1. In MRI signal-to-noise ratio (SNR) can be increased by all of the following except:
A. Switching from body to head coil.
B. Increasing the number of acquisitions.
C. Increasing the static magnetic field strength.
D. Decreasing the slice thickness
E. Decreasing the receiver bandwidth.
2. Chemical shift artifacts in MR imaging are:
A. Secondary to using gadolinium contrast.
B. More severe for lower magnetic field strength.
C. Not affected by the gradient strength.
D. Seen in the frequency encoding direction.
E. Eliminated by swapping the phase and frequency gradient directions.
3. The figure below shows a cross-sectional view of a superconducting magnet. The RF coils are represented by
4. ___________ is the fastest acquisition pulse sequence to produce MR images.
A. Inversion recovery spin echo
B. Echo planar
C. Fast spin echo
D. Gradient recalled echo
5. What is common among the following MR spin echo pulse sequences: Inversion Recovery (IR), Short Tau IR
(STIR), and Fluid Attenuated IR (FLAIR)?
A. The use of a very long inversion time (TI) for fat suppression.
B. The use of a very short inversion recovery time (TI) for nulling CSF signal.
C. They employ a 180-90-180 degree RF pulse sequence per TR.
D. They employ a 90-180-180 degree pulse sequence per TR.
E. They produce T2-weighted images.
6. In MR imaging, selective observation of the interaction between protons in free water molecules and protons
contained in the macromolecules of a protein forms the base for producing ________.
A. Magnetization transfer contrast
B. Perfusion contrast
C. Diffusion contrast
D. Functional MR images
E. Blood oxygen level contrast
7. Image intensifiers, CRTs and scintillation cameras should be outside the _________ magnetic field gauss line.
8. Cortical bone does not show strong signals in a proton density weighted scan because:
A. it has a low attenuation coefficient for RF energy.
B. Of a very long T1 value.
C. It is made up of a large amount of paramagnetic substances.
D. Of a lack of mobile protons.
E. RF cannot penetrate to flip the spins in the transverse plane.
9. The change in magnetic field strength from one end of the imaging volume to the other with an applied
frequency-encoding gradient is closest to _________ gauss for a 30 cm FOV.
10. The T2 relaxation time of a tissue is about 60 mS on an 0.5 Tesla MRI system. On a 1.5-tesla MRI, one might
expect the T2 relaxation time to:
A. Increase significantly.
B. Increase slightly.
C. Remain the same.
D. Decrease slightly.
E. Decrease significantly.
11. In MRI, which of the following states is True regarding the function of RF coils?
A. The same coil may be used to transmit the RF pulses and to receive the MR echoes.
B. Separate coils may be used to transmit the RF pulses and to receive the MR echoes.
C. A surface coil would be optimal for imaging structures deep within the brain.
D. They digitize the received echoes.
E. A and B.
12. Which one of the following is not characteristic of superconducting magnets?
A. Requires the use of cryogen.
B. Shuts off as soon as the electrical current supply to its windings is stopped.
C. Available in clinical systems to produce a magnetic field of 1.5 Tesla.
D. Uses electrical current running through wire that has no electric resistance when kept cold.
E. Often used in mobile MRI settings.
34. When changing from a 2562 matrix with 2 NEX (SNR1) to 256 x 128 with 1 NEX (SNR2) with phase encoding
along the 128 matrix dimension (everything else unchanged), the SNR relationship will be:
