mri for technologists mri systems and coil … mri systems and coil...mri for technologists . mri...

NO

T FO

R D

ISTR

IBU

TIO

N

MRI for Technologists

MRI Systems and Coil Technology PROGRAM INFORMATION MRI for Technologists is a training program designed to meet the needs of radiologic technologists entering or working in the field of magnetic resonance imaging (MRI). These units are designed to augment classroom instruction and on-site training for radiologic technology students and professionals planning to take the review board examinations, as well as to provide a review for those looking to refresh their knowledge base in MR imaging.

Original Release Date: Material Review Date Expiration Date:

October 2006 October 2012 November 1, 2019

This material will be reviewed for continued accuracy and relevance. Please go to www.icpme.us for up-to-date information regarding current expiration dates.

OVERVIEW

The skill of the technologist is the single most important factor in obtaining good quality diagnostic images. A successful MRI examination is the culmination of many factors under the direct control of the technologist.

MRI for Technologists: MRI Systems and Coil Technology introduces the learner to different bore types and field strengths, types of MR coils, and appropriate coil selection for optimizing the MR examination.

After completing this educational material, the reader will be able to:

• State the advantages of the short bore MR system

• Compare advantages and disadvantages of the low-field and high-field strength magnets

• Decide what type of magnet is most suitable for a specific MR examination

• Describe the major types of MR coils and when they should be used

• Discuss advantages and disadvantages of specific MR coils

• Select appropriate coil configuration to optimize the MR examination

EDUCATIONAL CREDIT This program has been approved by the American Society of Radiologic Technologists (ASRT) for 1.0 hour ARRT Category A continuing education credit.

http://www.icpme.us/

NO

T FO

R D

ISTR

IBU

TIO

N

ii

HOW TO RECEIVE CREDIT

Estimated time to complete this activity is 1.0 hour. The posttest and evaluation are required to receive credit and must be completed online.

• In order to access the posttest and evaluation, enroll in the online course at www.icpme.us.

• Read the entire activity. • Log in to your account at www.icpme.us to complete the posttest and evaluation,

accessible through the course link in your account. • A passing grade of at least 75% is required to be eligible to receive credit. • You may take the test up to three times. • Upon receipt of a passing grade, you will be able to print a certificate of credit from your

online account.

ACKNOWLEDGMENTS Our appreciation goes to Thomas Schrack, BS, ARMRIT for his review and update of this material. We would also like to acknowledge the original authors of this material: Jeffrey J. Brown, MD, FACR, MBA Washington University School of Medicine St. Louis, MO Thomas Schrack, BS, ARMRIT Fairfax Radiological Consultants Northern Virginia Community College Fairfax, VA

Alan H. Stolpen, MD, PhD University of Iowa Hospitals and Clinics Iowa City, IA Daniel R. Thedens, PhD University of Iowa Iowa City, IA

For their contributions to this material, special thanks go to: Stephen Dashnaw, ARMRIT Columbia University, New York, NY

Mark Flyer, MD Maimonides Medical Center, Brooklyn, NY

DISCLAIMER

Participants have an implied responsibility to use the newly acquired information to enhance patient outcomes and their own professional development. The information presented in this activity is not meant to serve as a guideline for patient management. Any procedures, medications, or other courses of diagnosis or treatment discussed or suggested in this activity should not be used by clinicians without evaluation of their patient’s conditions and possible contraindications on dangers in use, review of any applicable manufacturer’s product information, and comparison with recommendations of other authorities.

SPONSORED BY SUPPORTED BY AN EDUCATIONAL GRANT FROM



NO

T FO

R D

ISTR

IBU

TIO

N

FIELD STRENGTH

Low-field Scanners

Low-field scanners can be built around any ofthe three primary magnet types. Supercon-ducting magnets are used for field strengthsat the upper end of the low-field category(approximately 0.5 T and above) will be dis-cussed in the high-field scanner section.

Permanent Magnet Systems

In a permanent magnet system, the magneticfield is generated vertically between a pair of plates, and the patient is positioned hori-zontally between those plates. Commercial permanent magnet systems are capable ofgenerating a magnetic field of up to approxi-mately 0.3 T. The magnetic field in a perma-nent magnet system is always in effect and

MRI Systems and Coil TechnologyAfter completing this educational material, the reader will be able to:

State the advantages of the short bore MR system

Compare advantages and disadvantages of low-field and high-field strength magnets

Decide what type of magnet is most suitable for a specific MR examination

Describe the major types of MR coils and when they should be used

Discuss advantages and disadvantages of specific MR coils

Select appropriate coil configurations to optimize the MR examination

MRI SYSTEMSThree primary types of magnets are used in MRI systems: permanent, resistive, andsuperconducting. The operating principles and construction details of each type of magnetare detailed in other materials in this series. In this unit we will discuss the particularabilities and disadvantages of each magnet when performing different types of MRexaminations.

In addition to the primary type of magnet, clinical MR scanners can be broadly dividedinto two categories: low-field strength and high-field strength. Scanners having amagnetic field strength below 1.0 T are considered low-field, while scanners with fieldstrength of 1.0 T or greater are high-field systems.

NO

T FO

R D

ISTR

IBU

TIO

NMRI SYSTEMS AND COIL TECHNOLOGY 2

requires no power or cooling to maintain. The magnetic field is stable, provided that itstemperature is precisely controlled. However,permanent magnets have poorer field homo -geneity compared to superconducting mag-nets, and the volume of homogeneous field is also somewhat smaller. The weight of apermanent magnet is also enormous at15,000– 20,000 kg and may require rein -forced flooring to site the system, particularlyif the magnet is not located at ground level.

Resistive Magnet Systems

The magnetic field in a resistive magnet isgenerated with an electromagnet. The elec-tromagnet requires a large electric current(approximately 50 kilowatts for a 0.15 T field)to be passed through electrical windings toproduce the magnetic field. As a result, largeamounts of heat are generated, so the scan-ner must have a substantial cooling system.Like permanent magnet systems, the homo-geneity of the field and the overall volume ofhomogeneous field of a resistive magnet arereduced compared to a superconductingmagnet. Resistive magnets are available infield strengths to about 0.3 T. While a resistivemagnetic field can be turned off when not inuse, the cost of electricity when in use (inaddition to expensive cooling systems) can be quite high. Moreover, when the system isfirst turned on (meaning electrical current isturned on after being off for some amount oftime), it requires as much as 30 to 60 minutesto bring up the magnetic field and then tostabilize it, which causes a loss in productivityand efficiency.

Low-field scanners are used mostly in systemswith an open bore design, in systems designedto simultaneously perform interventional procedures, and for small specialized systemssuch as extremity MRI systems.

Superconducting Magnet Systems

High-field scanners are always built aroundsuperconducting magnets, and the majority ofinstalled MRI systems use this type of magnet.Clinical scanners with superconducting mag-nets are commercially available with fieldstrengths up to 3.0 T, and whole-body systemscurrently exist at research facilities with fieldsas high as 9.4 T.

