CT Exam 4 Chpt 5 & 6
Object model
Deals with radiation attenuation through the patient
IR (iterative reconstruction) algorithm was first used by
Dr. Godfrey Hounsfield
All of the following describe basic problems with the back-projection algorithm EXCEPT:
compares assumptions with measured values
Fourier coefficients
constants used in the Fourier series making reconstruction of a CT image possible and easy
Which of the following uses a digital filter to modify an image matrix?
convolution
FBP is also referred to as the:
convolution method
Filtered back-projection is also known as
convolution method
What step in the typical IR process generates the initial CT image from an FBP algorithm?
input
3D reconstruction technique allows the user to:
interactively visualize, manipulate, and measure large 3D objects on general purpose work stations
Fourier reconstruction technique does not use any filtering because:
interpolation produces a similar result
Convolution kernels can only be applied to____
raw data
Examples of Studies in 2012
• ASIR reduced image noise 26% to 30% • ASIR resulted in a dose reduction • 46.4% reduction in CT chest exam • 38.2% reduction in CT abdomen exam
The textbook highlights two IR algorithms available commercially today:
• Adaptive statistical iterative reconstruction (ASIR)-V • Adaptive iterative dose reduction (AIDR) 3D
Examples of Studies in 2013
• Algorithms involved in these studies • ASIR • Model-Based Iterative Reconstruction (MBIR) • Sinogram Affirmed Iterative Reconstruction (SAFIRE) • iDose
Input
• CT scanner produces raw data or measured projection data • Raw data set is then subjected to the standard FBP algorithm to generate the initial CT image
3 essential steps in the IR loop:
1. forward projection is applied to the initial CT image to create artificial raw data (simulated projection data) 2. simulated raw data are compared with the measured raw data projection and the difference is calculated to create an updated image. (correction term) 3. the updated image is back-projected (FBP algorithm) to the current CT image.
4 strategies used to reduce noise during CT image reconstruction:
1. projection space denoising 2. analytical FBP denoising using standard reconstruction kernels 3. image space denoising 4. Iterative reconstruction
The horizontal ray sums for the pixels shown are: > 2 3 0 1 5 2 >
13
3D imaging uses:
3D surface and volumetric reconstruction
As documented by a 2012 performance evaluation, what percentage of dose reduction did the ASIR algorithm result in when compared to the FBP algorithm for a CT chest exam?
46.4%
Toshiba's most recent IR algorithm
AIDR 3D
GE's next generation of ASiR
ASIR-V
What was the first IR algorithm commercially available and when?
ASiR algorithm, 2008
Optics model
Addresses the x-ray source, image voxel, detectors, and the like
Which of the following refers to a defined set of instructions for solving a problem?
Algorithm
Which of the following occurs as the last step before the image is created using the FBP algorithm?
Back Projection
The first operation to which raw data are subjected is referred to as:
Back-projection
Which of the following describes the beam that diverges from the xray tube to the outer edges of the detectors?
Cone beam
The first operation to which raw data is subjected is referred to as:
Convolution
Which of the following is a digital image processing technique used to modify images through a filter function?
Convolution
which is a class of cone beam algorithms available for use with the new generation of multi slice CT (MSCT) scanners?
Exact cone-beam algorithms
Physics model
Examines the physics of CT data acquisition
CT manufacturers offer each of these of IR algorithms commercially EXCEPT:
FBP algorithm
FBP, ASiR and VEO comparison
FBP-standard image noise, limited dose reduction opportunity, diagnostic image quality at standard dose ASiR- lower image noise, lower dose, improved image quality at lower dose VEO- lowest image noise, lowest dose, improved image quality at lowest dose
The image reconstruction algorithm used by current scanners:
Filtered back-projection
Which of the following belongs to the class of analytical algorithms for CT:
Filtered back-projections
Which of the following changes a time domain signal into a frequency domain signal?
Fourier transformer
Which is a mathematical technique used to estimate the value of a function from known values on either side of the function?
Interpolation
In a 2015 performance evaluation, it was noted that the reduced dose iterative model reconstruction (RD IMR) algorithm proved significantly better than reduced FB (RD FBP) algorithm in each of these pathological processes EXCEPT:
Intracranial hemorrhage
why is back-projection not used in CT?
It does not produce a sharp image of the object
A notable difference between SSCT and MSCT
MSCT uses multiple detector rows that cover a larger volume at an increased speed and therefore require new algorithms
What area of modeling does the adaptive statistical iterative reconstruction-V (ASIR-V) algorithm not focus?
Optics
Which model addresses the X-ray source, image voxel, and detectors?
Optics model
Which of the following is NOT an essential step in an IR algorithm without modeling?
Predetermined image quality criteria are not necessary
Noise model
Takes into consideration the x-ray photon statistics
What does the IR loop form?
The basis of the iterative reconstruction
Statistical methods with modeling use:
X-ray photons (photon counting statistics) during the IR process.
What is the output of an IR loop
a back projection that results in the final volumetric image
A measure of the total absorption of x rays along a straight line is referred to as?
a ray sum
Fourier transform divides a waveform into:
a series of sine and cosine functions of different frequencies and amplitudes.