A. SNR1 = 1/2 SNR2
B. SNR1 = 1/4 SNR2
C. SNR1 = SNR2
D. SNR1 = 2 SNR2
E. SNR1 = 4 SNR2
14. After collecting an echo from one slice, what must the MRI do to collect an echo from the next adjacent slice?
A. Increase or decrease the range of frequencies encoded into the echoes.
B. Increase or decrease the center frequency at which the RF energy is transmitted
C. Select slice with a different gradient coil.
D. Frequency encode with a different gradient coil.
15. In MRI, which of the following statements is False?
A. T1 relaxation and T2 relaxation occur simultaneously.
B. T2 relaxation usually occurs much more quickly than T1 relaxation.
C. T1 relaxation refers to longitudinal recovery of magnetization, while T2 relaxation refers to decay of the
D. Echoes collected in the transverse plane demonstrate only T2 relaxation differences.
16. Concerning image artifacts in MRI, which of the following may be used to eliminate wrap-around artifacts?
A. A larger field-of-view.
B. A saturation pulse over the unwanted anatomy.
C. Selecting a “no frequency wrap” and/or a “no phase wrap” option on the system.
D. A and B.
E. A, B, and C.
17. Which of the following is a characteristic of MR angiography acquisition?
A. The images are post-processed to make projection images.
B. Moving blood signal intensity is reduced to match the intensity of stationary tissue signals.
C. A set of flow-sensitive images is acquired.
D. The stationary tissues are made to have a high signal intensity.
E. A and C.
18. A rapid spin echo sequence is faster than a conventional spin echo pulse sequence because:
A. T2* contrast takes less time.
B. A shorter TR is usually used.
C. Several echoes for each slice are collected during each TR period.
D. A higher matrix is used.
19. __________ is the fastest MRI imaging technique available today.
A. Spin echo
B. Inversion recovery
C. Single-shot EPI
D. Rapid spin echo
20. For MR imaging, the patient’s weight must be entered accurately during patient registration to ________.
A. Enable calculation of the specific absorption rate (SAR).
B. Prevent damage to the RF coils.
C. Prevent artifacts in the images.
D. Prevent damage to the patient table.
21. In an MRI gradient recalled echo (GRE) acquisition sequence:
A. A 180o refocusing RF pulse is used to induce the formation of an echo.
B. Images produced have a significant dependence on magnetic field non-uniformities.
C. Tissue contrast is a strong function of T1.
D. “Spoiled” GRE is used to produce mostly T2* images.
E. “Spoiled” GRE has less sensitivity to chemical shift artifacts.
22. ____has the highest gyromagnetic ratio and thus the highest potential sensitivity for in vivo NMR spectroscopy.
23. Assuming accurate flip angles are applied, an inversion-recovery sequence is the most accurate way to measure:
C. Diffusion coefficient.
24. A signal void at the interface between an organ and adjacent fatty tissue is most likely caused by:
B. Chemical shift artifact.
C. Insufficient K-space sampling.
D. Poor signal-to-noise ratio.
25. In an NMR spectrum, the field homogeneity can be assessed by measuring the:
A. Chemical shift difference between two peaks.
B. Resonance frequency of a standard compound.
C. Ratio of two peak heights.
D. Line width of a single peak.
26. The sequence with fewest artifacts due to magnetic susceptibility differences (air-tissue interfaces) would be:
A. Gradient echo.
B. Fast gradient-echo.
D. Echo-planar imagine (EPI).
27. To increase SNR, the total number of k-space points acquired is kept the same, but more points are sampled in
the central region than at high k-space values. What happens to the resulting image?
A. More motion artifact
C. Ghost artifacts
D. Smaller field-of-view
28. FDA recommendations for___ limits the amount of power which may be applied during a given pulse sequence.
D. Larmor frequency.
29. Calcium deposits in acute hemorrhages can cancel T2 shortening effects of blood degradation products due to__.
C. Magnetization transfer
D. Hysteresis effects
30. The “lattice” in spin-lattice relaxation time referes to:
A. The digital image matrix
C. An imaginary frame.
D. A set of adjacent nuclei.
31. A Spin echo sequence with TE=200 mS and TR=3 seconds will yield an image that is primarily_____weighted.
B. Proton density
D. T1/ T2.
E. Magnetization transfer.
32. A gadolinium contrast agent will most likely cause one of the following:
A. Decreased proton density signal.
B. Increased proton density signal.
C. Increased T2
D. Increased T1.
E. Decreased T1.
33. The dominant artifact affecting image quality most often in the phase encoding direction is:
A. Wrap around.
D. RF zipper.
34. In MRI, the _______ gradient is on during the time the echo is measured.
A. Slice selection
B. Phase encoding
C. Frequency encoding
35. For hydrogen imaging in a 1.0 T MRI unit, the frequency of the RF signal is about _______ .
A. 40 Hz
B. 40 kHz
C. 40 MHz
D. 400 MHz
E. 4 GHz
36. Contrast in MRI is created by all of the following except:
A. Differences in hydrogen content.
B. Differences in T1 time of tissues.
C. Differences in T2 time of tissues.
D. Administration of a contrast agent.
E. Differences in atomic number.
37. In MRI the signal-to-noise ratio can be increased by all of the following except:
A. Switching from a volume to a surface coil.
B. Increasing the number of acquisitions.
C. Increasing the static magnetic field strength.
D. Decreasing the slice thickness.
E. Increasing TR.
38. Following a localizer image of cylindrical phantom (24 cm FOV, 256 2 matrix, 6 mm slice thickness, 1 average),
a high resolution image is obtained (12 cm FOV, 256 2 matrix, 3 mm slice thickness, 4 averages). All other
parameters are unchanged. What is the relationship between SNR in the 1 st and 2nd images?
