Basic physics of multidetector computed tomography ( CT Scan) - how ct scan works, different generations of ct, how image is generated and displayed and image artifacts related to CT Scan.
2. TOPICS TO BE COVERED
1. Basic Principles
CT concept
2. Generation of CT image
Scan/ data acquisition
Reconstruction
Display
3. Image Quality
Quantitative measurements
Image artifacts
3. Basic principle
CONCEPT:
Internal structure of an object can be
reconstucted from multiple
projections of the object.
CT image is a display of the anatomy of a thin slice of the body developed
from multiple x-ray absorption measurements made around the body’s
periphery.
Image in conventional tomography blurring out the information from
unwanted regions
In CT constructed using data arising only from section of interest.
Routine CT generated axial image image is reformatted to coronal /
sagittal.
4. Generation of CT image
Scan / data acquisition:
Components:
Scan frame:
early Ct scanners – rotating frames and recoiling system cable step and shoot
methods
Current systems slip rings (transmitting electrical energy across rotating)
X ray Generators:
operating frequencies 5 to 50 kilohertz
power 15 to 60 kW
80 to 140 kV and 30 to 500 mA
Computer modulated generators:
1. ability to obtain high x-ray flux when needed
2. provision for tube heat management
3. ability to change the current as needed to maintain high image quality within
the context of ALARA.
5. X-ray Tubes:
rotating anodes, unique cooling methods
2 MHU with 300kHU/min 7 MHU with 1 MHU/min
enhance system’s ability to cover large areas of anatomy at
diagnostic levels of x-ray output.
Double tube designs:
increase x-ray source, faster acquisition time
spectral absorption differences of tissues
characterising tissues and abnormalities – fat content, stones.
Data acquisition system (heart of CT system):
1. Detector system
2. Analogue to digital conversion
3. Data processing
6. Detectors:
1. Scintillation crystals:
produce light when ionizing radiation reacts with them.
Scintillation detector
1st
2 gen used Thallium-activated NaI
Disadv- hygroscopic and very long afterglow.
Replaced by silicon photodiodes (a/c solid-state)
CeI, gadolinium oxysulfide and Cd tungstate. ( no afterglow)
2. Xenon Gas Ionization Chambers:
limited to use in rotate-rotate type scanners.
Should have- anode and cathode, inert gas, voltage, walls
separate, window
disadv- inefficiency
xenon (heaviest), compressing to 8 – 10 atm, long chamber size.
10. MDCT sequential (axial) scanning
Using sequential (axial) scanning, the scan volume is covered by
subsequent axial scans in a ‘‘step-and-shoot’’ technique.
In between the individual axial scans the table is moved to the next
z-position.
With the advent of MDCT, axial ‘‘step-and-shoot’’ scanning has
remained in use for only a few clinical applications, such as
- head scanning,
- high-resolution lung scanning,
- perfusion CT
- interventional applications
11. MDCT spiral (helical) scanning
Spiral/helical scanning is
characterized by continuous
gantry rotation and
continuous data acquisition
while the patient table is
moving at constant speed.
Advantages:
1. Minimize motion artifacts
2. Decreased incidence of mis-
registration between
consecutive axial scans
3. Reduced patient dose
4. Improved spatial resolution
in z- axis
5. Enhanced multi-planar or 3D
rendering.
12. For general radiology applications, clinically useful pitch
values range from 0.5 to 2.
For the special case of ECG gated cardiac scanning, very low
pitch values of 0.2–0.4 are applied to ensure gapless volume
coverage of the heart during each phase of the cardiac cycle.
13. Reconstruction Process
A cross sectional layer of body is divided into many tiny blocks.
Then each block is assigned a number proportional to degree that
the block attenuated the x-ray beam voxel
Their composition and thickness, along with the quality of the
beam, determine the degree of attenuation.
Linear attenuation coefficient (µ) is used to quantitate
attenuation.
14. Algorithms for Image Reconstruction
1. Back projection
2. Iterative methods
3. Analytical methods
BACK PROJECTION
oldest method. Not used now, but easiest to explain, prototype
Block is scanned from top and left side.
Heights of steps is proportional to the amount of radiation that passed
through the block.
Rays from two projections are superimposed or back projected , they produce
a crude reproduction of the original object.
In practice, many more projections would be added to improve image quality,
but the principle is the same.
15.
16. ITERATIVE METHODS:
• Starts with an assumption ( eg all points in the matrix have the same value)
and compares this assumption with measured values, make corrections to
bring the two into agreement, and then repeat the process over and over
until the assumed and measured values are the same or within acceptable
limits.
• Three variations:
1. Simultaneous reconstruction: all projections for the entire matrix are
calculated at the beginning of the iteration, and all corrections are made
simultaneously for each iteration.
