Part III Magnetic Resonance Imaging
MRI Acquisition monitor check
0/5% and 95/100% patch, gray level steps distinct. If fails, call service and consult w/ supervising radiologist
ACR sagittal localizer sequence (lg phantom)
1 slice, sagittal spin-echo, TR=200 ms, TE=20 ms, slice thickness=20 mm, FOV=25 cm, matrix=256 × 256, NEX=1, scan time: 51-56 seconds
MRI Geometric distortion test
1) Ensure gradient amplifiers have been on for at least an hour 2) Measure length or diameter of ACR phantom Tolerance +/- 2 mm of true values
What should you do to set a non-acceptance baseline for MRI LCD?
1) Have Service run mfg's diagnostic tests and confirm it meets all specs 2) Reviews the tests and results 3) Have supervising radiologist look at several clinical images and confirm that IQ is as good as expected for this make and model 4) Collect QC data for 10 days
Spectral Peak method of B0 Uniformity
1) Position uniform, spherical phantom at center 2) Obtain a spectrum using manual tuning or prescan. Freq res should be << expected peak width 3) Measure FWHM using Larmor eq
Recommendations for MRI Acceptance (Report 100)
1) Vibration and RF shielding (prior to install) (Vendor?) 2) Magnetic fringe field mapping 3) Inventory 4) Mechanical checks 5) Emergency system checks 6) All ACR Annual Tests 7) Eddy current compensation (Vendor) 8) EPI: Ghosting, Geometric distortion, stability 9) Spectroscopy: VOI location accuracy, Spectral quality
MRI artifact evaluation weekly test
1) adjust WW/WL to show full range of pixel values by find approx. brightest area, use ROI, set WW to that value and WL to half that value 2) look for geometric distortion, ghosts, streaks, bright/dark spots, unusual features
Center frequency test procedure
1) find where center frequency is noted 2) display the central, sagittal slice through phantom acquired in localization test. For lg phantom, recommended 11 slices, starting at vertex of crossed wedges at the inferior end to crossed wedges at superior end 3) set up scan, note center frequency selected
What should the MRI QC log contain?
1. A section describing the facility's QC policies and procedures for the equipment covered by the log 2. A section of data forms where QC procedure results are recorded for each piece of equipment covered by the log 3. A section for recording notes on QC problems and corrective actions
ACR's 4 Steps of Quality Control
1. Acceptance testing of new units or major repairs 2. Establishing baseline performance 3. Detection of changes before they appear in imgs 4. Verification of correction of issues
MRI Weekly LCD test
1. Choose a specific slice, most sensitive to deviations in system performance (first in which you can or can't see all the discs)t 2. Adjust WW/WL for best view. Adjust for each slice, but consistent week to week 3. Record total complete spokes in each slice (irregular holes count, but compare w/ bkgd) Criteria: >=3 fewer spokes than determined by physicist, or however physicist specifies
When might a physicist want to be more strict with MRI geometric accuracy?
1. Images used for stereotactic surgical or radiation therapy planning 2. Assessment of the geometrical reproducibility of pulse sequences that use extremely high gradient amplitudes and/or switching rates (e.g., EPI) 3. Co-registration of images acquired at various time points and/or from multiple scanners
ACR MRI Physicist Tests
1. Magnetic field homogeneity evaluation 2. Slice-position accuracy 3. Slice-thickness accuracy 4. RF coil checks, including signal-to-noise ratio and image intensity uniformity of volume coils 5. Soft-copy (monitor) QC 6. MR safety program assessment
MRI low-contrast test
1. Physicist chooses one slice, the first without all spokes 2. count *complete* spokes Criteria: no change?
MRI Weekly QC Procedure
1. Position the lg phantom in the head coil or, for extremity MRI systems, position the sm phantom in the knee coil 2. Record center freq and transmitter attenuation (or transmitter gain) 3. Analyze images. Measure phantom dimensions. Assess high-contrast res and LCD. Note artifacts
MRI geometric accuracy test
1. Set WW to 0 or 1, adjust WL until about 1/2 of fluid is white 2. Set WW to WL in 1, then halve WL 3. Measure distance (length and transaxial)
MRI low-contrast objects
1. Slices 8-11 in lg phantom (6-7 sm) 2. 10 spokes of 3 disks each in each slice. 3. All spokes on a given slice have same contrast 4. All discs in a given spoke have the same dia. 5. Diameter of discs progresses from 7-1.5 mm
MRI High contrast test
1. display image of slice 1 2. Magnify between 2-4x 3. Set WW to a small value (<10% of signal range.) Adjust WL until holes are individually displayed. Must use one WW/WL setting for analysis 4. If all 4 holes in any row are distinguishable, the image is considered resolved L-R at this hole size. 5. Repeat for vertical Criteria: 1/0.8 mm in both directions (lg/sm)
Recommended max FOV for coil checks
2 x (phantom width in phase-encode)
Minimum SNR for coil checks
20
Diameter of MRI body phantom
27 cm
Typical FWHM of GE Spectral peak
55 Hz
For a 1.5 T system, the resonance freq. should be approx ___
63.87 MHz
How big should the Mean SNR ROI be for volume coils?
