The document summarizes the history and technology of computed tomography (CT) scanners. It describes how CT was developed in the 1970s by Godfrey Hounsfield and Alan Cormack, who were later awarded the Nobel Prize. It outlines the key innovations in each generation of CT scanners, from the first generation's pencil beam geometry to later generations' use of detector arrays and helical scanning, which reduced scan times. The document also discusses the components of a CT scanner, including the x-ray tube, detectors, and techniques for image reconstruction and calibration.

1. COMPUTED TOMOGRAPHY
SCANNER
SABARI KUMAR P
M.Sc. Radiation Physics

2. Historical Perspective
In 1917, Radon, an Austrian mathematician proved that image of
internal structure of a three-dimensional object could be reconstructed
from its multiple projections.
In 1963, Alan Cormack, a physicist from Tufts university developed the
mathematics used for reconstruction of image from multiple
projections.
In 1970s,Godfrey Hounsfield, an Engineer with Electro Musical
Instruments Ltd. (EMI), developed a scanner and demonstrated the
reconstruction of image from multiple projections.
Godfrey Hounsfield and Alan Cormack Shared Nobel Prize in Medicine
in 1979 for this invention.

3. Conventional radiographic image is a two-dimensional projection of
a three- dimensional object.
Hence in conventional imaging techniques, image contrast is
reduced by :
1.Superimposition of overlying structures and
2.Scatter from the body parts other than the region of clinical
interest, which cannot be avoided in projection radiography.
Computed Tomography (CT) scanners provides cross sectional
images of body parts.
Cross-sectional imaging maximally avoids the body parts other than
the area of clinical interest, and hence solves to a greater extent the
above cited limitations of conventional radiography.
Contrast resolution of CT is superior to that of conventional
radiography.

4. CT was the first modality that made it possible to probe the inner
depths of the body, slice by slice
The development of the CT scanner has been one of the most
explosive phenomena in modern disease diagnostic area
Tomography = Body section radiography
CT is a digital imaging technique involving the following steps:
1.Scanning of the patient through special movement of x-ray tube and
detectors
2.Conversion of transmitted x-rays, which carries projected information,
into electrical signals using detector system.
3.Sampling and digitization of analog signals from the detectors
4.Reconstruction of image and display

5. CT EQUIPMENT
Gantry
(X-ray tube &
detector
array)
Control
Console
Patient
Couch

6. Slice :
It is a cross sectional portion of the body which is scanned for the
production of CT image. It has width and hence volume
The CT image is presented as a
matrix of numbers and each
number represents the value
of the image at that location
The 3D volume element in the
image matrix is called Voxel
The 2D representation of Voxel
is called Pixel (Picture Element)
The process of CT image acquisition
involves the measurement of X ray
transmission profiles through a patient
for a large no. of views

7. Parallel Beam Geometry: all rays in projection are parallel to each other
Fan Beam Geometry: the rays at a given projection angle diverge
and have the appearance of a fan
Depending upon source-detector configuration and their relative
movement for scanning and data acquisition, CT scanners are classified
into different generations.
These modifications are aimed at reducing the patient scan time and
to reduce thereby the movement un-sharpness.
Reduced scan time is advantageous to increase patient throughput
too.

8. Features:
 Rotate and translate Mechanism
Single X ray detector used to
measure transmission of X rays
through the patient (in EMI scanner
two detectors)
Used Pencil Beam Geometry,
starting at a particular angle, the X
ray tube and detector system
translated linearly across the field of
view
The Scanning time about
4min/slice and 1min reconstruction
time
 Excellent Image quality (minimum
scatter inclusion
First Generation

9. Features:
 Movement is Rotate and Translate
 Linear detector Array of 30
detectors were used
 Increased the utilization of X ray
beam by 30 times compared to
First Generation Scanners
 Narrow Beam of 100 angle
Scan time 18 sec/slice and 15
times more faster than earlier
Use of narrow beam of X rays,
Scattered Radiation much
pronounced
Second Generation

