DETECTOR distance and divided by
source to ISCOCENTRE distance
So due to magnification of the divergent beam, slice thickness from the isocentre and NOT necessarily the detector width
Helical/Spiral CT acquisition
Pitch = (table speed x rotation time)/isocentre beam width
Pitch = 1 No gaps
Pitch >1 = Gaps but less patient dose
Spiral data is compressed and slices are INERPOLATED
With higher pitches - more gaps
but the data in the gap can be interpolated from the two adjacent slices
Helpful as Dose α 1 / pitch
Adaptive beam collimation to prevent “half xrays” that won’t contribute to image at either end reduces the dose in helical equipment
Linear attenuation coefficient = attenuation through PE or scatter (R or C)
Rest will be transmitted
Fraction of attenuation of a MONOCHROMATIC beam over a KNOWN distance = LAC
Less attenuation:
High energy
Lower density
Lower atomic number of tissue
Lower electron density
Attenuation = fraction removed over every unit distance = exponential decrease
You can lower the LAC
- Reduce kVP/photon energy - MORE PE + Scatter
LAC CHART changes depending on xray intensity
The HU is standardised to water. So the HU for water = 0
Window Width = range of HU represented
Window Level = Centre point of range
Narrowing down the window allows
CT actually uses a polychromatic beam = some tissues will preferentially attenuate the lower energy photons.
The emergent beam can then have a HIGHER AVERAGE energy = BEAM HARDENING
Simple back projection - reconstruction using multiple views - but has smearing / blurring of data
But you can filter it to remove the blurring
Frequency domain
High frequency areas of image = where there is an abrupt change in attenuation values .e.g at boundaries / details
You can do a 1D or 2D fourier transform
You apply a RAMP filter to reduce the LOW FREQUENCY data - amplifies edges