Of the three magnet types, superconductingmagnets are capable of generating the mosthomogeneous magnetic field over the largestvolume. The magnet orientation is nearlyalways a traditional cylindrical bore configura-tion, although the length of the bore mayvary considerably. Superconducting magnetswith open bore design have appeared inrecent years as well, usually with fieldstrengths between 0.5 T and 1.0 T (Figures 1and 2).

A superconducting magnet is an electromag-net where the windings are made of a super-conducting material. Materials become“superconductive” when cooled to tempera-tures approaching 0° Kelvin. At this tempera-ture all molecular motion stops. Since the

Figure 1. Example of an open borepermanent magnet system.

NO

T FO

R D

ISTR

IBU

TIO


preponderance of any compound is actuallyspace created by the electron cloud, any elec-tricity passed through such a material willencounter no resistance. The material used insuperconducting MRI systems becomes super-conductive at 4° Kelvin when immersed inliquid helium. As long as this temperature ismaintained, the magnetic field will also persistwithout any additional electrical power. As aresult, the magnetic field is always present,without the need for continued electric volt-age. In essence, the electricity continues toflow indefinitely, even without a primarysource. In an emergency, the field can beturned off by slowly discharging the currentflowing through the windings, which mayrequire 30 minutes or more. This is known as“ramping down” the magnet. An immediaterelease of the cryogens—whether on purposein response to a life-threatening emergency orby accidental induction of excessive heat intothe magnet—is known as a “quench.” Anuncontrolled quench can be powerful anddangerous because super-cooled liquid heliumexpands exponentially and displaces breath-able oxygen from the room. If the cryogenventilation system is overpowered by thesheer amount of escaping helium, the oxygen

in the room can drop to levels below thatneeded to sustain human life.

The structure that contains the magnet andthe cryogens is called the cryostat. The cryo-stat is insulated from the surrounding environ-ment as much as possible to reduce heating ofthe cryogens. However, the liquid helium willslowly boil off and eventually requires refilling.In modern systems, a helium refill may not berequired for as much as a year or more.

FIELD STRENGTH CONSIDERATIONSAND COMPARISONS

The magnetic field strength of a scanneraffects a wide variety of imaging factors thatmust be taken into account when consideringthe most appropriate system and protocol touse for any specific examination (Table 1).

Signal-to-Noise Ratio

The signal-to-noise ratio (SNR) is significantlygreater at high fields as compared to lowfields. In general, the signal-to-noise ratioincreases in direct proportion to the mainmagnetic field strength. For example, with all other factors remaining equal, the SNR of

Figure 2a-c. Examples of high-field 1.5 T MRI systems.

a cb

NO

T FO

R D

ISTR

IBU

TIO


an exam performed at 3.0 T will be twice thatof the same exam at 1.5 T. However, otherhardware considerations such as bore size and coil design may also affect SNR betweensystems as well.

The increased SNR available at high fields for equivalent examinations may also permittrade-offs in the parameters for the exam.Image SNR is affected by the resolution,signal bandwidth, and other parameters. Thehigher SNR at high-field can be “traded” forimprovements in one or more of these fac-tors. For example, at 3.0 T it should be possi-ble to acquire images with the same SNR andexamination time, but with higher resolutionthan at 1.5 T. Alternatively, the same SNR and resolution of the 1.5 T exam could beacquired in a shorter amount of time. Coilsand other configuration details may affect thedegree to which this is possible.

Specific Absorption Rate

A limiting factor of high-field imaging, espe-cially at 3.0 T, is the increased energy disposi-tion for the same types of RF pulses. For anincrease of a factor of two in field strength

from 1.5 T to 3.0 T, RF energy depositionincreases by a factor of four. Simply put, thehigher the field strength, the more RF energyis needed to flip a proton away from a longi-tudinal orientation. This is especially trouble-some for pulse sequences that use many 180ºpulses, such as fast spin-echo (FSE). Restric-tions on SAR may reduce the number of slicesor increase the repetition time (TR) required(increasing the scan time) to achieve thedesired coverage.

Modifications to the acquisition protocol thatcan reduce SAR for high-field exams include:

• increased RF pulse spacing • decreased flip angles • use of hyperechoes • use of parallel imaging

Field Homogeneity

Imaging at higher field strengths such as 3.0 Tcan introduce greater magnetic field inhomo-geneities compared to 1.5 T imaging. Anotherfactor is the varying magnetic susceptibility inthe human body itself, which has a greaterinfluence at higher field strengths. Inhomo-

As Field StrengthParameter Increases Notes

Acoustic Noise Increases Higher field exerts greater force on conductors

Fat/Water Separation Increases Easier to perform fat saturation, but increases chemical shift artifact

SAR Increases 4X increase for 2X increase in field

SNR Increases Directly proportional to field strength

Susceptibility Increases Yields more artifacts in fast scanning but improves fMRI

T1 Increases Requires longer TR or reduced flip angle for same contrast

Table 1. A summary of the effects of different field strength systems.

NO

T FO

R D

ISTR

IBU

TIO


geneities in the main magnetic field can resultin image distortions, blurring, loss of resolu-tion, and a reduction in the usable field ofview (FOV) in the magnet. Inhomogeneitycan also adversely affect fat saturation due tothe corresponding chemical shifts. High qual-ity shimming is a necessity to minimize theeffects of inhomogeneity at higher fields.

Magnetic Susceptibility Effects

Magnetic susceptibility is the ability of a tissueto magnetize. Some tissues magnetize to agreat degree (blood, fat), while other tissues(compact bone) do not. Magnetic susceptibil-ity effects can be either desirable or undesir-able. The effect of susceptibility can createneeded contrast in tissues or cause artifactsthat are detrimental to the image quality.

Susceptibility effects are increased as imagingmoves to higher field strengths. This can beeither a disadvantage or an advantage,depending on the specific imaging application.

In applications with areas of high susceptibil-ity—such as the air-tissue interfaces aroundthe sinuses and near the heart-lung boundaryin cardiac imaging—many acquisitions, espe-cially fast imaging acquisitions like echo-planar and spiral imaging, may suffer greaterdistortions and signal loss compared to imag-ing at lower field strengths.

Some techniques to reduce susceptibility arti-facts include:

• reduction in echo time • reduction in echo spacing • increase in bandwidth • use of parallel imaging

Pulse sequences that rely on susceptibilityeffects to generate image contrast, however,benefit from higher field imaging. BloodOxygen Level–Dependent (BOLD) imaging

contrast results from the susceptibility effectsof oxygenated blood, which are increased athigher field strengths. Perfusion studies usingdynamic susceptibility contrast (DSC-MRI)may also benefit from the greater susceptibil-ity effects. MR spectroscopy is well known toimprove greatly at higher field strengths.There are two main reasons: higher SNRallows signals of weaker metabolites to bevisualized, and higher field strength equatesto greater chemical shift between metabolites,yielding more distinctive spectroscopy peaksfrom neighboring metabolites.

Other Imaging Effects

The main magnetic field strength affects several tissue parameters. T1 increases withhigher field strength. T2 is generally reduced,though the magnitude of the change is highlyvariable. T2* (T2 star) is also reduced, prima-rily due to inhomogeneity and susceptibilityeffects. The chemical shift between fat andwater grows proportionately with fieldstrength.