Which of the following is NOT a key feature of FBP algorithm-measured projection data (raw data)?
adaptive section collimation
Image data are obtained:
after back projection
The following refers to IR algorithms currently available from CT manufacturers:
all are correct: 1. statistical methods w/modeling photon statistics 2. iterative without modeling 3. model-based methods
The back projection technique results in:
all are correct: classical star pattern, image blur, poor quality and no diagnosis can be made
A 3D reconstruction technique for surface display is based on;
atleast two processes; preprocessing and display and consist of the following operations; interpolation, segmentation, surface formation, and projection
algorithms applicable to CT:
back projection iterative methods analytical methods
Algorithms applicable to CT
back projection, iterative methods, analytical methods
The criteria used to evaluate IR's performance in various studies include each of the following EXCEPT:
data acquisition speed
Edge enhancement algorithms-
emphasize the edge of structures and improve detail BUT create image noise.
advantages of ASIR-V algorithm:
faster image reconstruction, can be used with low dose CT examinations and delivers low radiation dose to patients.
Compared to FBP algorithm, IR algorithms result in all of the following EXCEPT:
faster production of the final CT image
CT scanners now use the:
filtered back projection algorithm
How is ASIR-V algorithm different from VEO CT model based IR?
it features more sophisticated noise modeling and object modeling including a bit of physics modeling, and does not focus on the system optics modeling
What term is known as a calculation process using a series of operations that are repeated several times?
iterations
Types of data
measurement data, raw data, filtered or convolved data, image or reconstructed data
output of the IR loop is dependent on
modeling
Convolution: The process involves:
multiplication of overlapping portions of the filter function and the detector response curve selectively to produce a third function which is used for image reconstruction.
FBP limitations:
noise and artifact creation
A IR algorithm that models the X-ray tube, image voxels and detectors, etc. is referred to as the:
optics or system model
Hounsfield's first scanner used _________________ beam projections acquired through a 180-degree rotation.
parallel
A set of ray sums is referred to as:
projection
The following are assumptions of the FBP algorithm EXCEPT:
projection data is shaped by a special filter
Which of the following is NOT a primary goal of iterative reconstruction (IR) algorithms?
reduce artifacts caused by patient motion
Primary advantages of iterative reconstruction algorithms
reduce image noise and minimize the high radiation dose that is inherent in the filtered back-projection algorithm
Smoothing algorithms do what?
reduce image noise and show good soft tissue anatomy
How does convolution improve image quality?
removal of blur; sharpens image
A radiograph is considered and image in the:
spatial domain; shades of gray representing various parts of the anatomy
Limitations presented by the filtered back-projections (FBP) algorithm include all of the following EXCEPT:
speed
The FBP algorithm is recognized for its:
speed and "closed for solution" which is available from a single pass over the acquired data.
back-projection is also called:
summation method or linear superposition method
With IR algorithms without modeling:
the FBP algorithm is applied to the raw data to produce an initial CT image.
Modeling is used to refer to:
the characteristics of the CT imaging system and the image object and includes what has been popularly referred to as system optics, noise statistics, and object and physics model.
Fan beam approximation algorithm require that
the data be consistent; the xray beam from the tube to the detector and the section being imaged must be in the same plane
An important step in the implementation of IR algorithms is:
the need for performance evaluation or these algorithms, particularly in CT clinical practice
Filtered back-projection
the projection profile is filtered or convolved to remove the typical star like blurring that is characteristic of the simple back-projection method
Raw data
the result of preprocessed scan data and are subject to the image reconstruction algorithm used by the scanner.
Performance evaluation studies have compared IR algorithms with FBP algorithms on all of the following characteristics EXCEPT:
the speed of obtaining the final image
Goals of IR algorithms compared to FBP algorithms:
to improve image quality (spatial resolution or sharpness), reduction of noise, better low contrast detect-ability (LCD), and artifact reduction.
The basic problem in CT is to calculate:
us for tissues from a large set of projection data.
The most recent method to reduce noise in low dose CT imaging is to:
use an interactive reconstruction (IR) algorithm
Fourier Transform
used to reconstruct images of a patients anatomy in CT and also in MRI
Examples of Studies in 2014
• New performance evaluation studies • AIDR 3D • Image Reconstruction Image Space (IRIS) • The basic conclusion of these studies is that both IR algorithms reduced noise and improved image quality compared to the FBP algorithm
Iterative Reconstruction Algorithm Goals
• Reducing image noise when using lower exposure technique factors while preserving image quality • Reducing artifacts due to the presence of metal implants, beam hardening effects, and photon starvation
Examples of Studies in 2015
• Study compared performance of a knowledge-based iterative model reconstruction (IMR) algorithm with that of the FBP algorithm using a standard dose technique and a reduced dose technique • The IMR reduced dose detected the same lesions as the FBP standard dose • The IMR reduced dose produced acceptable image quality, whereas the FBP reduced dose was unacceptable for subjective image quality
assumptions of FBP algorithms
• The x-ray focal spot is a point source • The size and shape of each detector cell is not taken into consideration • "All x-ray photon interactions are assumed to take place at a point located in at the geometric center of the detector cell, not across the full area of the detector cell" • A pencil beam is used to "represent the line integral of the attenuation coefficient along the path" • The size and shape of the voxels are "infinitely small points located on a square grid" • The projection data are accurate and not affected by x-ray photon statistical changes and electronic noise
