Ch 16-Tomography in Nuclear Medicine

What is the basic problem with conventional radionuclide imaging?

images being obtained only with two-dimensional (2-D) projections of three dimensional sources (3-D).

Why is the problem to obtain 2 dimensional images?

potential to be obscured by superimposed images of overlying and/or underlying structures

How we overcome the 2 dimensional imaging situation?

taking projection images from the posterior, laterals, and oblique views.

Why do we take different projections of the 3d object?

most structures that are deep-lying have overlying structures from all projection angles

What are tomographic images?

are 2-D representations of structures lying within a selected plane in a 3-D object.

What are the scanners that obtain images that the scanner rotates and get tomogrphic images?

• CT

• PET

• SPECT/CT

Mathematical algorithms

are then used to reconstruct images of selected planes within the object from the projection data

What is Emission Computed Tomography (ECT)?

is the process of reconstructing images of selected planes within the object from the acquired projection data.

What is the process of the Emission Computed Tomography (ECT)?

we can produce images where the activity from overlying (or adjacent) cross-sectional planes is eliminated from the image.

Does the ECT increases the contrast-to-noise ratio?

Yes

Reconstruction of images from transmitted emissions from an external source like an X-ray tube, is known as

Transmission Computed Tomography

Does the SPECT and PET improve the contrast-to-noise ratio?

True

Why does the SPECT and PET improves the contrast-to-noise ratio?

primarily from the ability to quantify the exact location and depth of the activity within the body.

When the mathematical algorithm underlying reconstruction tomography was first published?

Johann Radon in 1917.

However, it wasn’t until the 1950’s and 1960’s that radio-astronomy and chemistry allow for practical application.

True

Does PET and SPECT use the same mathematical formulas to process the images?

True

Data is collected with a standard gamma camera fitted with a conventional parallel-hole collimator

True

What limit the radiation and what creates?

Cylinders form the collimator and creates the line of response for the collimator hole

What is the measured quantity is referred?

to as the line integral for the line response (can be counts recorded or radioactive content)

A full set of line integrals recorded across the detector is called a

Projection or a projection profile

Images will be collected from several angles around the body

True

PET systems use of detector elements in a ring or around the body

stationary arrays, hexagonal pattern

Evenly spaced intervals help recreate the object with better visual distinction.

True

How is the data collected for tomograpghic reconstruction?

data collected from the object correlates to a slice through the object perpendicular to the bed.

What is the angle direction for this scan?

transverse or transaxial direction

What is the direction along the axis of the bed which defines the location of the slice?

Axial direction

However in the clinic and for practical applications usually 360 degree images are obtained

True

One common way to display a full set of projection data is in the form of a 2D matrix

Sinogram

What is the row mean in the sinogram?

intensity display across a single projection

Why there each successive row is taken from top to bottom?

indicating the different angles images were acquired.

What is the general goal of reconstruction tomography?

is to generate a 2D cross sectional image of activity from a slice within the object.

How does this general goal of reconstruction tomography is done?

done using the sinogram or set of projection profiles for that slice

What is the most basic approach for reconstructing an image?

simple back projection

How does the simple back projection function?

It functions similar to a Sudoku square with squares being filled in or adding intensity the more slices that include overlap

What is back projection?

Profile elements are divided uniformly amongst the pixels that fall within its projection path.

What means when the back projections for all profiles added together?

there is an approximation of the distribution of radioactivity within the scanned slice that is obtained

The image that is obtained resembles the true source distribution, yet there are still some obvious artifacts present in the counts that are projected outside the true location of the object.

True and this lead to blurring of the image

How can we increase the quality of the image?

increasing the number of projection angles and the number of samples along the profile.

Does the Direct Fourier Transform Reconstruction avoid blurring?

True

Does the direct fourier transform reconstruction serves as an alternative method for representing spatially varying data?

True

What is Direct Fourier Transforrm reconstruction indicate?

indicates the Fourier transformation (FT) of a profile is equal to the value of the FT of the object measured through the origin and along the same angle

Is DFT the creation of 1-D profile of a 2D object?

True

The drawback of fourier reconstruction is the interpolation of data from polar to rectangular coordinates in K-space is computationally intensive.

True

Does the Fourier Reconstruction generates artificats in the image?

True

What is the method called to get rid of the artifacts for the Fourier Reconstruction?

Filtered Back Projection

What are the steps for filtered back projection?