These effects, as well as many of the otherchanges that depend on field strength mentioned above require that imaging protocols be designed for the specific fieldstrength to be applied. Protocols that yieldacceptable images at one particular fieldstrength may give poor results or may not be possible at a different field strength. Image contrast can change considerably ifappropriate changes are not made in theimage acquisition parameters.

Coil Selection and Availability

Coils for specialized applications must bedesigned for a specific MR system model andfield strength. Since 1.5 T systems continue tobe the most commonly used, coil availabilityand quality is greatest for these systems.Lower field strength scanners are normally

NO

T FO

R D

ISTR

IBU

TIO


used for routine examinations. In the case ofspecial purpose systems, such as extremityscanners, coil selection is not a large issuebecause the scanner is designed for only anarrow range of applications.

3.0 T scanning has become widely clinicallyavailable in the last few years. Coil selectionfor 3.0 T scanners is presently more limitedthan that for 1.5 T. Initial applications of clini-cal 3.0 T imaging have been mostly in thehead, so head coils were the first to be devel-oped. Other areas, such as musculoskeletalimaging, lagged in coil development. Thetrend towards more 3.0 T clinical sites hasspurred the development of newer high-chan-nel coils. Now that coil development for theseapplications is occurring, applications such asbody and MSK imaging are greatly benefitingfrom 3.0 T imaging. As a result, 3.0 T imagingis now a mainstay in many imaging facilities.

Acoustic Noise

The source of acoustic noise during an MRexamination is the interaction of the largestatic magnetic field with the gradient coils.The pulsing of current through the gradientcoils in the main magnetic field causes thecoils to vibrate, much like a loudspeaker. Thevolume of the noise depends on both theconstruction and mounting of the gradientcoils, as well as the gradient amplitudes andtiming for any particular pulse sequence.Increasing the magnetic field will also increasethe amplitude of these forces, and thus thenoise. Patient use of earplugs or headphonesis vital to prevent hearing damage, and usagebecomes more important at higher fieldstrengths and during examinations usinghigher performance gradients. Soundproofingthe magnet room is also important for MRpersonnel in the surrounding environment.

SUMMARY OF FIELD EFFECTS

It is important to understand the physicallimits of lower field strength systems. All MRimaging requires the technologist to balancethe elements of spatial resolution, SNR, andscan time, although what is considered anacceptable balance is somewhat subjective.Higher field strength systems have inter -mittently higher SNR. Low-field strengthscanners have an important place in medicalimaging; however, it must be rememberedthat to achieve the same diagnostic quality asa high-field system using a low-field system(all things being equal), at least one of thefollowing, or some combination thereof, must occur:

• the scan time must be increased

• the images will have lower SNR and/or resolution

• fewer images will be obtained

SCANNER BORE DESIGN

Scanners can also be categorized by the typeof bore for patient entry into the system. Thetwo main types are cylindrical bore and openbore systems.

Cylindrical Bore

Virtually all high-field systems (1.5 T andhigher) are cylindrical bore systems. Acylindrical bore magnet has a tube-shapedopening in the main magnet. The patient isplaced on a motorized table that slides theregion of interest (ROI) into position in thecenter of the tube, so that the patient issurrounded on all sides. The length anddiameter of the bore vary widely amongcurrently used MRI systems.

NO

T FO

R D

ISTR

IBU

TIO


Older high-field MRI scanners (typically pro-duced before 1995) often have a longer bore.The bore in these systems may be as long as250 cm, which means the patient will becompletely inside the bore for almost all typesof examinations. The length was required togenerate a field in the center of the borehomogeneous enough for high quality imag-ing. However, the longer bore magnets canbe problematic for even mildly claustrophobicpatients.

In recent years a great deal of effort has goneinto magnet design to shorten the length ofthe bore. The theory is that the shorter borewill result in the lower likelihood of claustro-phobia for patients. However, to date therehave been no studies that either prove or disprove this theory.

Newer systems have a much shorter borelength, typically in the range of 150–180 cm;a 1.5 T magnet with a bore length of just 125cm was recently released. The smaller overallsize of the machine makes it easier to install,because less space is required for siting. Thehomogeneous region of the field remainsabout the same as in older long bore systems;however, the shorter bore magnet designshave a more difficult time maintaining highhomogeneity over a large FOV. As a result, forsome shorter bore systems the maximum use-able FOV is reduced as compared to older,long bore systems.

Another important variable in cylindrical boresystems is the bore diameter. The diameterreflects the size of the opening for the patient;typically, this is in the range of 55–60 cm,though sizes as large as 70 cm have recentlybecome available for clinical use. The actualFOV that can be achieved for imaging issomewhat smaller, around 40–50 cm.

Open Bore

An open bore magnet design uses a pair ofmagnet coils arranged vertically, with a rela-tively large gap in the center for patientaccess. The patient slides in from the side andis not completely surrounded by the system.The open bore accommodates larger subjectsand lessens the feelings of claustrophobia insome patients. Patients who could not bescanned in a cylindrical bore system due tosize or claustrophobia may be able to toleratean open bore magnet.

Open bore systems can use any of the threemajor magnet types, although permanentmagnets are most common. Open bore systems are most often low-field, which makes them easier to install and popular inoutpatient settings due to the relatively smallsize of the system. Due to the limitations oflow-field systems, however, open bore scan-ners are not always an option for patients whomight prefer them for reasons of comfort.

Considerations

As a rule, the higher field strength and betterfield homogeneity of cylindrical bore magnetsresults in considerably higher image qualitywhen compared to open bore magnets. Thetwo main advantages of open bore systemsare for obese patients and for those whoexperience claustrophobia. The design and“feel” of open bore systems may permit manypatients to be scanned that might otherwiseneed sedation or be unable to tolerate thescan. Open bore designs can accommodatemuch larger patients than some closed boresystems. Low-field systems also offer advan-tages such as less visible motion artifacts, lessSAR to the patient (allowing more slices perTR), and a far lower acoustic noise level forthe patient. However, the lower field, longerexamination times, and limited range of imag-

NO

T FO

R D

ISTR

IBU

TIO


ing protocols mean that open bore systemsmay not be appropriate for a particular exami-nation. The continuing trend of cylindricalbore systems with shorter and wider bore sizes(comparable to a CT scanner) may eventuallypermit high-field scanning of both large andpotentially claustrophobic subjects.

GRADIENTS

The gradient system of an MR scanner gener-ates an additional set of magnetic fields thatvary depending on location within the magnet.The dependence of these fields on location isused to determine the spatial variation in thesignals given off by the tissues. Therefore, it is the gradient fields that make it possible forthe system to create 2D and 3D images of tissues.

The gradient system of a scanner is made upof three sets of coils. Each coil can generate amagnetic field that varies along one of thethree principal directions of X, Y, and Z (Figure

3). The change in the magnetic field producedby the gradient is directly proportional to thedistance from the center of the magnet andthe strength of the gradient applied. The gra-dients are switched on and off at a rapid ratethroughout the scan, and the amplitude of thefields they generate can be varied to acquireenough data to generate a 2D image or 3Dvolume. Several characteristics of the gradientsystem impact their performance and thetypes of pulse sequences that can be applied.

