Computed Tomography

Computed Tomography

imaging using x rays and computer processed measurements to produce cross-sectional images acquired in axial plane

Why is CT used?

projection radiography limited - structures superimposed, good differentiation of structures with low contrast resolution

CT Detector

converts x-rays into electrical signals

Xenon Detector

60-70% efficiency

Solid State Detector

98% efficiency

CT Detector Characteristics

small - better spatial resolution
high detection efficiency and high sensitivity
fast response time with negligible afterglow
high stability to function under high speeds of rotating gantry

Modifications of CT Tube

larger and thicker
absorbs and dissipates large amounts of heat
modified cathode assembly produces smaller focal spot
metal envelope and ceramic insulators
flying focal spot

Attenuation

measure of x-ray penetration through material, quantifies how much beam is weakened by material

Attenuation Coefficient

fraction of x-ray absorbed or scattered per unit thickness

What is the purpose of Slip Ring Technology?

transmits power and electrical signals from stationary structure to rotating structure, gantry can rotate without wired connections

Ray Sum

transmission of x ray beam through body section

Attenuation profile

all the ray sums coming from the volumes of material in a rotation

Back Projection

reconstruction algorithm producing cross sectional images, creates stripes of grey associated with certain attenuation

What data does back projection use to produce images?

attenuation profile, angle data was acquired in

Filtered Back Projection

altering projection data before the back-projections, fixes blurring problem in standard back projection

Sharpening Filter

picks up sharp edges in algorithm within projection and subtracts out extra smearing caused by back projection

Image Filtering

using algorithms to increase sharpness/smoothness

What effect is created when a sharpness filter is increased too much?

increasing sharpness filter too much increases noise

Iterative Reconstruction

image reconstruction algorithm that begins with image assumption and compares it to real time measured values while making constant adjustment

Iterative Reconstruction Process

computer expects generic set of data common to anatomy being scanned based on past images, uses real time measured values to build on existing data

Benefits of Iterative Reconstruction

quicker acquisition as is building on already present data
widely used due to improvement of computer processing power
overcomes noise associated with filtered back projection - improves image quality reduce artifacts and decrease radiation dose

Image Matrix

square matrix of pixels representing tissue

Pixel

picture element

Voxel

volume element

Hounsfield Units

linear scale of grey scale values (densities) based on measured attenuation coefficients, CT numbers

How is CT Image Displayed

each voxel represents section of body part imaged and has associated number based on the x ray attenuation of that section

Benefits of CT Numbers/Hounsfield Units

removes subjectivity of the level of grey
allows some insight of tissue make up of anatomy

CT Windowing

greyscale manipulated via CT numbers, change appearance of image to highlight structures

Window Width

range of CT numbers in an image, controls contrast

Small Window Width

short grey scale
small block of CT numbers assigned grey levels
small transition zone of white to black level
centred near average CT no. of organ interested

Large Window Width

long grey scale
large block of CT number assigned grey levels
large transition zone of white to black
used where large latitude is required
used to simultaneously display tissues greatly differing in attenuation

Window Level

midpoint of CT number range, controls brightness.

Helical CT

x ray tube continuously rotating in same direction within gantry whilst patient is continuously moving along z axis

Axial CT

tube stops and rotates to acquire data from single slice, patient movement stopped and started

Helical CT Requirements

slip ring technology
high power x ray tubes
interpolation algorithms

Helical CT Advantages

faster
avoids motion artefact from breathing
controlling pitch can reduce scan time and radiation dose
can control slice thickness
more effective use of contrast

Pitch

measure of overlap through scanning, defined by speed of table movement through the gantry

Pitch Calculation

table travel per rotation/x ray beam width

Higher Pitch

pitch number greater than 1, table travels more than width of beam, gaps between data acquired

Benefits of Higher Pitch

lower dose because missing sections of patient, quicker scan

Limitations of Higher Pitch

lower image quality as fewer projections obtained

Lower Pitch

pitch number less than 1, table travels less than width of beam, overlap in data acquired

Advantage of Lower Pitch

better z-axis resolution so better image quality

Limitation of Lower Pitch

higher patient dose

Interpolation

image reconstruction mechanism that estimates values using known data from nearby points
creates virtual slices
converts helical path to transverse slices

Advantage of Interpolation

data not missed without patient dose increasing

Disadvantage of Interpolation

overall image will be diagnostic but not of same quality as if directly acquired

Interpolation Algorithm

mathematical process required to reconstruct axial images from spiral volume data set

Slice Thickness

determines trade-off in image quality between spatial resolution and image noise determined by collimation and detector selection