75% of area
Definition of Percent Signal Ghosting
<2.5% for head coil
MRI low-contrast detectability (Physicist)
>9 spokes in slices 8-11 w/ lg phantom (>9 in slices 6-7 w/ small). For 3T magnets >37 spokes. Should change less 3 spokes or less. Note recommended slices and total # of spokes is given in ACR manual
MRI Monitor brightness limit
>90 cd/m^2, min <1.2 Cd/m2
What is the MRI phantom filled with?
A solution of nickel chloride and sodium chloride
Parallel Imaging (MRI)
A technique that fills k-space by using the response of multiple receive RF coils that are coupled together with independent channels, so that data can be acquired simultaneously. Reduces FOV by 1/(# of channels). Reduction factor = amount scan time (PE steps) is reduced (# of channels). Requires processing to avoid wrap-around that looks at the sensitivity profile of each coil (signal changes by how close it is to each coil). Compare to standard imaging, in which signal from each coil is combined and analyzed.
Why might an MRI scanner's resonant frequency change?
Changes in the resonance frequency reflect changes in the static magnetic field (B0). Changes in B0 may be due to superconductor "run down" (typically < 1 ppm/day on superconducting magnets), changes in current density due to thermal or mechanical effects, shim-coil changes, or effects due to external ferromagnetic materials.
What should a physicist do if the LCD test fails on repeat?
Check SNR. If SNR change is acceptable, establish new action limits. Document the change
Why is using the spectral peak a crude and insensitive method of assessing homogeneity?
Contains no info regarding spatial variations of B0
Alternative way of measuring SNR in MRI
Create difference image and find standard deviation in phantom ROI
What can cause inappropriate image blurring or enhanced truncation artifacts in MRI?
Excessive filtration. Use of zero-fill interpolation or filters that enhance spatial res tends to cause truncation artifacts to become more apparent. Filters that enhance SNR tend to result in increased img blurring
MRI Slice Position Accuracy
Made on slices 1 and 11 of large phantom. Positioned so they are aligned with vertices of the crossed 45-deg wedges on either end of phantom. If the slice is displayed superiorly w.r.t. center, the bar on your right will be longer. Measure difference between them. Limit: magnitude < 5 mm (corr. to 2.5 mm displacement)
What to measure on surface coils
Max or average SNR
Why is transmitter (TX) gain or attenuation important?
Measure of power needed to nutate bulk magnetization by 90 deg. For the same coil and phantom, TX should remain constant if MR normal. A change is directly related to SNR and may indicate a problem in some part of the RF transmitter and/or its associated coils
Which MRIs are resonance frequency checks especially important on?
Mobile units and resistive magnet systems that undergo frequent ramping of the magnetic field
How can you remove phase variations not due to B0 inhomogeneity?
Most are not affected by changes in echo time. Can be removed by acquiring 2 GRE imgs, reconstructing phase maps, then subtracting the 2 maps. The resulting phase diff will be proportional only to mag field variations and diff in echo times
Convert ppm to Hz
Multiply by Larmor freq (for 1.5 T scanner, 64 MHz); 1 ppm = 64 Hz).
What standards organization does ACR base their SNR test on?
NEMA
Where does the ACR draw their MRI criteria from?
NEMA standards and AAPM report
What should a physicist do if they can't measure B0 uniformity?
Note in report, arrange for service engineer to provide most recent field map (from past 6 mo).
What should an MRI Quality Assurance Committee be comprised of, according to the ACR?
One or more radiologists, a QMP, a supervisory MR tech, other personnel involved in caring for pts (such as nurse, desk attendant, secretary), personnel outside radiology (such as medical/paramedical staff, referring physicians)
Why is checking MR resonance frequency important?
Operating an MRI off-resonance reduces an image's SNR, adversely affecting low-contrast detectability
What can cause MRI LCD to fail?
Phantom tilt, too low an acquisition matrix, excessive sharpening filters (excessive truncation artifacts resulting in poor visibility of outer disks), excessive ghosting, improper slice positioning
Which method provides the most accurate measurement of B0 homogeneity?
Phase-difference method. Can be performed using either 2-D or 3-D GRE sequences
What can cause dotted-line artifacts in MRI in the phase encoding direction?