Amplitude

The gradient amplitude measures the maxi-mum variation in the magnetic field that thegradient coils are capable of creating. In thepast this was referred to as gradient strength.However, because other gradient attributesalso describe gradient performance, eg, risetime and slew rate, employing the term“strength” to characterize gradient amplitudeis not very useful. The maximum amplitude ismeasured in units of mT/m, with typicalvalues ranging from 10–60 mT/m in high-field clinical scanners; however, the gradientcoils can be programmed to create any ampli-tude up to the maximum value. For example,a 10 mT/m gradient amplitude using the Xgradient (the X direction is the left-right direction looking into the scanner) means thatthe magnetic field will be 1.0 mT higher at alocation 10 cm to the right of the center ofthe magnet (10 mT/m times 0.1 meter).

The amplitude of the gradient affects the spatial resolution and FOV that the system is capable of creating. A higher maximumgradient will allow the system to acquireimages using thinner slices. Higher gradientamplitudes are also needed to perform rapidimaging sequences such as echo-planar imaging (EPI).

B0

x

y

z

Figure 3. The three primary directions inthe cylindrical, or traditional, magnet.

NO

T FO

R D

ISTR

IBU

TIO


Slew Rate

The gradient fields are not turned on con-stantly but pulsed on and off during imageacquisition to obtain the needed data. Switch-ing from the off condition to the maximumgradient cannot happen instantaneously; ittakes some time. The time required to switchthe gradients from zero amplitude to themaximum can be measured two ways: risetime and slew rate.

The rise time is defined as the time requiredto switch the gradients from zero to theirmaximum amplitude. A typical rise time isbetween 0.1 to 1.0 ms on many commercialsystems. This time lag is particularly importantfor fast sequences such as EPI, where the gra-dients are switched back and forth manytimes during a single data acquisition. Thefaster the rise time, the more rapidly theimage can be acquired for these sequences. InEPI, where the gradients go from max positive

to max negative, the switching of the gradi-ent from –max (max –G) to + max (max +G)is referred to as ramp time (Figure 4).

Another way of describing gradient perform-ance is the slew rate of the gradients, whichmeasures the maximum possible change inthe gradient per unit time. The units of thismeasurement are then either milliTesla permeter per millisecond (mT/m/ms) or Tesla permeter per second (T/m/s); these units willactually have the same value. For example, agradient system with 40 mT/m maximumamplitude and a 0.4 ms rise time would havea maximum slew rate of 100 mT/m/ms or100 T/m/s (40 mT/m divided by 0.4 ms, or0.04 T/m divided by 0.0004).

The slew rate will also determine how fast themagnetic field changes while the gradientsare switching on and off. For some pulsesequences, the slew rate that can actually beused may be restricted by FDA regulations.This restriction is designed to limit peripheralnerve stimulation, not the capabilities of thegradient coils and amplifiers.

Duty Cycle

To rapidly switch the gradients on and offrequires large amounts of power to be pro-vided to the gradient amplifiers, resulting inthe generation of substantial amounts of heat.

The duty cycle of the gradients specifies thefraction of time that the gradients can beturned on compared to the total scanningtime. A duty cycle of 100% means that thegradients can be switched on and off continu-ously. A duty cycle of less than 100% meansthat the gradients must “rest” for some timeduring the pulse sequence in order to preventthe accumulation of too much heat, whichcould damage the system. The duty cycle maydetermine the shortest TR that is possible with

Figure 4. Gradient rise time and gradientramp time should not be confused. Risetime is the time required for a gradient togo from zero to its maximum amplitude(max G). Conversely, ramp time is the timerequired for a gradient to go from its max-imum positive amplitude (max +G) to itsmaximum negative amplitude (max –G).Since EPI sequences rely heavily on gradi-ent reversals in the echo train, ramp timebecomes far more relevant than rise time.

Rise time

Max +G

Max –G

Ramp time

0

NO

T FO

R D

ISTR

IBU

TIO


a particular pulse sequence to make sure thatthe gradients are not switched on too often.All major manufacturers of clinical MRI sys-tems currently report 100% duty cycles ontheir systems.

IMAGING CAPABILITIES

As noted above, the capabilities of the gradi-ent system will partly determine the imagingparameters that can be achieved. The maxi-mum gradient amplitude directly affects theminimum slice thickness that can be pre-scribed. Along with the RF system, it may alsoinfluence the highest resolution the system iscapable of in a given scan time. The slew ratewill also have an effect on how fast imagescan be acquired. In most sequences, no imagedata are acquired while the gradients areramping up to their maximum or back downto zero, so a high slew rate minimizes thisdead time in the acquisition and may permit ashorter repetition to be achieved. The dutycycle allowed sets limits on the TR and acts asanother factor in determining the rate atwhich images can be acquired.

The primary applications that demand high-performance gradients are fast scanning tech-niques such as EPI, fast 3D gradient-echoimaging, and extremely rapid FSE imaging. InEPI, the amount of data acquired after eachexcitation pulse is much greater than for astandard spin-echo, or even for a gradient-echo pulse sequence. EPI is achieved by veryrapidly switching the gradients from positiveto negative amplitudes to acquire multiplelines of data from the same excitation. In theextreme case, entire images can be acquiredin a single excitation, which requires thesystem to switch the gradients from positiveto negative 128 or more times within 50–200ms. Such fast acquisition requires large ampli-tude gradients to acquire data with high

bandwidth, high slew rates to minimize theamount of time between the positive andnegative gradient acquisitions, and a highduty cycle to repeat the process many timesin a short repetition time.

SITING CONSIDERATIONS

The strong magnetic field associated with theMR scanner requires that care be taken in thesiting of the magnet. It is important to con-sider the location and the surrounding envi-ronment, both to assure that interference andartifacts do not arise as a result of outsideconditions, as well as to guarantee that themagnetic field neither poses a safety hazardnor affects other nearby instrumentation.

Magnetic Shielding

Magnetic shielding refers to the methodsused to keep the strong magnetic field con-fined to a small area surrounding the magnet,preferably within the room where the magnetis housed. The magnetic field surrounding themagnet is known as the fringe field. Thefringe field is primarily an issue of safety, asthe magnetic field can attract objects from adistance, causing them to fly towards themagnet. The fringe field can also causedamage to pacemakers and other implanteddevices. In practice, the safe area around a magnet is defined as the region outsidewhich the magnetic field is less than 5 Gauss(0.5 mT); this is considered the safety limit forunrestricted exposure to regular public traffic.In general, higher field strength magnets willhave a larger fringe field.

There are two general categories of magneticshielding: passive and active. Passive shieldingwas generally applied to magnets producedbefore 1990; it involves installing large ironplates in the walls, ceiling, and floor of theroom that houses the scanner. It is also

NO

T FO

R D

ISTR

IBU

TIO


possible to directly surround the magnet itself with iron plates, sometimes called self-shielding. The iron plates have the effect ofconcentrating the magnetic field, so that it is dramatically reduced outside the confinesof the shielding structure. The amount ofmetal required for passive shielding results in enormous weight and greatly increases the load on the floor. While passive shieldingis an effective method of containing thefringe field, over the course of approximately10 years the steel plates themselves canbecome magnetized and will eventually have to be replaced (Figure 5).