10. Features:
No Translation Movement
Movement mechanism is Rotate
and Rotate
Mechanically coupled X ray tube
& Detector array rotate together
More than 800 detectors with
wide fan beam were used to
allow the X-ray beam to scan
entire body
Scan time reduced to 2sec/slice
 A single detector towards the
end of array acts as reference
detector.
Third Generation

11. A major drawback of 3rd generation
scanners is that failed detectors
produce ring artifacts.
A failed detector includes no
transmission signal along a beam
direction in every projections
As a result a circular region of high
attenuation appears in the image
Here,
It = intensity towards center of array
I0 = intensity near edge of reference detector
g1 = gain of reference detector
g2 = gain of measuring detector
If there is any electronic drift on one of the detectors i.e. g1≠g2, hence
the μt value changes and results Ring Artifacts in back projection
t
Ig
Ig
ln
eIgIg
t2
01
t
01t2






 

12. Fourth Generation
Introduced towards the end of 1970s mainly to overcome the
problem of ring artifacts produced by third generation scanners.
Features:
Rotate –Stationery
Detector ring all around the patient
(about 4800 detectors)
X-ray tube alone rotates inside the
detector ring
Each detector acts as its own
reference detector, results g1=g2=g
then transmission measurement
becomes
Scan time is reduced to about 1 sec
Patient dose is more compared to that in other generation scanners.
t
Ig
Ig
ln
t
0










13. Fifth Generation (Electron Beam CT )
No moving parts & No x-ray tube. Instead, a large arc of tungsten
encircles the patient
Detector ring is positioned directly opposite to the tungsten arc.
Electron beam from an electron source is steered through a wave guide
to strike the tungsten arc through a bending magnet to produce x-rays.
During the x-ray production , patient couch is moved to image different
body cross sections.
Capable of 50 ms scan time
and can produce fast-frame-
rate CT movies of beating
heart.
Special applications in
cardiology.

14. Helical/Spiral Scanner:
Earlier, scanning of successive slices was done by rotation of source
detector combination through 360 0 in one direction for one slice and
360 0 rotation in the reverse direction for the next slice and so on
(because of the constraint on motion imposed by wires connected to
stationery scanner electronics)
With the introduction of slip ring technology continuous rotation of
source-detector combination was possible in CT.
During the continuous rotation of source-detector system, the couch
advances the patient into the axial direction.
As a result x-ray beam follows a spiral/ helical trajectory, hence the
name spiral/helical CT.

15. During spiral CT, image data is received continuously.
However, when the image is reconstructed the plane of the image will
not contain enough data for reconstruction.
Therefore image reconstruction in spiral CT is done by the
interpolation of data available 3600 apart.
Advantages:
Shorter scan time (typically 30 seconds)
or even within a single breath-hold
of patient.
Improved patient throughput
In contrast study, since data can be
obtained during the peak of enhancement,
use of IV contrast can be optimized
From the transmission data acquired during the spiral scan, body
slices can be reconstructed at arbitrary intervals. Hence provides
improved lesion detection.

16. Multi Slice Helical CT scanner:
Multi slice CT employs multiple detector channels.
Facilitate effective utilization of x-ray output
Examination can be completed in a single scan with less thermal load
to the x-ray tube.
Conventional CT scanners with a single detector channel, employs
wider size detectors (about 15 mm) and slice thickness is determined by
adjustable detector collimators.
Multiple detector arrays are a set of several linear detectors (solid
state detectors) tightly kept close together.
Facilitates grouping of detector
elements to have a variety of
combinations of slice widths
and number of slices per scan
With multiple detector arrays,
slice width is determined by
the detectors not by the collimator