RF interference. Check the integrity of RF shielding or identify the source (equipment or lighting in MRI room)
Distribution of background signal in MRI (function)
Rician
How is MRI LCD data analyzed?
Should be recorded weekly for trend analysis.
Where is high contrast resolution measured in the MRI phantom?
Slice 1
MRI geometry measurement image (transaxial)
Slice 5
What could cause a change in MRI low-contrast detectability?
Slice-positioning errors, intermittent ghosting, phantom tilting.
Causes of MR high contrast test failure
Spatial res should not change. Changes can be due to gradient field strength, eddy current compensation, unstable gradient amplifiers and/or the main (B0) magnetic field homogeneity being out of cal. Often will fail other QC tests
Ways of measuring MRI center frequency
Spectral beak, bandwidth difference, phase map, phase-difference map
ACR T1 Accreditation Parameters
TR = 500 ms, TE = 20 ms
Changes from 2004 MRI manual
Tests for interslice RF interference removed, alternative method for homogeneity and PIU limits, ghosting and low-contrast criteria
Why can't you use spin-echo to assess uniformity?
The 180 reverses and eliminates any effect of inhomogeneities
How is resonance frequency setting performed?
The phantom is positioned in the center of the magnet (with all gradient fields turned off), and the RF frequency is adjusted by controlling the RF synthesizer center frequency to achieve maximum signal.
MRI Laser light accuracy
The phantom is positioned so that the axial alignment light is on the superior edge of the grid structure. The location of the superior edge of the grid structure is then measured on the sagittal localizer image, and must be within ± 5 mm of the magnet isocenter)
What does the ACR manual say about MRI testing after major repair or upgrade?
The physicist SHALL repeat appropriate tests afterwards
How is an MRI artifact determined to be significant enough to be worthy of investigation?
The physicist decides
What is of note of the ACR's alternative phantom recommendations?
They use NEMA's recommendation of an electrical simulation of the body, rather than the AAPM's recommendation of nonconducting material
Typical MRI Bandwidth
Typical: 50 kHz Narrow: 5-20 kHz
How are the MRI high-contrast objects measured?
Upper Left = right-left Lower Right = up-down Center-to-center hole separation in a row is 2x hole diameter Hole diameter for large: 1.1 mm L pair, 1.0 mm C pair, 0.9 mm R pair (sm: 0.9, 0.8. 0.7)
What are good ways to keep tabs on MRI artifacts?
When they are noted, record any ancillary conditions that may be different from normal procedures. Techs should save raw data of img in which artifacts occur (can look in k-space)
How do you find the max/min signal in a volume coil?
Window level to get the last bright or first dark regions. Use ROI about .15% of the FOV or 1 cm2 for a 25 x 25 cm FOV
How to measure slice thickness accuracy
Window/level to WW 0, WL half of mean signal in ramps. Measure the length of each ramp.
DSV
diameter of spherical volume
What would cause an MRI geometry failure?
ferromagnetic object, one or more miscalibrated or insufficiently warmed-up gradient, too-low BW, improper adjustment of gradient offsets, passive or active magnet shims
Is every MRI problem fixed by the service engineer?
no, should also include recommendations concerning use of RF coils, appropriateness of pulse sequences, image processing, viewing conditions and the QC process.
What's a good trick for aligning the MRI phantom?
observe the laser on the top of the grid structure inside the phantom. use a 3 plane localizer
MRI visual checklist
pt. bed transport, alignment and system indicator lights, RF room integrity, emergency cart, safety lights, signage, monitors are present, working properly, are mechanically and electrically stable Anything not passing or missing should be replaced or corrected immediately
What should an MRI QA procedures manual contain, according to the ACR?
responsibilities and procedures for QC, records of most recent QC tests, description of orientation for MRI operators, procedures for use & maintenance of MRI, techniques to be used (including positioning, coils, pulse sequences, and contrast), precautions to protect pt, records of service & maintenance & QC meetings, procedures for cleaning/disinfecting MRI and equip.
MRI high-contrast spatial resolution test (Physicist)
use FOV of 250 mm w/ 256x256 matrix on a normal phantom, should be able to resolve 1-mm pattern
Why are changes in TX gain coarse?
usually reported in dB, made over entire volume of central slice
The large MRI phantom is used for ____ magnets, and the small phantom is used for _____ magnets
whole body, extremity
Center frequency of a 1.5 T magnet
~63.87 MHz
Pass criteria for MRI geometry test
+/- 2 mm
Criteria for MRI table position accuracy
+/- 5 mm
Why is the transmitter gain or attenuation test necessary?
After establishing the center frequency, the system acquires several signals while varying the transmitter attenuation (or gain) level so that imaging can proceed using the proper flip angles. Significant fluctuations in the transmitter attenuation (or gain) levels suggest problems with the RF chain.