Nearly all new high-field magnets utilize activeshielding to restrict the magnet’s fringe field.Active shielding in high-field magnets consistsof an extra set of windings that surround thecoil and generate its own magnetic field. Thisextra set of windings creates another mag-netic field that is the polar opposite of themain magnetic field and which cancels out themain field in the area surrounding themagnet. As a result, the fringe fields are con-

siderably smaller surrounding the system. Thereduction in the fringe fields is enough thatthe 5 Gauss line can easily be located withinthe scanner room without the need for anyadditional shielding structure. Active shieldingreduces the weight, cost, and complexity ofthe scanner installation (Figure 6).

RF Shielding

In addition to confining the magnetic field,the design of the scanner room must alsominimize interference from RF energy. Thesignals generated and recorded in MRI are inthe same frequency range as those used byradio, television, and other electronic devices.The coils that receive the signal that generatesimages are sensitive to any source of signalsin this frequency range.

RF shielding is used to prevent outside signalsfrom entering the scanner room and interfer-ing with the desired signals for image acquisi-tion. RF shielding can consist of a “cage” ofcopper sheeting or copper mesh that sur-rounds the room or the magnet. If built into

Figure 5. Example of a 1.5 T system with apassive shield design.

Figure 6. Example of a 3.0 T activelyshielded system.

NO

T FO

R D

ISTR

IBU

TIO


the room, it must be applied to the windowsand doors of the room as well as to the wallsto prevent RF interference from all pathways.

Without proper shielding, external RF energyfrom common devices found in the hospitalsetting can generate signals that are as largeas, or larger than, the signals generated forimaging. The outside signal source may be amonitoring device, or even a light bulb withinthe scan room. The result of RF interferenceare images that are noisy or display character-istic line artifacts in the image, often referredto as a zipper artifact, discussed in anotherunit in this series.

SUMMARY

High-field scanners remain the most com-monly used MRI systems, with 1.5 T fieldstrengths still the most widely used in clinicalpractice. However, the emerging trend istowards a greater number of 3.0 T systems,particularly in new installations. The growingpopularity of 3.0 T systems carries the poten-tial for higher quality and higher resolutionimaging, but there are both advantages andlimitations to consider when imaging withhigh-field systems. It is important to realizethat 3.0 T systems require extra diligence infollowing proper screening and facility safetyprocedures because of their extreme torque(or pull). However, this should not be con-strued to imply that safety procedures can bemore lax at lower field strengths. Low-fieldsystems still function in specialized areas, soan understanding of their applications andcapabilities is important for a completeassessment of the most suitable MR examina-tion technique.

NO

T FO

R D

ISTR

IBU

TIO


COIL TECHNOLOGYThe main magnetic field strength that sets up the alignment of the protons is a majordeterminant of the quality and SNR the system is capable of providing. The performance ofthe gradient systems that provide the ability to localize signals and generate images alsoimpacts the speed and resolution possible for the MRI system. The other major subsystemof electronics in an MRI system is the RF system. The RF system is composed of the elec-tronics and coils that generate and record the MR signal that is decoded (reconstructed) toform an image.

BASIC OPERATION OF RF COILS

The RF system performs two primary tasks:exciting the tissue protons by generating aresonating RF signal and converting thealtered signals received from the excited tissueinto a form used to create the image. Putsimply, it sends signal and then receives thesignal back in a different form.

B1 Field Generation

The RF signal transmitted to the body is anoscillating magnetic field at the resonant frequency of the tissue of interest. This oscillating magnetic field is referred to as the B1 field (recall that the main magneticfield is referred to as the B0 field.) The first ofthe two primary tasks of the RF system is togenerate an RF signal, a small, oscillatingmagnetic field that sends energy to the protons in the body tissue such that theybegin to spin or precess. This process is calledexciting the protons and is accomplished bypassing electrical current through the wind-ings of the transmitter RF coil. The transmitter coil is usually cylindrical, as inthe case of the body coil or the transmit/receive head coil.

Receivers and Preamplifiers

The task of the receiver system is to convertthe analog signals received by the RF coil intodigital form suitable for the computer to use

to generate and display an image. A block dia-gram of the signal path from receiver coil toreconstruction computer is shown in Figure 7.

The signal conversion process is complicatedby two factors. First, the signals picked up by the RF coil generate voltages only in therange of microvolts. To be useful, these sig-nals must be greatly amplified. Second, the

Figure 7. The receiver path signal conver-sion process is shown beginning at thereceiver coil and ending at the reconstruc-tion computer.

ReconstructionComputer

RF receiver coil

Reference (center)frequency

RF preamplifier

Phase sensitive detectorPhase sensitive detector

Audiofrequency filtersAmplifierSamper

Audiofrequency filtersAmplifierSamper

NO

T FO

R D

ISTR

IBU

TIO

N

signal is oscillating as it is received at thesame frequency as the B1 field generated inthe transmit phase, in the radiofrequencyrange of megahertz. For example, for a 1.5 Tsystem the signals recorded will oscillate at afrequency of about 63 MHz. Filtering toreduce noise and other unwanted signal is dif-ficult in this frequency range. Converting theanalog signal to the digital form needed forimage reconstruction is also impractical.

The first operation of the receiver system is toamplify the signal generated in the receivercoil to a usable range. This is performed bythe preamplifier. The preamplifier strengthensthe entire signal received by several orders ofmagnitude to raise the signal levels from themicrovolt range to the millivolt range. One ofthe most important functions required of thepreamplifier is to introduce an absolute mini-mum of noise into the signal. Otherwise, thenoise will be passed along the signal path andgreatly degrade the resultant image quality.

The next step is to shift the frequency of thepreamplified signal from the RF range of

megahertz down to the audio frequencyrange of kilohertz. The frequency shiftingoperation is accomplished by mixing the pre-amplified signal with the original transmittedfrequency of the RF excitation. In effect, thisprocess subtracts the transmitted frequencyfrom the received signal. Before the mixingoperation, the signals of interest are all in anarrow range around the transmission fre-quency, so this process converts the oscillationfrequency of the signals of interest to a rangeof several kilohertz (a factor of about 1000)lower than the original. The precise range offrequencies considered is determined by thebandwidth and gradients used during thesignal readout. The signals in the resultinglower range are much more convenient toprocess, because they are in the same rangeas audio signals and can be processed bycomputer in much the same way. In particular,the signals can be further amplified, filtered toeliminate noise and other unwanted signals,and digitized to permit computer processingand reconstruction to generate images.

Figure 8a-b. (a) The diagram on the left demonstrates a quadrature coil design that is 90 degreesout of phase. (b) The diagram on the right shows two RF signals. While these signals are the same frequency, the RF pulses are 90 degrees out of phase with each another.

