CT Physics Review Questions

Backprojection and Image Reconstruction
Helical CT and Pitch
Radiation Dose in CT
Patient size and Dose in CT
Dose Reduction in CT
Cardiac CT
Beam Hardening and Dual-Energy CT

Take these questions as a:

Backprojection and Image Reconstruction

Back to section.

The 'filter' in filtered backprojection refers to
1. Bowtie filter between the beam and patient
2. Conversion between attenuation and Hounsfield units
3. Conversion between fan-beam and parallel geometry
4. Fix for the blurring inherent to backprojection
The filter in filtered backprojection compensates for blurring or smearing inherent to the process of backprojection.
Changing filters (aka kernels) in filtered backprojection results in
1. Trade-off between image sharpness and noise
2. Different window levels in CT images
3. Different patient dose
4. Different reconstructed field of view (FOV)
Changing the filter in filtered backprojection allows more or less high-frequency features through. High frequencies (just like the periphery of k-space in MRI) are responsible for image sharpness - but also noise. Thus, sharper kernels (e.g. bone filter or lung filter) result in noisier images. Soft tissue kernels blur out the noise for a smoother image. You can adjust the windowing on reconstructed images however you like, displaying bone filtered images on soft tissue windows for example.
The main advantage of iterative reconstruction techniques versus filtered backprojection is
1. Better depiction of bone detail
2. Does not require specification of reconstruction kernel or filter
3. Better handling of noisy images
4. Faster reconstruction
Iterative reconstruction techniques use cycles of simulation to converge on the 'best' solution for what the CT image should be. These are less sensitive to noise than filtered backprojection and thus allow better-looking reconstructions from low-dose (thus noisy) CT scans. They also work better for metal artifacts. Unfortunately, they are slower, and they do require specification of reconstruction kernels as well as reconstruction 'strengths' typically.

Helical CT and Pitch

Back to section.

In helical CT, a single transverse slice represents
1. A plane through the body perpendicular to the scan axis
2. A plane through the body oblique to the scan axis
3. A reconstruction made from projections at neighboring scan axis positions
In helical CT, the scanner never images a single slice. Instead, the slice is reconstructed by averaging CT projections taken 180 degrees apart - which are separated along the scan axis based on the value of the pitch.
In helical CT, pitch is defined as
1. Table movement in 360 degrees / beam width
2. Patient dose in 360 degrees / beam width
3. Reconstructed slice thickness / beam width
4. Gantry angle with respect to the scan axis
Pitch is defined by how much the table moves during the time it takes to cover 360 degrees divided by the width of the beam. Dose is related to pitch by CTDIvol = CTDIw/pitch. Slices can be reconstructed at any thickness or interval (with minimum being the collimator thickness). The angle of the gantry is unrelated to pitch.
High-pitch techniques are useful for
1. Very small findings (e.g. nondisplaced fracture)
2. Gated cardiac CT
3. Accurate multiplanar reconstructions
4. Fast scans
The advantage of a high pitch is that it reduces scan time and dose. However, it will blur small findings such as tiny liver lesions or nondisplaced fractures. Retrospective gated cardiac CT is done with very low pitch because the scanner must acquire images of the entire heart during diastole so it has to cover each point multiple times. Multiplanar reconstructions need relatively closely spaced data, so the higher the pitch the more likely for artifacts to occur.
In helical CT, a low-pitch technique might be most helpful for
1. Scanning a tachypneic patient
2. Detecting a non-displaced fracture
3. Detecting a subtle liver lesion
4. Scanning a young child
Low-pitch is most helpful for detecting small lesions, e.g. a non-displaced fracture (especially parallel to the transverse plane of scanning); low pitch is also used in retrospectively-gated cardiac CT for temporal resolution. However, it results in a slower scan, thus more respiratory motion artifact in a tachypneic patient. It has no real effect on contrast resolution, so it would not be helpful for subtle liver lesions. Finally, low pitch yields higher radiation doses, not good for scanning children.

Radiation Dose in CT

Back to section.