17. X-ray tubes used in CT scanners
 Exposure time is more in CT scanning. Hence heat storage and
dissipation is the major concern in the design of x-ray tubes used in CT
scanners.
Anode heating capacity must be at least 1 MHU for conventional CT
tubes and for spiral CT it must have at least 4 MHU capacity.
Massive anodes are used in the x-ray tubes to increase its heat storage
capacity. This increases the size of x-ray tube.
X-ray tubes used in spiral CT are especially bulky.
High speed anode rotation is used in CT tubes for efficient heat
dissipation.
In order to reduce the thermal input to the anode, some CT machines
employs pulsed x-ray beam.
For most of CT imaging protocols, energizing voltage of about 120 -
140 kV is used.
X-ray beam in CT is heavily filtered ( about 6 mm Al equivalent).

18. A Teflon bow- tie filter is also used
to match with the contour of the
body to reduce attenuation
un-sharpness.
(Due to Heal Effect)
Detectors used in CT Scanner:
NaI / CsI : Hounsfield’s first scanner. Excellent spatial resolution but
expensive photo multiplier and have not been made smaller than about
1cm in diameter
Xenon gas Detectors: high pressured Xenon gas detectors of 6cm thick
(to compensate relative low density and high geometric efficiency). But
these are directional dependency. So, couldn’t used from fourth
generation scanners
Solid State Detectors: scintillator coupled tightly to a photo detector.
Size of detector becomes 1.0mm to 1.5mm. Top of the detector is flat
and therefore it is capable of X ray detection over a wide range of angles

19. Detector Calibration:
The large no. of detectors used in a scanner cannot be mass produced
to have identical sensitivities
To overcome this, they are exposed with no patient in the beam and
the readings recorded by the computer, then it determines a calibration
factor for each
In fourth generation, this procedure carried out by making the beam
wider than then patient so that during one complete rotation, every
detector will see the source directly at least twice – thus allowing for a
calibration before & after each scan
In third generation, calibration must be made before or after the
patient is removed from the scanner

20. Pitch:
It is an important component of the scan protocol and it influences
radiation dose to the patient , image quality and scan time
In helical CT scanner with one detector array, The pitch is called as
collimator pitch and defined as
Collimator pitch = table movement per 3600 gantry rotation/collimator width at isocenter
Normally its value lies between 0.75 to 1.50
Greater pitch value benefits faster scan time, less patient motion
In helical CT scanner with multiple detector array,
Detector Pitch = table movement per 3600 gantry rotation / detector width
Generally its value lies between 3 to 6
The relation between detector and collimator pitches is defined as,
Collimator pitch = detector pitch/N
Here N = no. of detector arrays

21. Image Reconstruction:
I0
It
I0
It
It = I0 e - t
ln (I0 / It ) = t
 values for each ray can be computed by measuring Io and It
μ values depends on the composition of material, the density of
the material and photon energy
As X ray beam transmitted through the patient, different tissues
with different densities are encountered. So these  values for
each ray is used in CT reconstruction algorithm.

22. That means, the basic data needed for CT are the intensities of the
attenuated and un- attenuated X ray beams
Transmission measurement is made at different angles about the
object.
Transmission measurements are back projected on to a digital matrix.
Areas of different attenuations will be positively reinforced to
different degrees through the back projection and provide the picture of
internal structure.

23. Algorithm:
Simple back projection
Each ray in each view represents an
individual measurement of μ,
acquisition angle and position in the
detector array
Simple back projection starts with an
empty image matrix and the μ value
from each ray in all views is back
projected onto the image matrix.
Simple back projecting comes very
close to reconstructing the CT image as
desired, but a characteristic 1/r blurring
is a bi-product

24. 2. Filtered Back projection
In filtered back projection the raw view
data are mathematically filtered (modified)
before being back projected on to the
image matrix.
Mathematically, it reverses the image
blurring and this procedure involves
convoluting the projection data with a
convolution kernel
The kernel refers to shape filter function
in spatial domain.
In general, filtering will perform in
frequency domain.
Fourier transforms are used to convert
spatial domain into frequency domain.
Once filtration completed, inverse Fourier
transforms are used to define in spatial
domain