Thinner Slice Thickness

increase in spatial resolution but increases image noise

Acquired Slice Thickness

thickness of each slice set in scan parameters

What is the thickness of the smallest acquired slice?

smallest thickness cannot be less than smallest detector size

Reconstruction Slice Thickness

reconstructing acquired slices into thinner slices, depends on acquired slice thickness, determined in reconstruction parameters

Slice Interval

distance between centres of adjacent slices, determines number of images in series

Contiguous Slices

interval equals thickness, no anatomy missed

Non-contiguous Slices

interval greater than thickness, some anatomy missed, less images in series

Overlapped Slices

interval less than thickness, some anatomy shown in adjacent slices, more images in series

MDCT

Multi Detector CT, multiple slices in one acquisition.

Advantages of MDCT

faster screening time - fewer motion artefacts
reduced patient risk - less time on table
longer scans possible - more slices acquired per rotation
less sedation
less contrast
thinner slices - improved z axis resolution

Uniform Detector

all detector rows have same width

Non-uniform Detector

central rows thinner, wider towards edges fewer septae improves dose efficiency

Hybrid Detector

same width detectors but central detectors narrower than outer rows

Beam Pitch

table distance traveled per rotation divided by total thickness of all simultaneously acquired slices

Cone Beam Acquisition

more detectors and more slices create wider beam width, cone beam required to cover whole detector width

Cone Beam Artefacts

misrepresentation occurs due to angle of beam as it diverges

When are Cone Beam Artefacts more likely to occur?

when objects caught by the periphery of the beam

Titled/Oblique Reconstruction

produces non-axial images filtered into standard axial.

Feldkamp Algorithm

uses 3D back projection to reconstruct cone beam acquisition
divided into 3D voxels not 2D matrix of pixels

Good Image Quality

measure of how suitable an image is for its intended diagnostic purpose
visibility of anatomical structures, various tissues and signs of pathology

What makes good image quality?

good low contrast detectability
low noise
good high spatial resolution
artefact free

Desirable Attributes of CT Imaging

good image quality, fast scanning

Factors Affecting Spatial and Contrast Resolution

scanner design (hardware and software)
scan acquisition parameters
reconstruction parameters
patient factors

Factors Only Affecting Perception of Image

post-processing parameters
image viewing conditions
display resolution
observer performance

What is the effect of Scanner Design on resultant image?

influences spatial and contrast resolution

Scan Acquisition Parameters

Settings that affect image capture quality.

Reconstruction Parameters

Settings that influence image reconstruction quality.

Radiation Dose Trade-off

keeping patient dose as low as possible whilst maintaining diagnostic image quality

Thicker Slices

lower noise but lower spatial resolution

Post-processing parameters

settings applied after image acquisition to enhance quality

Contrast Resolution

ability to distinguish differences in image shades

Low Contrast Detectability

ability to detect an object with small difference in attenuation coefficient from its homogeneous background influenced by image noise

Noise

variation in CT numbers not related to true attenuation coefficient
due to not enough photons reaching detector

Stochastic noise

from random variations in detected photon numbers

Electronic noise

noise from measuring system

Structural noise

noise from reconstruction algorithm

Quantifying Noise

measured as standard deviation of pixel values in uniform medium

What does a more standard deviation mean in terms of noise?

higher noise has more SD and spread of values

Factors Affecting Noise

scanner specifications and designs
scanning acquisition parameters
reconstruction parameters
patient factors

Factors Affecting Noise - Scanner Specifications and Designs

efficiency of detectors
x ray beam filtration
scanner geometry

Factors Affecting Noise - Scanning Acquisition Parameters

tube voltage/current
scan time
slice thickness
pitch

What is the relation between mAs and kV and noise?

increasing mAs and kV reduces noise

Factors Affecting Noise - Reconstruction Parameters

back projection algorithms, noise filters

Factors Affecting Noise - Patient factors

patient size, to maintain constant noise mAs must be doubled for every extra 4cm of tissue

Tube current

amount of current affects photon production

Spatial resolution

measure of how far two objects must be apart before they can be seen as separate details in image
differentiating between two points close together

Line pairs

measured in lp/cm, indicates spatial resolution

Transaxial resolution

resolution across patient's axial plane

Z-axis resolution

measure of sensitivity of scanner to objects along z-axis
affects visualisation of small objects
important for 3D reconstructions
affected by slice thickness

What is the effect of small focal spot size on image resolution?

gives higher transaxial resolution

What is the effect of large focal spot size on image resolution?

increase geometrical unsharpness - penumbra