A. SNR1/SNR2 = 1/4.
B. SNR1/SNR2 = 4.
C. SNR1/SNR2 = 2.
D. SNR1/SNR2 = 1.
39. In MRI, pure water will have a ________ T1 and a _______ T2.
A. Long, long
B. Long, short
C. Short, long
D. Short, short
40-42. Match the following MRI terms. (Answers may be used more than once.)
A. Shim coils
C. Gradient fields
40. ____Used to localize MR signal
41. ____Used to tip the net magnetization of spins
42. ____Used to adjust magnetic field uniformity
1. D Signal-to-noise ratio (SNR) is proportional to voxel volume, which increases with slice thickness. SNR
is proportional to the square root NEX (number of signal-averaging acquisitions). Proximity of the
receiver coil to the volume of interest increases the coil quality factor (which increases the SNR). Body
receiver coils have moderate quality factors. SNR is inversely proportional to square root of the
bandwidth (range of frequencies around the center frequency).
2. D Chemical shift refers to small differences in resonant frequency for protons of different chemical. Fat
protons precess ~3.5 Hz per Mhz slower than water protons (3.5 Hz/Mhz x 42.58 Mhz/T x 1.5T = 224 hz
at 1.5T). Since frequency differences are used to encode position along the frequency encoded (read-out)
direction, chemical shifts may cause misregistration: i.e., signals from different chemicals (i.e., fat and
water) in the same physical voxel are reconstructed into different image pixels in the read-out direction.
Since resonant frequency increases linearly with B o, so do chemical shifts. To avoid misregistration,
frequency-encoding gradients must be steep enough to impose voxel-to-voxel frequency differences
larger that the chemical shifts. For a 30 cm FOV and 256 2 matrix, a 5 mT/m gradient creates a voxel-to-
voxel difference of ~250 Hz (0.005 T/m x 42.58 mHz/T x 0.3 m/256 voxels). Alternatively chemical
shift in certain areas may be reduced by swapping phase and frequency directions.
3. C Proximity of the receiver coil to the patient’s volume of interest increases the coil quality factor, which
increases the SNR.
4. B Single or multiple Echo Planar Imaging (EPI) acquisition provides extremely fast imaging times. EP
images typically have poor SNR, low resolution (64x64 or 64x128 matrix) and exhibit chemical shift
artifacts. Nevertheless, EPI offers real-time “snap-shot” image capability with 50 mS or shorter total
imaging time. EPI is most useful for time-dependent physiological processes and functional imaging.
5. C All Inversion Recovery (IR) acquisitions emphasize T1 contrast by extending the longitudinal recovery
(T1) with an initial 180o RF inversion pulse. Variations in time of inversion (TI) controls image contrast.
Short Tau Inversion Recovery (STIR) uses very short TI. During longitudinal recovery, all tissues pass
through a “bounce point” or tissue longitudinal magnetization null. Judicious TI selection can suppress a
given tissue such as fat or chemical shift artifacts. Fluid Attenuated Inversion Recovery (FLAIR) uses a
longer TI to suppress signals from tissues with long T1 (fluids like CSF and other water-bound anatomy.
6. A Perfusion and Diffusion imaging use endogenous or exogenous “tagged” tracers to study delivery of
oxygen and nutrients to cells and removal of carbon dioxide from cells. Blood oxygen level-dependent
(BOLD) and “functional MR” imaging rely on differential contrast generated by blood metabolism in
active areas of the brain.
7. B Instruments employing a focused electron beam are very sensitive to magnetic fields, even the earth’s
magnetic field. Therefore, fringe fields greater than 1 gauss can cause sever problems. The safety limit is
5 gauss (pacemakers, etc.). The earth’s magnetic field is about 0.5 gauss.
9. B A typical gradient strength is 5 mT/meter = 0.005 T/meter = 0.5 g/cm; 0.5 g/cm x 30 cm = 15 gauss.
Gradient strength ranges from about 0.1 to 5 gauss/cm (1-50 mT/m). Expressing this in Hz:
0.005 T/m x 42.58 x10 mHz/T x 0.3 m = 0.06387 mHz = 63,870 Hz
10. C T2 relaxation only changes slightly with Bo< 4T. To a 1st approximation, we can assume it is unchanged.
11. E An RF coil may be used to transmit RF pulses and receive echoes. Alternately, one coil may transmit RF
pulses and another coil used for reception. Surface coils are optimal for imaging shallow structures.
brain imaging is best accomplished with a volumetric coil. Received signals are digitized by the analog-
to-digital converter, not the RF coils.