Receiver coil #2

2 RF signals 90 degrees out of phase with each other

Receiver coil #1

90 degree offset

MRI SYSTEMS AND COIL TECHNOLOGY 14

NO

T FO

R D

ISTR

IBU

TIO

N

Quadrature detection

In many MRI systems, two detectors are usedin the process of converting the signal from itsoriginal RF range to the audio frequencyrange. The two detectors differ by a 90° phasechange in the RF signal used for referencewith the incoming signal, and the audio fre-quency range signals are processed separatelybefore image formation. Figure 8 illustratesthis concept. The advantage is that the signaloutput from these two detectors is highly cor-related, but the noise in each path is random,resulting in the same effect as two signal aver-ages with respect to the noise. The outcome isan improvement in the SNR by a factor ofabout 40%.

TRANSMIT AND RECEIVE COILS

Transmit and receive coils use a single coil unitto perform both the excitation (transmit) andthe signal readout (receive) functions.

Volume Coils

Common volume coils include the body coil,head coils, and some extremity coils. Volumecoils are designed to have a uniform responseand SNR over the volume covered, as com-pared to receive-only surface coils. However,the compromise required for large volumecoverage is a reduced SNR over the volume ofinterest. Transmit and receive volume coilscommonly use the quadrature techniquesdescribed above to maximize the potentialSNR for large volume acquisitions.

Volume coils have several designs:

• Birdcage coil. The birdcage coil designis widely used in transmit/receive bodycoils, head coils, and extremity coils. Itconsists of a cylindrical arrangement ofwires along the longitudinal direction.The arrangement of the wires provides

a highly uniform field for RF excitationand signal readout.

• Saddle coil. The saddle coil geometry iscomposed of two loops of a conductorwrapped around opposite sides of acylinder. The saddle geometry wasamong the earlier designs for volumecoils. It is a simple design, but it suffersfrom greater RF inhomogeneity whencompared to birdcage coils of similar size.

• Solenoidal coil. A solenoidal coil is builtwith closely spaced windings or coppertape on a cylindrical form. Solenoidalcoils have a small diameter compared to their length. Solenoidal coils are best suited for use on longer areas ofana tomy (greater length) where deeppenetration is not required due to thesmall diameter of the coil. Elbows andknees are examples of anatomy thatbenefits from the use of solenoidal coils.

RECEIVE-ONLY COILS

Receive-only coils perform only the signaldetection function of the MR signal acquisi-tion process. Receive-only coils must be usedin combination with a separate transmit coil(usually the body coil) to generate the RFexcitation. Because they do not need totransmit RF energy for excitation, receive-only coils can be placed close to the anatomyof interest for imaging and can be muchmore specialized to achieve the desiredanatomic coverage.

Surface coils

Surface coils generally refer to the family ofcoils that act as receive-only coils and aredesigned with a specific anatomical part inmind. There are many different types of surface coils, including both single-coil and


NO

T FO

R D

ISTR

IBU

TIO

N

multi-coil designs. Surface coils are made in all shapes and sizes, depending on the area of interest.

All other factors remaining the same, thesmaller the coil size, the better the SNR. However, the smaller size also means asmaller detection area. Additionally, the depthat which the coil will detect signal is directlyproportional to the diameter of the coil; inother words, the smaller the coil diameter, theless the coil can penetrate tissue to capturesignal. In particular, the area in which the coilcan detect signal will extend to a distanceroughly covered by the diameter of the coilelement in the lateral directions, and to adepth approximately equal to the radius ofthe coil. Smaller surface coils are best used for anatomy near the surface of the bodybecause the sensitivity of the coil is greatestnear the coil and falls away as the distancefrom the coil is increased. The closer the coilcan be placed to the anatomy to be imaged,the more sensitive the coil will be to thatanatomy, and the SNR and image quality will be correspondingly higher.

The advantage of a surface coil is that the coilis sensitive to a considerably smaller localvolume, depending on the coil diameter. Bothsignal and noise are received only in a regionof the body near the coil. The higher SNR andreduced coverage means that the anatomy

of interest can be imaged with higher spatialresolution. Smaller surface coils are thereforeused when high resolution of small structuresis required.

The sensitivity of surface coils is not uniform.The signal intensity will be reduced as the dis-tance of the coil from the patient is increased.For surface coils with a single element, thissignal variation may cause intensity fluctua-tions over the anatomy of interest. Properplacement of the coil is important to minimizethese effects.

Surface coils come in a variety of shapes andsizes for specific anatomy or for general pur-pose usage. Examples of specific coils are pre-sented at the end of this material. Some ofthe general characteristics of large and smallsurface coils are shown in Table 2.

Surface Coil Correction

The increasing availability of clinical 3.0 Tscanners—as well as the development of 8,16, and 32 receiver systems—has resulted inincreased visualization of very high signalintensity in tissue close to the surface coil. In asimple example, a sagittal lumbar spine usinga phased array spine coil exhibits extremelybright signal in the posterior fat close to thecoil, yet the spinal structures further from thecoil and next to the fat are far less bright(phased array coils are described later). The


Large Surface Coils Small Surface Coils

More uniform sensitivity Sensitive only near the body surface

Lower SNR Higher SNR

Not suitable for small FOV Small FOV and high resolution imaging possible

Applications include chest, torso, hip Applications include small anatomy such as wrist, elbow

Table 2. General characteristics of surface coils.

NO

T FO

R D

ISTR

IBU

TIO

N

great difference in contrast can make it diffi-cult to change the brightness and contrastadjustments in image display for optimalappearance. If the technologist uses window/level for the spinal cord, for example, the pos-terior fat may be so bright as to be distractingto the radiologist; if window/level is used to“dampen” the extremely bright fat signal, thespinal structures may appear too dark forproper visualization.

Another unsettling example of the phenome-non of nonuniform contrast comes from theincreased use of high-element phased arraycoils (8 or more elements) that completelysurround the anatomy of interest. Two exam-ples are a 12-element head array coil and an18-element chest/torso array coil. The coildesign heightens the brightness of all thetissue close to the coil, giving the entire imagea haloed appearance, with anatomy closest tothe coil brighter than the anatomy toward thecenter of the image. This appearance can bevery distracting to the radiologist, and

window/leveling the image to create a moreuniform image is virtually impossible. Theimages that follow offer examples of thesignal intensity issues referred to here.

To correct, or at least reduce, the excessivebrightness near the coil, MR manufacturershave instituted numerous software-drivenprograms to correct the intensity. Surface coilintensity correction schemes are implementedeither in the reconstruction process or as apostprocessing feature.

The simplest and most common method ofsurface coil intensity correction postprocessinguses software to view a pixel intensity histo -gram across the image (see Figures 9 and 10).In the example of the lumbar spine discussedabove, the histogram shows very high signalintensity on one side of the image, whichthen sharply drops off because the signalfrom the anatomy is farther from the coil.Then the computer reorganizes the imagedata to reduce the brighter intensities and


Figure 9a-b. (a) This histogram across the image shown in (b) demonstrates the uncorrected highvariation in signal intensity across the image; (b) The sagittal T2-weighted image of the lumbar spine displays characteristic very bright signal from fat posterior and proximate to thespine phased array coil. The extremely bright signal can be visually distracting to the radiologist.

