Scribd
Upload
23 views31 pages

Saurav Gusain CT Image Reconstruction

Uploaded by

Swapnil Satapathy
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
23 views31 pages

Saurav Gusain CT Image Reconstruction

Uploaded by

Swapnil Satapathy
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 31

IMAGE RECONSTRUCTION IN

CT

SAURAV SINGH GUSAIN


230513009
MRIT 3rd SEM
TABLE OF CONTENT

• Introduction
• Algorithm for image reconstruction
• Back projection
• Analytic reconstruction
• Filtered back projection
• Fourier reconstruction
• Iterative method
• Reference
• Quiz
INTRODUCTION

• Image reconstruction in CT is a mathematical process that


generates tomographic images from X-ray projection data
acquired at many different angles around the patient
• The process of using raw data to create an image is called
reconstruction
• Image reconstruction has fundamental impacts on image quality
• It is done with the combination of complex computer
algorithms, mathematical equations and physics
• 1917 Radon has developed mathematical solution to the
problem of image reconstruction from a set of projection
• In 1961 this technique used in the field of medical
ALGORITHMS FOR IMAGE RECONSTRUCTION

• An algorithm is a mathematical method for solving a problem


• Thousands of equations must be solved to determine the linear
attenuation coefficients of all the pixels in the image matrix.
• Several different methods, or algorithms, are in current use all
attempt to solve the equations as rapidly as possible without
compromising accuracy.
• There are three main types of mathematical algorithms for
image reconstruction:

1. Back projection
2. Analytic reconstruction
3. Iterative reconstruction
BACK PROJECTION

• Back-projection, sometimes called the “summation method” is


the oldest means of image reconstruction
• First used by Oldendorf (1961) and Kühl and Edwards (1963).
• It’s method is analogous to a graphical reconstruction
• This involves "smearing back" the projection across the image
at the angle it was acquired.
• By smearing back all of the projections, you reconstruct an
image
GRAPHIC REPRESENTATION
• Consider four beams of x rays that
pass through an unknown object to
produce four projection profiles P1, P2,
P3, and P4
• The projected datasets are back-
projected (i.e. linearly smeared) to form
the corresponding images BP1, BP2,
BP3, and BP4. The reconstruction
involves summing these back projected
images to form an image of the object.
MATHEMATICAL REPRESENTATION
• The simple back projection has the problem of “star pattern” artifacts
caused by “shining through” radiations from adjacent areas resulting
in the blurring of the object
ANALYTIC RECONSTRUCTION

• Analytic reconstruction algorithms were developed to overcome the


limitations of back-projection and iterative algorithms and are used
in modern CT scanners
• All commercial scanner now uses this method
• It uses a mathematical technique known as Convolution or Filtering
• There are two major types
1. Filtered back projection
2. Fourier reconstruction algorithms
FILTERED BACK PROJECTION

• Filtered back-projection is also referred to as the convolution


method
• It is similar to the back projection except the image is filtered or
modified to exactly counter balance the density which causes
blurring (Star-pattern) in simple back projection
• Convolution filter refers to a mathematical filtering of the data
designed to change the appearance of the image
FOURIER RECONSTRUCTION

• Developed in 17th century by mathematician Baron Jean


Baptiste Joseph Fourier
• It is a mathematical procedure for breaking down a waveform
into a series of sine and cosine functions of different frequency
and amplitude
• It is a mathematical function that converts a signal in spatial
domain to a signal frequency domain
• This algorithm does not use any filtering as interpolation does
the work of rearranging the image components in a rectangular
grid
FLOWCHART OF FOURIER RECONSTRUCTION

The object to be scanned is represented by function (x,y)

Projection data are obtained, at least 180° represented as spatial


domain image

Each projected data are transformed into frequency domain by


fourier transform

CT uses fast fourier algorithm to acquire image in radial pattern thus


must be converted into rectangular grid by interpolation technique

Interpolation image is transformed into spatial domain image


through inverse fourier transform
ITERATIVE METHOD

• This is first used by Sir Hounsfield in his first EMI scanner


• An iterative reconstruction starts with an assumption (for
example, that all points in the matrix have the same value) and
compares this assumption with measured values, makes
corrections to bring the two into agreement, and then repeats
this process over and over until the assumed and measured
values are the same or within acceptable limits
• The primary advantages of iterative image reconstruction algorithms
are to reduce image noise and minimize the higher radiation dose
inherent in the filtered back-projection algorithm
• There are three variations of iterative reconstructions, depending on
whether the correction sequence involves the whole matrix, one ray,
or a single point
1. Simultaneous reconstruction
2. Ray by Ray correction
3. Point by Point correction
• Let’s illustrates a ray-by-ray iterative reconstruction for a four-
element square
• Consider the following numeric illustration
• Original projection datasets
(horizontal ray sum)

1 2 3
3 4
7

4 6
1. Initial estimate: Compute the average of four elements and
assign it to each pixel, that is, 1 + 2 + 3 + 4 = 10; 10/4 = 2.5
• New projection dataset
(Horeizontal ray sum)

2.5 2.5 5
2.5 2.5
5
2. First correction for error (original horizontal ray sums minus the new
horizontal ray sums divided by 2)
• (3 − 5)/2 and (7 − 5)/2 = −2/2 and 2/2 = −1.0 and 1.0
(2.5-1) (2.5-1)
1.5 1.5
(2.5+1) (2.5+1)
3.5 3.5

1.5 1.5

3. Second estimate: New projection dataset 3.5 3.5

(Vertical ray sum)


5 5
4. Second correction for error (original vertical ray sums minus
new vertical ray sums divided by 2)
• (4 − 5)/2 and (6 − 5)/2 = −1.0/2 and +1.0/2 = −0.5 and +0.5

(1.5-0.5) (1.5+0.5)
1 2
(3.5-0.5) (3.5+0.5)
3 4

1 2
• The final matrix solution is thus
3 4
• In the early years of CT use these iterative techniques were not
used in commercial scanners because of the following limitations:
1. It is difficult to obtain accurate ray sums because of quantum
noise and patient motion.
2. The procedure takes too long to generate the reconstructed image
because the iteration can be done only after all projection datasets
have been obtained, and because of the lack of computing power
to solve these equations quickly.
3. To produce a “true” image, there should be more projection
datasets than pixels
REFERENCE

• COMPUTED TOMOGRAPHY physical principles, clinical


applications, and quality control FOURTH EDITION dr. Euclid
seeram
• Christensen's Physics of Diagnostic Radiology
• https://www.slideshare.net/SantoshOjha15/ct-image-
reconstruction
• https://www.slideshare.net/avcracker/ct-image-reconstruction-
avinesh-shrestha
QUIZ TIME ….
THANK YOU

You might also like