12. B A superconducting magnet can only be shut off by service engineers. It takes several minutes to shut off
because there is no resistance in the magnet windings to dissipate electrical current. Instead, current must
be drawn away from the windings and dissipated externally. A resistive magnet shuts once the current
supply to its windings is stopped. Descriptions A, C, D and E all apply to superconducting magnets.
13. C Doubling voxel volume doubles SNR (SNR is directly proportional to voxel volume). However, the total
number of excitations is decreased by a factor of 4 (2 to 1 NEX and 256 to 128 phase-encode steps),
which decreases SNR by √ 4= 2 (SNR is directly proportional to the square root of the number of
excitations). The overall relative SNR of the images is therefore unchanged.
14. B RF energy is transmitted everywhere in proximity to the RF coil. To excite protons in the next adjacent
slice, the center frequency of the transmitted RF must be increased or decreased to match the Larmour
frequency at the center of the next slice. The range of transmitted frequencies at each slice position must
remain constant for constant slice thickness. The RF pulse center frequency determines the slice position
along the slice selection gradient axis while the range of frequencies within an RF pulse determines slice
thickness. Slice selection with a different gradient would change slice orientation. Frequency encoding
with a different gradient would swap the phase and frequency direction in the image matrix.
15. D Echoes collected in the transverse plane can contain T1 and/or T2 information, depending on TR and TE
selections. T1 and T2 relaxation occur simultaneously, although T2 relaxation is usually much quicker.
Once an RF excitation pulse tips the magnetization into the transverse plane, transverse magnetization
begins to shrink due to T2 relaxation. At the same time, longitudinal magnetization begins to grow due
to T1 relaxation. Typical T2 relaxation times are on the order of tens of milliseconds whereas typical T1
times are on the order of hundreds of milliseconds to seconds.
16. E Wraparound can occur if any patient anatomy extends beyond the specified field-of-view (FOV) within
the slice plane. In the frequency encode direction, wraparound may be caused by undersampling the
received RF frequencies, leading to aliasing. In the phase encode direction, anatomy within the FOV are
encoded with phases between 0 and 360°; anatomy outside the FOV acquire phases <0° or >360°, which
cannot be distinguished from the equivalent phase between 0 and 360. For either direction, the anatomy
outside the FOV is mismapped to the opposite edge of the image. Wraparound may be avoided by using
a larger FOV (with a time penalty in the phase encode direction is pixel size in unchanged, since more
phase encoding steps are needed). It may also be eliminated by using a saturation (deleting) pulse over
the unwanted anatomy. Finally, many systems offer “no frequency wrap” (which low-pass filterd the RF
waveform prior to sampling) and/or “no phase wrap” options to correct for wraparound artifacts.
17. E There are a few key features in all MRA exams. 1st, a set of “base” images are collected, which must
have maximal signal from the flowing blood while minimizing stationary tissues signals: i.e., these
images exhibit bright blood vessels and dark stationary tissues. Since base image “slice” may contain
only a tiny portion of the blood vessels of interest, base images are post-processed using a Maximum
Intensity Projection algorithm to generate three-dimensional projection views of the vessels.
18. C A rapid spin echo (RSE) or fast spin echo (FSE) sequence consists of an initial 90o RF excitation pulse
followed by a series of 180o RF pulses within a single TR period for each slice. Each 180o RF pulse
generates another spin echo to which a different phase-encoding has been applied. For a matrix with N
voxels in the phase-encode direction and M echoes per T1 period, M of the required N phase encoding
steps are accomplished during each T1 period, thus filling the raw data set each slice more quickly (by a
factor of N/M) than conventional spin echo (CSE) sequences. Note, however, that since TE for the last
collected echo per slice is longer the typical TEs for CSE, TR in FSE are typically longer than for SE to
allow sufficient time to collect echoes for all interleaved slices, thus reducing the time advantage a bit.
19. C Echo planar is the fastest MR imaging technique available today. It is capable of acquiring all echoes
needed to reconstruct an image following a single RF excitation in a single TR period (known as single-
shot EPI. Although single-shot EPI is the fastest acquisition scheme to date, it requires special hardware
components and is associated with some image artifacts. Alternately, the MR echoes may be collected in
a few groups, each beginning with an RF excitation. This is called multi-shot EPI.