2. Ray-by-Ray Correction: one ray sum is calculated and corrected, and these
corrections are incorporated into future ray sums.
3. Point-by-Point correction: calculations and corrections are made for all rays
passing through one point, and these corrections are used in ensuing
calculations, again with the process being repeated for every point.
17. Analytic methods:
used in almost all CT today.
Differ from iterative methods in that exact formulas are utilized for the
analytical reconstructions.
Types:
1. Two-Dimensional Fourier Analysis:
basis any function of time or space can be represented by the sum of various
frequencies and amplitudes of sine and cosine waves.
Ray projections with squared edges, are the most difficult to reproduce.
18. 2. Filtered Back projection:
is similar to back-projection except that the image is filtered, or modified to
exactly counterbalance the effect of sudden density changes, which caused
blurring (the star pattern) in simple back-projection.
Those frequencies responsible for blurring are eliminated to enhance more
desirable frequencies.
Inside margins of dense areas are enhanced while the centres and immediately
adjacent areas are repressed.
22. Display:
CT number (HU)
After a CT scanner reconstructs an image, the relative pixel values represent the
relative linear attenuation coefficients.
A CT numbering system that relates a CT number to the linear attenuation
coefficients of x-rays has been devised.
Where
K is magnification constant
µw attenuation coefficient of water
µ is the attenuation coefficient of the pixel in question.
Window level: the position describing the center of scale.
Window width: the range of CT numbers selected for gray-scale amplification.
23.
24. ADVANCED DISPLAY
Multiplanar reformatting
conventional CT study consists of several contiguous axial images
perpendicular to the long axis of the body,- coronal or sagittal images are not
possible except when the gantry is tilted or the body is positioned to show
the image in the desired plane.
This limitation of CT can be overcome by image manipulation commonly
referred to as multiplanar reformatting (MPR).
In this process, image data are taken from several axial slices and are
reformatted to form images.
25. In this viewing mode, the user defines the number of imaging planes and their
position, orientation, thickness, and spacing, and the reformatted image is
displayed in sagittal, coronal, or oblique planes.
26. 3D Shaded Surface Reconstruction
For a surface reconstruction, the user selects a threshold range. This allows
the user to select only the tissue (e.g., bone) to be rendered. The voxels with
Hounsfield values within the threshold range are set to the “on” state,
whereas the rest of the voxels are set to the “off” state.
The second step is to project rays through the entire volume. As the rays pass
through the data, they stop when they identify the first “on” voxel. For that
particular ray, this first “on” voxel is part of the surface; the other voxels are
ignored. This is done for all the rays, and all of the “on” voxels are used to
create the surface.
27. A, Sample of eight-voxel data set with displayed CT number values. B, User-
defined threshold of CT number values for tissue definition.
28. Three-Dimensional depth based shading
With this method, those voxels that are closer to the viewer are illuminated
at a greater intensity than those that are farther back.
As with sagittal and coronal images, the quality of the image is improved by
interpolating the data between slices to form a smooth, continuous image.
When various three-dimensional views are selected sequentially in time, the
image can appear to rotate.
29. Volume Rendering
technique that displays an entire volume set with control of the opacity or
translucency of selected tissue types. In this case, each voxel has an associated
intensity in addition to an associated opacity value.
advantage of VR over three-dimensional shaded-surface rendering is that it
provides volume information.
By using transparency, the operator can visualize information beyond the surface.
For example, VR is the preferred stent evaluation because stents can be made
transparent to view the lumen of vessels.
Other - ability to make plaque transparent for more accurate diagnosis of vessel
stenosis.
30. MIP (Maximum Intensity Projection)
Unlike three-dimensional shaded-surface and VR displays, no preprocessing is
required.
The rays are cast throughout the volume, and depending on whether it is maximum
intensity projection or minimum intensity projection, maximum or minimum values
along the rays are used in the final image display.
Using maximum intensity projection (MIP) for visualization permits easy viewing of
vascular structures or air-filled cavities.
MIP enables easy viewing of an entire vessel in one image.
This is because voxels representing the contrast-filled vessels are most likely to be
the ones with the highest values along the ray (assuming no bone along the ray).
Along the same lines, minimum intensity projection can be used to demonstrate
air-filled cavities.
31. Image Quality
Quantitative Measurements
Spatial Resolution:
measured by the ability of a CT system to distinguish two small, high-contrast
objects located very close to each other under noise-free conditions.
required for evaluating high-contrast areas of anatomy,
such as the inner ear, orbits, sinuses, and bone in general, because of their
complicated shapes.
Spatial resolution can be specified by spatial frequencies, which indicate how
efficiently the CT scanner represents different frequencies. Modulation
transfer function (MTF) describes this property
32. Filter effects on resolution
the major role of the convolution filter is to remove the image blurring
created by the back-projection process.