UNCORRECTED IMAGE

Pixelintensityacrossthe

image

FOV across the image

a b

NO

T FO

R D

ISTR

IBU

TIO

N

increase the darker intensities. Finally, the dataare reconstructed to display an image withmore uniform intensity level across the image.

Phased Array Coils

As mentioned earlier, smaller surface coils yieldhigher SNR but less coverage. Conversely, alarger surface coil will provide greater cover-age, but at the cost of higher SNR.

Initial attempts to create a surface coil withthe SNR of a small coil and the coverage areaof a large coil used several smaller coils joinedtogether such that they all received signal atthe same time. This arrangement is called amulti-coil configuration (Figure 11). Unfortu-nately, the resulting images yielded only extracoverage, with SNR measurements beingequivalent only to a single large coil. Theproblem with the multi-coil configurationarose from the design, in that all the coilswere connected to the same signal receiver.

When individual coils were connected to adedicated receiver the SNR gains were real-ized, along with the increased coverage. This


Figure 10a-b. (a) The corrected data corresponding to the filtered image. (b) The same lumbarspine image after correction using an image intensity filter. Note the suppression of the excessivefat signal. The images shown in Figures 9a and 10a have exactly the same brightness and con-trast levels.

CORRECTED IMAGE

Pixelintensityacrossthe

image

FOV across the image

a b

Figure 11 shows a multi-coil configuration,which is not the same as a phased arraycoil. Multiple coils are connected to a sin-gle receiver, so the system reconstructs animage that appears to be from a singlelarge surface coil.

RECEIVER

The result-ing imagehas a largecoveragearea, butwith lowerSNR than inFigure 12.

NO

T FO

R D

ISTR

IBU

TIO

N

discovery in surface coil technology revolu-tionized MR imaging and is known as aphased array surface coil design (Figures 12and 13).

Phased array coils are composed of severalindividual surface coil elements joined into asingle unit. Each coil element independentlycovers a limited volume. The signal from each element is separately recorded and processed; combining the signals from all coil elements creates the final image. Thiscoil design brings the SNR advantages of sur-face coils to examinations with a larger FOVand anatomic coverage.

The advantages of phased array coils are thatthe SNR characteristics are the same for eachcoil element, but the volume that can be cov-ered is greater. The design and arrangementof the coil elements also help to make the

final image intensity pattern more uniform.Phased array coils are frequently designed towrap around the body to cover the chest,torso, or hip with higher SNR than the bodycoil and more uniform coverage than a singlecoil element.

SPECIFIC COIL TYPES

Many types of coils have been created forspecialized applications to maximize the reso-lution and image quality for specific types ofexaminations. Some common types of special-ized coils are discussed below.

Head Coils

The brain has always been the most oftenimaged part of the human body using MRI.Therefore, it is logical that surface coil designsfor head/brain imaging are numerous and


Figure 12 shows a phased array coil config-uration. Each coil is connected to its owndedicated receiver, so that each image isprocessed individually. The individualimages are then combined in the lastrecontruction phase and a single image isdisplayed, having the SNR of a small coilbut the area coverage of a larger coil.

1

2

3

4

1

2

3

4

CompositeImage

Receivers Images

The finalcompositeimage hasthe samelarge

coveragearea, butcompara-tively higherSNR than inFigure 11.

Figure 13. An example of a phased arraycoil. The photo shows a six-element quad-rature cervical-thoracic-lumbar (CTL)phased array coil. The coil elements can beused in any grouping of 2, 3, or 4 coil ele-ments. Typically these combinations arethe top 2, 3, or 4 coil elements; the mid-dle 3 or 4 coil elements; or the bottom 2,3, or 4 coil elements. The particular combi-nation selected is dependent on the areaof interest.

NO

T FO

R D

ISTR

IBU

TIO

N

diverse. Typically referred to as “head coils,”they can be either single channel transmit-receive coils (typically in a birdcage design) ora multi-element, phased array, receive-onlycoil. The phased array coil designs have vari-ous sizes and shapes. Some phased arrayhead coil designs may have the same birdcageappearance as a single-channel transmit-receive coil, but every MR system designatesthe type of each coil. The technologist isresponsible for knowing and selecting the cor-rect coil for any examination based on thespecific anatomy being imaged and theSNR/resolution requirements for that type ofexamination.

Neurovascular Array Coils

One specialized type of head coil is the neuro -vascular array coil. The neurovascular arraycoil differs from the head-only coils men-tioned above in two areas: the coil is availableonly in a phased array design, and it hasgreater length to cover not only the head, but also the neck, cervical spine, and into themediastinum. The primary use of the neuro -vascular array coil is to allow imaging of thebrain, the intracranial vascular system, and the carotid arteries, all commonly requestedexaminations. Before the advent of the neuro -vascular array coil, these examinationsrequired the technologist to change coilsduring the scan process, switching from thehead coil to a cervical or neck coil. Thisprocess was time consuming and inefficient.Using the neurovascular array coil, all of therequired structures and arteries can be visual-ized without needing to change coils in themiddle of the scan. Moreover, because of thephased array design, the individual coil ele-ments can be selected so that only the brain is imaged for better SNR, or the entire headand neck area can be imaged for completecoverage. An example of a neurovasculararray coil is shown in Figure 14.

Spine Array Coils

Second to the head coil, the spine array coil isby far the most often used. The spine arraycoil is typically long enough to cover theentire length of an average adult spine andcontains a sufficient number of coil elementsto image any portion of the spine. Dependingon the particular manufacturer’s design, theentire spine can be imaged in one large FOV,or any portion of the spine may be imaged ina smaller and more detailed FOV.

Muscloskeletal Coils

Many different types of musculoskeletal coils,both phased array and nonphased array, existfor virtually all joints in the body. The mostcommonly used coils are for imaging theknee, shoulder, and wrist (Figures 15 and 16).


Figure 14. An example of a neurovasculararray coil. The coil elements are arrangedso the head can be imaged alone or incombination with the cervical spine, softtissue of the neck, or carotid MRA imagingsequence. The head-only part of the coilcan be activated while the neck portionremains inactive if imaging only the brain.When performing carotid MRA, the cover-age requirements are from the aortic archto the Circle of Willis; in this case both thehead and neck portions of this coil can beactivated together and receive signal inunison.

NO

T FO

R D

ISTR

IBU

TIO

N

Breast Coils

Specialized breast coils are receive-only coilsdesigned for both unilateral and bilateralimaging of the breasts, as well as of the chestwall and axillary tissue. The most commondesigns use multiple channels that can beswitched on independently for imaging of one

or both breasts. State-of-the-art coils aredesigned with an open architecture to allowaccess to breasts for patient positioning,needle localization, or MR-guided biopsy.

Peripheral Vascular Array Coil

Complete imaging of the peripheral vascula-ture for magnetic resonance angiography(MRA)—such as runoff examinations—requires a very large FOV to image from the renal arteries all the way to the feet.Peripheral vascular coils are multiple-element,receive-only coils that permit high SNR imag-ing of vascular structures. These coils can also be used for soft tissue and musculo -skeletal imaging in the lower extremities(Figure 17).