20. A SAR (specific absorption rate) is the amount of RF power transmitted into a patient over a given period
of time. It depends on patient weight as well as the time over which RF energy is transmitted, so patient
weight should be accurately entered during registration. Based on this weight, the MR system calculates
the specific absorption rate for each sequence before it is run. If a particular sequence exceeds the SAR
limits imposed by the FDA, parameter changes will be required in order to be able to run the sequence.
21. B GRE techniques use reversing magnetic field gradients, rather than 180o pulses as in SE, to form echoes.
Bo inhomogeneities and tissue susceptibilities are emphasized in GRE because dephasing and rephrasing
occur in the same precessional direction, so inhomogeneity effects do not cancel. Decay of transverse
magnetization is therefore a strong function of T2*. In “Spoiled” GRE, T2* effects can be eliminated to
yield mostly T1-weighted images. The downside of spoiled GRE technique is the increased sensitivity to
artifacts such as chemical shift and magnetic field non-uniformities as well as a lower SNR.
22. C H has the highest gyromagnetic ratio.
23. A By using various inversion times, inversion recovery pulse sequences may be used to measure T1.
24. B Different resonant frequencies (chemical shift) of protons on water and fat molecules may an artifact in
the frequency encoding direction. If the spectrometer frequency is centered on water, the reconstruction
algorithm spatially shifts the fat tissue in image, creating a signal void at the interface of the two tissues.
25. D The line width of a peak reflects field homogeneity: the narrower the line, the better the homogeneity.
26. C Except spin echo, these sequences use gradient-rephasings to form echoes, which do not refocus (cancel)
phase shifts within a voxel due to field inhomogeneities, such susceptibility effects. Spin echo pulse
sequences do refocus local field inhomogeneities and are thus less sensitive to susceptibility differences.
27. B Not sampling high k-space points means undersampling the high spatial frequencies, causing image blur.
High frequencies are required to accurately represent image areas which change rapidly (i.e., edges).
28. A SAR (specific absorption ratio) (watts/kg) has limits recommended by the FDA.
29. A Due to paired nuclear spins, calcium exhibits diamagnetism which diminishes the local magnetic field.
Paramagnetic agents, such as methemoglobin, augment the local magnetic field. In certain situations,
these properties offset, and the image does not exhibit the expected appearance of a hemorrhage.
30. B The lattice is the environment (the surrounding tissue) in which the spins exist.
31. C This “long” TE and “long” TR creates T2 weighting.
32. E Gadolinium, a paramagnetic agent, decreases T2 times because it creates magnetic inhomogeneities
which causes more rapid dephasing of transverse magnetization. It also interacts with the tissue lattice to
cause a decrease in T1 times; thus, the answer E.
33. C The phase encode gradient is applied with varying strengths throughout the acquisition; algorithms
assume no motion. Since even a slight motion will cause substantial phase shifts, most of the motion
artifacts are manifested in the phase encode gradient direction.
34. C During collection of an echo, the frequency-encode gradient is turned on to encode different frequencies
into the echo. The slice selection gradient is turned on during the application of the RF pulses. The
phase-encoding gradient is not turned on during the slice selection or frequency-encoding steps.
35. C The Larmour frequency for hydrogen is 42 MHz x magnetic field strength in Tesla.
36. E Unlike x-ray images, atomic number differences have no effect on MRI images
37. D Slice thickness reduces voxel volume, so reduces SNR. SNR is proportional to the square root NEX
(number of signal-averaging acquisitions). Proximity of the receiver coil to the volume of interest
increases the coil quality factor (which increases the SNR). Increasing TR may allow more longitudinal
magnetization to regenerate between acquisitions.
38. B In-slice voxel size is reduced by 4x and slice thickness by 2x, so SNR due to voxel size is reduced by 8x.
NEX increases by a 4x, so SNR increases by a factor of √ 4= 2. Therefore, SNR1/SNR2 = 8/2 = 4.
40. C Three perpendicular gradient fields in the three planes are required for spatial localization.
41. B The RF is used to excite (or tip) the hydrogen nuclei, and therefore the magnetization, out of
equilibrium with the main field.
42. A Shim coils and some cases, small blocks of metal are used to improve uniformity of the main (Bo)
Medical Physics/MRI Medical Physics Exam 2008/pf