Various filters control the amount of image blurring created by accentuating
high-frequency components found in the data.
For a crisp image, the high spatial frequencies are accentuated, and this has
the effect of sharpening the edges and improving resolution.
One pays for a crisp picture with a decrease in density resolution. Similarly,
by increasing density resolution, one pays by loss of some spatial resolution
and image crispness.
33. Opening size of Detector Aperture
The detector aperture MTF curve depends on the magnification
factor of the system and the physical size of the detector.
If the object being viewed is smaller than the width of the data
ring, it will be difficult to resolve because it occupies only a
fraction of the space seen by the detector.
Typical detector apertures of CT systems today range from less
than 1 mm to 1.5 mm, with center-to-center detector spacing
of approximately 1 mm.
34. Factors Affecting Spatial Resolution
Focal Spot Smaller focal spot, SR improves
Detector width Smaller Detector Width, SR improves
Number of Projections More projections, SR improves
Slice thickness Smaller ST, SR improves
Pitch Lower pitch, SR improves
Pixel Size Smaller Pixel Sixe, SR improves
FOV Decreasing FOV(everything else constant), SR
improves
Patient Motion Decreased Patient motion, SR improves
35. Pixel Size
the spatial resolution can be no greater than the size represented by the pixel
length.
In reality, pixel size should be 1.5 to 2 times smaller than the desired
resolution.
Unless a matrix element exactly coincides with an object, the object
representation will be averaged over two or more pixels and thus may not be
visualized.
It must be realized that the pixel size refers to the FOV (or body), not the
viewing screen or film.
36. Contrast Resolution
ability to differentiate the attenuation coefficients of adjacent areas of
tissue.
In the computation of any single pixel value, there is error in the form of
statistical variation; it is this variation that limits the ultimate contrast
resolution.
This variation (called image noise) is manifested as a grainy background, or
mottle.
The parameter used to evaluate this variation is the standard deviation (SD).
37. NOISE can be reduced by
Increasing tube voltage, tube current, scan time, FOV &Slice thickness
Using reconstruction filters
38. Factors Affecting Contrast Resolution
mAs More mAs, CR improves
Pixel Size FOV and pixel size increase, CR improves
Slice thickness ST increases, CR improves
Reconstruction filter Using Soft tissue improves CR
Patient Size For larger patients, at same technique, more
attenuation, detected photons decreases, CR
degrades.
39. Temporal resolution
refers to the ability of a CT scanner to capture objects that change shape or
position over time and depends primarily on the gantry rotation speed and the
reconstruction method used.
depends on:
1. gantry rotation speed
2. spiral interpolating algorithm used during reconstruction.
40. Ring Artifacts
usually the result of difficulty with the detector.
each detector is associated with a data ring. A malfunction of any one
detector incorrectly back-projects along the data ring to produce the ring
artifact.
If a detector is not matched or is not intercalibrated accurately, the back-
projection for each data ring will be slightly different, causing multiple rings.
Detectors in the center of the detector arc are most sensitive.
41. Metal and Bone Artifacts
The presence of objects having an exceptionally high or low attenuation can
create artifacts by forcing the detector to operate in a nonlinear response
region.
Because this incorrect response occurs at specific directions of the beam
through the object, incomplete cancellation of the back-projected rays during
reconstruction occurs and yields streaking artifacts.
42. Beam-Hardening artifacts
result from the preferential absorption of low-energy photons from the beam.
average beam photon energy is progressively increased.
toward the end of the x-ray path, the attenuation is less than at the
beginning because the attenuation coefficient is smaller with higher energy.
The reconstruction program, however, assumes a monochromatic beam and
attributes any change in beam intensity to a change in tissue composition
rather than to the result of a shift in average photon energy.
The assigned attenuation coefficients are thus in error, and the densities seen
on the image are in error.
The effect is most pronounced in regions of large attenuation, such as bone.
43. Stair-Stepping Artifact
occur when in one direction the pixel of the reformatted image has the same
length as the axial image
but in the other direction the pixel length is the same as the slice thickness.
Because pixel length in most scans is considerably smaller than slice
thickness, the reformatted scan has an unusual appearance.
Uncommon in modern CT.
44. Thank You
References :-
CT and MRI of the Whole Body: 5th
edition; HAAGA
Christensen’s Physics of Diagnostic Radiology.
Recent Advances: AIIMS, PGI, MAMC Series.
Volume CT: State-of-the-Art Reporting. AJR 2007;
189:528–534
Three-dimensional volume rendering of spiral CT data:
theory and methods. Radiographics 1999;745-764.
Developments in CT. Imaging, 18 (2006), 45–61