Which of the following numbers reported by the scanner best reflects the total amount of radiation delivered to the patient
1. Dose
2. CT Dose Index (CTDI)
3. Pitch
4. Dose-length product (DLP)
5. Effective mAs
All measurements of dose, such as CTDI, reflect the average energy delivered to a volume of tissue, i.e. J/kg. They do not tell you how much tissue was exposed. DLP is the CTDI*scan length, i.e. the total energy delivered (to the phantom). Ideally, one would convert the CTDI to an SSDE and multiply that by the scan length to correct for patient size.
The following term refers to a measure designed to reflect stochastic (cancer) risk from radiation
1. Dose
2. CT Dose Index (CTDI)
3. Average dose
4. Effective dose
5. Dose-length product (DLP)
Effective dose, measured in Sieverts (Sv), is a parameter designed to measure the stochastic risk from a radiation-based exam. It is calculated with tissue weighting factors for radiation sensitivity and requires detailed simulation or measurement of dose to each organ.
Deterministic effects in CT would best be measured by looking at
1. CT Dose Index (CTDI)
2. Dose-length product (DLP)
3. Effective dose
4. Pitch
While none of these options is ideal (none accurately reflect skin dose), the closest and best option is CTDI. This represents the dose to a phantom tissue simulator from the scan. Using conversion factors, one could estimate skin dose or - given scan length and body part scanned - even estimate effective dose, although these are highly dependent on patient size.
Dose in CT can be reduced by which of the following parameter adjustments (assuming other factors constant):
1. Increasing kV
2. Increasing mAs
3. Increasing Pitch
4. Increasing scan length
Increasing pitch reduces patient dose. Increasing kV will increase dose if mAs is kept constant; however, mAs can be adjusted to give the same dose (or even a lower dose) for any given kV setting. Increasing mAs increases photon flux and therefore increases dose. You have to be careful with scan length: changing scan length does not affect CTDI but does affect DLP. Thus, 'dose' (as in J/kg) does not change, but effective dose does change.
How is radiation dose distributed in the patient's body in CT?
1. Mostly at the skin surface
2. Mostly in the center of the body
3. More at the skin surface than in the center
4. Entirely uniformly from the skin to the center of the body
More dose is deposited at the skin and less in the center since x-rays are attenuated as they pass to the center of the body. However, this difference is decreased by the presence of the bowtie filter, which attempts to make the dose difference more uniform across the patient.

Patient size and Dose in CT

Back to section.

With the same CTDI (i.e. same scan parameters), what happens to radiation dose with changing the patient size?
1. Increasing patient size causes increased dose
2. Increasing patient size causes decreased dose
3. Increasing patient size does not change dose
Because of the bowtie filter, and to a lesser extent patient attenuation, larger patients will receive less dose than expected with the same CTDI.
The dose parameter used to correct CTDI for patient size is the
1. Effective dose
2. Dose-length product (DLP)
3. Dose-area product (DAP)
4. Size-specific dose estimate (SSDE)
5. CTDI Volume (CTDIvol)
The size-specific dose estimate (SSDE) applies a conversion factor to the CTDIvol based on patient diameter. Note that this does not represent the effective dose! That would need an estimate of dose to each organ, which obviously changes based on patient anatomy. While effective dose can be estimated for an 'average-sized' patient, these estimates will not be accurate for larger or smaller patients.
Positioning the patient off-center within the CT gantry will result in
1. Decreased dose
2. Increased dose
3. Reconstruction artifacts
4. Incorrect Hounsfield unit numbers
Because of the bowtie filter and (potentially) automated exposure compensation, CT dose will increase if the patient is off-center. Noise will also be worse.

Dose Reduction in CT

Back to section.