1. Acquire the projection profiles at N projection angles

2. Compute the 1-D FT of each profile to generate values of the FT for a line in the K-space

3. Apply a ramp filter to each K-space profile

4. Compute the inverse FT for each filtered FT

5. Perform convention back projection using the filtered profiles

What is the purpose of the ramp filter?

enhance high spatial frequencies and to suppress low spatial frequencies.

What is the result form the ramp filter?

image that contains less blurring

What happens to the negative portions of the filtered profiles? (Ramp Filter)

subtract out some of the projected intensity to reduce blurring of the object

Why do we use the Filtered Back Projection?

• Due to its speed and widely used reconstruction method

• a single 2-D slice is reconstructed quickly on a standard computer

• It produces an accurate distribution of the radioactivity within the slice

What are the drawbacks of the filtered back projections?

• limitations of major artifacts if data form the object is not measured incompletely

• poor counting statistics or random noise spikes which can create streak artifacts

• cannot be easily modified

What is the other reconstruction method that fixes the drawbacks of filtered back projection?

Iterative reconstructive techniques

What applies to single-slice imaging?

Back Projection and Fourier-Based Reconstruction

What does the PET and SPECT acquire?

data simultaneously for multiple sections through the body.

How does multislice imaging work?

Projection data is obtained from each section and are reconstructed into sliced then subsequently stacked to form a 3D Dataset

What can be reconstructed with the newly generated 3D data set?

• Coronal

• Transverse

• Sagittal

What is the other name for Coronal plane?

Frontal plane

The animal plane is used for research

True

The image quality for FT & FBP and the effects of sampling on image quality?

• Is point by point samples of projection profiles

• Distance between the sample point is the linear sampling distance

• There is a finite number of angular sampling intervals around the object

Courier Sampling, images are shown with less samples per iteration

True

FBP there is a trade off between image detail and signal to noise ratio

True

What is Iterative Reconstruction?

A viable and increasingly used alternative to FBP

Is the Iterative reconstruction more intensive than FBP?

Yes, has been slower to adapt into the clinical setting due to the increase time needed for processing

What is the general idea of Iterative Reconstruction?

estimating the true image and forward projection, which is the inverse of backprojection

How does forward projection work?

is performed by summing up the intensities along the potential ray paths for all projections through the estimated image.

If the foreword projection doesn’t work?

The estimated projection is compared to the actual projection if they don’t match up it takes the profile again and estimates it

What are the steps for Iterative Reconstruction Steps?

1. Image is obtain

2. Projection data is measured

3. Profile is compared to the image, estimation will occur

4. New projection file is calculated

5. The image will be reconstructed after passes the algorithm

What is the method for comparing the estimated and actual profiles?

The cost function

How the cost function work?

which measures the difference between the profiles generated by forward projections through the estimated image and the profiles actually recorded from the scanned object.

What is the Search or update function?

which uses the output of the cost function to update the estimated image.

What is the goal of Iterative Reconstruction?

to generate versions of these functions that produce the true function as quick and accurately as possible.

What does the IR algorithms require for an acceptable image?

several iterations to converge to an acceptable image and each iteration is equivalent to a separate back projection procedure.

FBP was the most time consuming portion of FBP processing filters

True

Forward projection is similarly time consuming?

True

What is the IR incorporate factors?

Collimator size, scatter and system geometry all increase the processing time for the image

What is ordered subsets?

is a method that increases processing time by limiting projection angle processing for the initial iterations, allowing the first few iterations to be processed faster. The entire projection profiles are used later to refine out small details in the image.

What is expectation-maximization (EM)?

incorporates statistical considerations to compute the “most likely” or maximum likelihood (ML) source distribution that can create the observed projection data.

How does EM work?

Specifically it assigns greater weight to high count elements of a profile and less weight to low-count regions.

What does back projections typically assigns?

equal weight to all elements of a profile.

What does the ML-EM help on?

increase image detail and shorten processing time.

What are the fan beam collimator made for?

allow for great static image contrast to be obtained

What are coned beam collimators made for?

allow for only 180 degrees of rotation in order to reconstruct the image

What is the cone collimator directed to?

toward or away from a common focal point

How the holes work for the cone beam collimators?

Each row of holes across the center of the collimator provides a projection profile, but the profiles all intersect at the center.

Are there multiple detector rings situated through the gantry?

True

The Projection data is acquired within a given detector ring and transverse images are reconstructed

True

What are the planes of interest for PET scans?

Coronal, transverse, and sagittal