25. Convolution can be represented by
P’(x) = FT-1{FT[P(x)] (X) k(x)}
Here P(x) = Projected data at a given angle
k(x) = spatial domain kernel
P’(x) = filtered data in spatial domain
In clinical CT scanners, various filters used such as
oBone filters/Bone kernel
oSoft tissue filters/Soft tissue kernel
Bone kernels have less high frequency roll off where as soft tissue
kernels have more roll off at higher frequencies since for clinical
applications, high contrast resolution is more important than spatial
resolution

26. CT Number:
In image reconstruction, the attenuation coefficients are represented
by a relative quantity called CT number
CT number is the  value of the tissue corresponding to the pixel that
is normalized to the  value of water (the major component of tissue)
CT Number = [μtissue – μwater /μwater ] x 1000
Here μtissue – linear attenuation co-efficient of tissue
μwater – linear attenuation of co-efficient of water
CT number is expressed in Hounsfield Unit (HU) in honor of the
inventor Godfrey Hounsfield.
CT number for water = 1000 [ (w - w) /  w] = 0 HU
CT number for bone = 1000 [ ( 2 w - w) /  w] = 1000 HU
( because bone (0.528/cm) is about twice w (0.206/cm)
CT number for air = 1000 [ ( 0 - w) /  w] = - 1000 HU
( because  air is taken as 0 )

27. Relation between Attenuation & CT number
CT images are produced with a highly filtered high KV X ray beam with
average energy of 75KeV
At this energy range, in
o muscle tissues – 91%,
o fat – 94%
o bone structures – 74%
Compton scatter predominates
Thus, CT numbers and hence images derive their contrast mainly from
the physical properties (Physical Density and Electron Density) of tissue
that influence the Compton Scatter
CT numbers are quantitative and this property leads to more accurate
diagnosis in clinical settings

28. Relationship between CT number & Gray Scale
Tissue type CT numbers Appearance
Cortical bone +1000 White
Muscle +50 Gray
White matter +45 Light Gray
Gray matter +40 Gray
Blood* +20 Gray
CSF +15 Gray
Water 0 (base line)
Fat -100 Dark gray to black
Lung -200 Dark gray to black
Air -1000 Black
*White if iodinated contrast media is present
Relative attenuation (shades of gray) is based on CT numbers (HU)

29. Radiation Dose:
The radiation dose distribution in CT is markedly different than in
Radiography, because of the unique way in which dose is deposited
1. Single CT image is acquired in a highly collimated manner, the
volume of tissue is irradiated by primary X ray beam is
substantially small
2. The volume of tissue that is irradiated in CT is exposed to the X
ray beam from all angles during the rotational acquisition, this
makes the dose evenly distributed to the tissues
3. the radiation dose to the slice volume is higher due to higher KV
and mAs
The Compton Scattering is the principal interaction mechanism in CT,
so the dose attributable to scatter radiation is considerable, and it can
be higher than the radiation dose from the primary beam
Multiple Scan Average Dose (MSAD) is the standard for determining
radiation dose in CT

30. MSAD is defined as the average dose, at a particular depth from the
surface, resulting from a large series of CT slices
MSAD can be estimated with a single scan by measure of CT dose index
(CTDI)
When slices are contiguous, CTDI can be measured with long thin pencil
ionization chamber of length 100mm
CTDI = f. (X/T) . L
f = Conversion factor (air kerma to Dose mGy/mGy)
/ (Roentgen to mGy)
X = exposure in R or air kerma in mGy
L = length of pencil chamber (mm)
T = Slice Thickness (mm)
In helical scanning,
Dose (helical) = Dose (axial) x 1/Collimator pitch
Radiation dose is proportional mAs
i.e if mAs doubles then Radiation dose will also doubles