Decreasing kV in CT is advantageous because
1. X-ray penetration improves
2. Tissue contrast improves
3. Scan times are reduced
4. Metal streak artifacts are improved
Lower kV improves contrast between soft tissues as well as between iodine contrast and soft tissues. Penetration is worse with lower energy x-rays. Metal streak artifacts (beam hardening) are worse at lower x-ray energies. Note that lowering kV and keeping mAs constant yields a lower dose, but mAs may change if automated exposure compensation is on.
Automated exposure compensation uses the topogram (or scout) image to determine
1. Tube current (mA)
2. Pitch
3. Reconstruction filter
4. Bowtie filter
The scanner uses the topogram to determine the tube current to use at each position along the scan axis. The bowtie filter, pitch, and reconstruction filter are typically determined by the CT protocol settings.
The goal of automated exposure compensation is
1. To generate images of similar noise in different patient sizes
2. To scan patients of different sizes with the same kV and mAs settings
3. To obtain pretty, low-noise images
4. To eliminate the radiation risks from CT examinations
The goal of AEC is to generate images with relatively similar noise characteristics regardless of patient size. Radiation risks are still present, and mA settings are increased in large patients to compensate for increased attenuation. Low-noise images require high doses.

Cardiac CT

Back to section.

The major technical challenge in cardiac CT is
1. Spatial resolution
2. Temporal resolution
3. Contrast resolution
4. Three-dimensional reconstruction
The major challenge in cardiac CT is 'freezing' the motion of the heart so that static images can be obtained of the coronary arteries. This requires very good temporal resolution.
The major determinant of temporal resolution in CT is
1. Gantry rotation speed
2. Reconstruction algorithm
3. Fan-beam angle
4. Detector collimation
5. Computer processing power
Temporal resolution represents the amount of time needed to reconstruct a single slice - which is the time needed to obtain 180 degrees (+ fan angle) of data. This is directly determined by the gantry rotation speed, which represents a major engineering challenge in scanner design. While fan-beam angle is relevant, it plays only a minor role in temporal resolution. Pitch is also a relevant factor.
Increasing the number of rows in MDCT principally allows for
1. Greater spatial resolution
2. Greater temporal resolution
3. Greater axial coverage
4. Greater contrast resolution
The number of rows in MDCT determines the axial coverage in any given rotation of the scanner. With enough rows (e.g. the 320-row cone-beam scanner), the entire heart can be captured in a single rotation. Temporal resolution is still determined by the rotation time, however.
A dose-reduction strategy in prospectively gated cardiac CT would be
1. Increasing tube mAs
2. Increasing tube kV
3. Utilizing ultrahigh pitch technique
4. Utilizing a scanner with more detector rows
5. Changing the heart rate
Ultrahigh pitch mode allows for lower dose, as the pitch is increased substantially. Even though these scanners do have 2 x-ray tubes, the resulting scans have lower dose. Increasing mAs would lead to increased dose; increasing kV would lead to poorer contrast resolution (dose may or may not increase, depending on dose-modulation techniques). Detector rows and heart rate would not have a substantial impact on dose.
An advantage of a cone-beam scanner (versus standard MDCT) in cardiac CT is
1. Improved spatial resolution
2. Improved temporal resolution
3. Improved contrast resolution
4. Improved patient motion artifacts
As a cone-beam scanner can image the entire heart within one heart beat, thus not requiring imaging over several heartbeats in between which the patient may move or breathe. Temporal resolution is determined by the gantry rotation time.
High radiation doses seen in retrospectively gated cardiac CT, as compared to prospectively gated CT, are a result of
1. Increased scanner mAs
2. Increased scanner kV
3. Increased scanner pitch
4. Decreased scanner pitch
5. Increased axial coverage
6. Improved spatial resolution
The high radiation doses seen in retrospectively gated cardiac CT reflect the low pitch needed to image each slice of the heart through the entire cardiac cycle. The other scan parameters (e.g. mA, kV, axial coverage, and spatial resolution) are not changed.
Retrospective gating (versus prospective gating) would be particularly useful in a patient with
1. Large body habitus
2. Slow heart rate
3. Irregular heart rate
4. Anomalous coronary artery origin
The advantage of retrospective gating is the ability to 'edit' the acquired data to exclude certain heartbeats such as PVCs, which in prospective gating may result in acquisition of images in systole instead of diastole. Patients with a slow, regular heart rate are perfect candidates for a prospective gating strategy. Very rapid heart rates may see an advantage in retrospective gating if multi-segment reconstruction is used, as this strategy can improve temporal resolution.
A dose reduction strategy in retrospectively gated cardiac CT would be
1. ECG dose modulation
2. Multi-segment reconstruction
3. Increasing kV
4. Decreasing heart rate
ECG dose modulation turns down the mA during systole so that overall radiation dose is less for the scan; we do not need as much dose during systole since those images are typically not used for coronary artery evaluation. Multi-segment reconstruction requires a _lower_ pitch and thus higher dose; slower heart rates require lower pitch as well. Increasing kV would decrease contrast resolution.