What should be included in the physicist's review of MRI QC?
An evaluation of the site's safety guidelines, practices, and policies
What can can cause bright line artifacts in MRI?
DC offsets on MRI signal, especially on imgs w/ no signal averaging. Typically view-to-view phase alternation allows them to be located off to the side of the img and do not affect the utility. A less frequent source is an imperfect 180° in a SE acq. The position of the line depends on the value of the read-out gradient & can affect clinical img quality. Interference from ext sources of RF can cause linear "single frequency" or broadband artifacts.
What can cause a single bright/dark pixel at the center of an MRI image matrix?
DC-offset errors (dark if overflow or img processing has occurred). Due to improper scaling of low-freq components (typically DC) in the Fourier transformation of the time-domain signal
What is the drawback of the phase map method of assessing B0 uniformity?
Differences in B0 are not the only causes of spatial variations in phase. If the echo peak is not exactly centered in sample window, it will cause a linear ramp across the FOV. Also at high B0 strengths the RF does not penetrate water-filled phantoms as well as at low fields. These differences, as well as susceptibility differences, result in changes of phase by distance
When is MRI transmitter gain set?
During the pre-scan
How does the ACR handle flex and multi-channel coils?
Flex = surface coils. Multi-channel = recommend test each channel separately
ACR's definition of major MRI repairs
Gradient amplifiers, gradient coils, magnet, RF amplifier, digitizer boards, signal processing boards, coils
Why should MRI artifact images be checked for geometric distortion?
Gross geometric distortion can occur even on a system that passes the geometric accuracy test b/c the measurements are only along the primary axes of the phantom
What can cause an MRI image to appear to have a very bright background that is smooth (not speckled like random noise)?
If RF attenuation/gain is not set correctly during the prescan, signals acq during 1 or more P-E steps could be larger than the max allowable digitization step. This "saturates" the receiver so that the signal is not accurately digitized and img is not properly displayed following the inverse FT. Also spike signals caused by malfunctioning electronics
B0 spectral peak uniformity theory
If a B0 is perfectly homogeneous, all p+ precess at the same freq, get a delta function for FT. Random spin-spin interactions cause p+ to precess faster/slower, spreading the peak with the FWHM related to the avg T2. Long T2 = narrow peak (little spin-spin). Imperfections also broaden peak
B0 GRE uniformity theory
Ignoring T2 effects, if B0 is perfectly homogeneous then at the time of a gradient echo, all p+ would be completely in phase w/ each other. Spatial variations cause p+ in diff. parts of FOV to precess at slightly diff rates. The greater the diff in B0 across phantom volume, the greater the diff in p+ freq and the diff in phases at echo time. Echo time of 10 ms = 100 Hz (1/0.01 s), so a spin 25 Hz faster than center will be pi/2 (90°) off center
What to measure on volume coils
Image uniformity, SNR, percent signal ghosting
What problems could cause a change in TX gain/attenuation?
Impairment of RF transmission field, degradation of B0 homogeneity or noise added by RF receiver chain (active electronic components such as PIN diodes or inadequate isolation between TX and receiver (RX) channels)
Possible issues with B0 uniformity
Imperfections in magnet, external ferromagnetics, presence of pt, lack of shim correction. Causes difficult in obtaining uniform fat suppression, geometric distortion (particularly at low B0), img uniformity, makes wrap artifacts worse, compromises SNR in fast sequences
What can cause zipper artifacts?
In a SE sequence due to transverse magnetization being produced by imperfect slice excitation of the 180° refocusing pulse. The signal is constant from phase-encoding view to phase-encoding view so that it presents as a single freq. line of alternating intensity
How is change in transducer gain or attenuation usually measured?
In dB
What can cause MRI high-contrast spatial resolution to fail?
Inappropriate filtering, poor eddy current compensation, gradient miscalibration, B0 inhomogeneity, low acquisition BW
Causes of MRI geometric accuracy failure
Insufficiently warmed/stabilized gradient amplifiers, very low receiver BW, improper adjustment of gradient offsets, improper adjustment of passive/active magnet shims, ferromagnetic objects
What can cause MRI ghosting?
It is a very non-specific issue. However, it can be caused by motion or poor RF connections
MRI Monitor luminance uniformity
Limit <=30%
Definition of Percent Integral Uniformity
Limit >87.5% for head coil for 1.5 T systems >82% for 3T systems. PIUs less than 90% uncommon
ACR MRI Slice Thickness Calculation
Limit: 5 mm +/- 0.7 mm
For what kind of MRI QC failures does ACR recommend consulting with the supervising physician on if imaging can proceed?
Low contrast (physicist), film (technologist)