32. Image Quality:
Image quality is the visibility of important structures in the CT image
Majorly three factors that affect the CT image quality:
Resolution (Spatial / Contrast)
Noise
Patient Exposure
Resolution:
It is the quality of image how it appears.
Two types of resolutions in imaging such as Spatial and Contrast
Compared to X ray radiography, CT has significantly worse spatial
resolution but better contrast resolution

33. Spatial Resolution:
It is the ability of the CT scanner to display separate images of two
objects placed close together
In CT maximum high spatial resolution that may achieved is about
2lp/mm
Few of the factors that affect the Spatial Resolution in CT scanner are:
1. Detector Pitch: influences the view sampling
2. Detector Aperture: smaller detectors improves the spatial resolution
3. No. of views: More number views causes improved spatial resolution
4. Focal spot: smaller one increases higher spatial resolution
5. Slice thickness: smaller the thickness cause higher spatial resolution

34. Contrast Resolution:
It is the ability of the CT scanner to display an image of a relatively
large object that is only slightly different in density from its surroundings
Fundamentally, CT contrast is tied with SNR
SNR is much related to the no. of X ray quanta/ pixel in the image
If no. of rays/ pixel increases, the contrast also increases but causes
more patient dose
The factors affecting the Contrast resolution are:
1. mAs: mAs increases, more X ray photons, then contrast also
increases
2. Slice thickness: higher slice thickness projected by more X ray
photons
3. Patient Size: larger patient causes more X ray attenuation then
reduces contrast
4. Gantry rotation Speed: faster speed reduces mAs utilization and
causes low contrast

35. Noise:
It refers to the variation in the levels of grey in the image that are
distributed over its area but unrelated to the structure being imaged
The effect of noise on image quality:
It reduces the contrast resolution of small object
Worsen the spatial resolution of low contrast objects
Three types of Noise occur in CT imaging such as
1. Electronic Noise : Related measuring devices
2. Quantum Noise: Related to detection of X ray photons
3. Structural Noise: Related to algorithm

36. Artifacts:
Artifacts are distortions or errors in image that are unrelated to the
object scanned
The effect of Artifacts:
Deteriorate image quality
Subject information is lost
Most common artifacts in CT are:
1. Motion Artifacts
2. Streak Artifacts
3. Beam Harding Artifacts
4. Ring Artifacts

37. Motion Artifacts:
Causes: Patient movement , cardiac motion,
Mechanical misalignment.
Appearance: Blurred / streaks / ghost images
Rectification: Reduction in scan time
Clear and concise instruction to the patient
Proper patient immobilization
Administration of sedatives/anti peristaltic drugs

38. Streak Artifacts:
Cause: Presence and movements of objects of very high
density (contrast media/metallic implants/dental
amalgam/surgical clips)
Appearance: Streaks ( dark & light lines)
Rectification: Remove the offending object if possible
Use a smoothing algorithm (Standard algorithm)

39. Beam Hardening Artifacts:
Cause :
CT uses a poly-energetic X ray spectrum, with energies ranging
from 20KeV to 150KeV and Attenuation coefficients are energy
dependents
As the beam passes through the patient, low energy photons are
filtered out and the beam becomes harder
This results decrease in attenuation coefficient & CT number of
given tissue along the beam path
The tissue towards the center of the patient are invariably crossed
by harden beams compared to those near the surface.
 The result is that CT numbers are lower in the center of the
patient than should be.
Appearance: Wide dark streak
Rectification: Beam hardening correction algorithm

40. Partial Averaging Artifacts:
Cause :
CT number in each pixel is
proportional to the average μ
in the corresponding voxel
For voxels containing all one
type tissues, μ is representative
of that tissue
For voxels contain a mixture of
different tissue types, μ is not
representative to either tissue
but instead is a weighted
average of the two different μ
values
This leads to misdiagnosis
Rectification: Thinner CT slices

41. Ring Artifacts:
Cause : Detector failure or mis-calibration of a detector
Appearance: Ring
Rectification : Regular Quality Assurance Checks

42. Thank you…