Beam Hardening and Dual-Energy CT

Back to section.

The most important physical process responsible for tissue contrast in CT imaging is
1. Coherent (Rayleigh) scatter
2. Photoelectric absorption
3. Incoherent (Compton) scatter
4. Characteristic x-ray production
Photoelectric absorption strongly depends on the material (atomic number and k-edge energies). Thus, especially at lower x-ray energies, attenuation because of photoelectric absorption is much different among different materials. Coherent scatter is pretty much irrelevant at diagnostic energies. Compton scatter is another important source of attenuation but does not distinguish between different materials very well.
Iodine is a good CT contrast agent because
1. We can achieve very high concentrations in target organs
2. The k-edge (33 keV) is much lower than x-ray energies produced by a 120 kV tube
3. The k-edge (33 keV) is near the average diagnostic x-ray energy produced by a 120 kV tube
4. It simulates the attenuation of bone
The k-edge of iodine (33 keV) is very close to the average diagnostic x-ray energies (1/3 of the maximum, in the 30-40 keV range) - thus, diagnostic x-rays are strongly absorbed by iodine-containing structures. While we can achieve decent concentrations of iodine, the concentration is not so high as to cause attenuation by simply being present. Bone contains calcium, which is attenuating because of density and higher atomic number, although the k-edge is quite low, below relevant energies.
The CT number (Hounsfield unit) of fat depends on
1. kV
2. mAs
3. Reconstruction algorithm
4. Nothing - it is constant
The attenuation of water and air are set - but not that of any other tissue. Fat's attenuation depends on the x-ray energy used in the scan. mAs changes the number of photons but not their energy. While the reconstruction algorithm changes the way an image looks (noise characteristics), the CT number of the tissue does not change.
Beam hardening in x-ray imaging refers to
1. Decreasing x-ray beam strength as it passes through a dense material
2. Decrease in a photon's energy as it is scattered by a dense material
3. Increase in a photon's energy as it is scattered by a dense material
4. Decreased average x-ray energy as a beam passes through a dense material
5. Increased average x-ray energy as a beam passes through a dense material
As a polychromatic x-ray beam (i.e. a beam with a spectrum of energies) passes through a material, the lower energy x-rays are attenuated more than the higher energy x-rays. Thus, the resulting beam has a higher average energy - it is _hardened_.
A hardened beam has what effect on the CT appearance of soft tissue
1. No change
2. Darker with poorer contrast
3. Darker with better contrast
4. Brighter with the same contrast
A hardened beam has fewer low energy x-rays - those x-rays that are able to distinguish different tissues. Thus, contrast is worse. Additionally, the higher energy x-rays are able to penetrate much better overall, making the tissue appear darker (i.e. 'less dense').
Two x-ray energies are used in DEXA bone densitometry in order to
1. Determine the minerals present in the bones (calcium vs. phosphorus)
2. Cancel out soft tissue attenuation
3. Find the optimal energy to image the bones
4. Decrease radiation dose
The two energies are used in order to determine attenuation at each energy - a simple set of equations can be solved to find the fraction of bone (which attenuates more at lower energies) versus soft tissue.
Advantages of dual-energy CT include all the following except
1. Virtual non-contrast images
2. Separation of dozens of different materials
3. Organ perfusion maps
4. Calcium subtraction
Virtual non-contrast, organ perfusion, and calcium subtraction all involve separating the dual-energy scan into three tissue components (soft tissue, iodine, and bone). However, with 2 energies, that is pretty much the limit. In order to separate more tissues, we need measurements at more x-ray energies - this may be accomplished with spectral CT.