Figure 15. This 3-element array coil isdesigned specifically for imaging the wrist.The small coil element design ensures highSNR, enabling high resolution images.

Figure 16. This quadrature design transmit-receive coil is used for knee and foot/ankle imaging. Note the column at the topfor distal foot imaging and right-angleankle positioning.

Figure 17 shows a 12-element peripheralvascular array coil designed for use duringperipheral run-off MRA examinations. Thecoil is divided into anterior and posteriorhalves. The patient lays on the larger poste-rior half of the coil and the anterior half isplaced on top the patient. The coil elementsare activated in three groups during theimaging sequence: top, middle, and bottomstations. This coil is designed to providecoverage from the renal arteries to the feet.

NO

T FO

R D

ISTR

IBU

TIO

N

The most common configurations of periph-eral vascular array coils use two or more com-ponents to “wrap” the entire lower body ofthe patient. The complete patient setup can beperformed at the beginning of the examina-tion with no additional adjustments of the coilplacement. The multi-element coils improveSNR and uniformity of coverage. Peripheralvascular coils are most suitable for high resolu-tion MRA and bolus tracking studies.

Endocavitary Coils

Endocavitary coils are specialized disposablereceive-only coils for very specific clinical indi-cations. Endocavitary coils are used primarilyfor imaging structures in the pelvic regionsuch as the prostate gland, the cervix, and the colon. For each of these applications, thecoil is designed to be positioned inside therectum, which brings it near to the organ orpathology to be imaged. By bringing thereceive element much closer to the organ ortumor, high quality, high resolution imagescan be acquired over a small FOV (down toabout 8 cm) to identify important structuresfor diagnosis.

The basic design of coils for all endocavitaryapplications is similar, but the design is indi-vidually optimized for prostate, cervical, orcolon imaging. For example, the endorectalcoil is inflatable and has adjustable deflectionfor precise positioning closest to the anatomyof interest (Figure 18).

SPECIALIZED ARRAY COILS FORPARALLEL IMAGING

Parallel imaging has been a breakthroughtechnology for high speed imaging. Parallelimaging has the potential to dramaticallyimprove scan speeds for many types of rou-tine clinical examinations and make higherresolution possible in a reasonable scan time.

Parallel imaging is also becoming importantfor 3.0 T imaging to overcome the limitationsimposed by FDA-regulated SAR. SAR is themaximum amount of RF energy that can betransmitted to the patient within a designatedamount of time according to the patient’sweight as defined by the FDA. Parallel imaging uses multiple-elementcoils (phased array coils) for signal receptionand multiple receiver channels to independ-ently process the image data from each coilelement. The data from the individual coil elements are then combined to reconstructthe final image.

Concept and Theory of Operation

The fundamental concept of parallel imagingis based on the fact that the different ele-ments of a multiple-element receiver coil willhave different sensitivities to a particular loca-tion in the body. The amplitude of the signalproduced in a coil element from any specificlocation in the body decreases as the distancefrom that location increases. Historically, thischange in amplitude was considered a draw-


Figure 18.Endorectalcoils.

NO

T FO

R D

ISTR

IBU

TIO

N

back of multiple-element coils because itintroduces a nonuniform intensity variationover the FOV; however, this phenomenon canbe exploited to allow much faster imaging.

The key to parallel imaging is to recognizethat spatial differences in coil sensitivities canbe used as a way of encoding spatial informa-tion. Each voxel of tissue will generate a dif-ferent signal in each coil element, dependingon the position of the tissue relative to thatcoil. By comparing the responses in each ofthe coils, some information about the spatiallocation of the tissue can be generated. Thetissue will generate more signal in the coil ele-ments it is closer to, as compared to the coilelements farther away. Therefore, not only isspatial information generated by frequencyand phase-encoding gradients but also fromthe arrangement of and signal in the coil ele-ments. Special reconstruction software canextract the spatial information from the sig-nals arising from the independent coils anduse it to replace some of the information thatwould normally be acquired by phase encod-ing. The end result is that all of the neededspatial information can be acquired usingfewer phase-encode lines, which means thatthe acquisition time is shortened. Ideally, thenumber of phase-encode lines could bereduced by a factor equal to the number ofcoil elements, but in practice the reductionfactor that can be achieved is somewhatlower. The trade-off for a shortened acquisi-tion time is a lower SNR.

To apply parallel imaging, the particular sensi-tivity pattern in each of the coils must beknown. The sensitivity information can betaken either from the known and fixed designof the coil or (more commonly) from a lowresolution calibration scan that estimates thesensitivity pattern during the examination.

These data are needed during the imagereconstruction to reduce the acquisition timeand still generate a full FOV image.

Although parallel imaging can accelerate animaging examination by a factor equal to thenumber of coil elements in the array coil(since each coil element provides unique spa-tial information), in practice this accelerationfactor is rarely achieved. Additionally, thereare trade-offs that must be considered whendetermining what acceleration factors can beused to produce a high quality examination.

The biggest drawback in a parallel imagingacquisition is the reduction in SNR as com-pared to a fully sampled scan. SNR is relatedto the total duration of the signal encoded bythe readout gradient. Thus, if an accelerationfactor of four is used for a parallel acquisition,a loss of a factor of at least two can beexpected in SNR; this decrease is in additionto the SNR loss related to the coil geometry.The loss of SNR coupled with the dramaticincrease in imaging speed makes parallelimaging perfect for applications like bodyimaging, where FOV and slice thickness yieldsufficient SNR to allow parallel imaging with-out a prohibitive loss of SNR. Musculoskeletalimaging applications may not be good candi-dates for parallel imaging, since they are typi-cally very high resolution (small FOV and thinslice thickness) and therefore already tend toexperience low SNR. Nevertheless, manytypes of acquisitions (such as FSE techniques)have sufficient SNR that even with the SNRloss due to the reduced scan time and coilgeometry, high quality images are stillacquired in a much shorter time, permittinghigher resolution, or reduced motion andother artifacts.


NO

T FO

R D

ISTR

IBU

TIO

N

Special Coil Design Issues

Since it is the arrangement of coil elements—and their relative responses—that make parallel imaging possible, it is not surprisingthat the geometry of the coils is very impor-tant in achieving maximum benefit from parallel imaging. Theoretically, the number of coil elements determines the maximumamount of acceleration possible. However, in an actual coil design there will be someoverlap of the areas covered by the individualcoil elements, and the information acquiredfrom each coil will not be completely inde-pendent. This means that using the maximumacceleration factor (of the number of coil elements) will likely leave some artifact in theresultant images.

Also, because the coil response is not uniform,noise in the images will be amplified in someareas in the final image reconstruction. Theamount of increased noise and its distributionin the FOV is measured by the geometryfactor (or g-factor) of the coil arrangement.

Ideally, the g-factor would be equal to one,indicating no additional noise enhancementfrom the coil arrangement. The g-factordepends very strongly on the exact geometryof the coils and usually must be measuredwith simulations. Careful attention to the coil arrangement can minimize the g-factorand permit acceleration factors close to themaximum (Figure 19).

Vendor-specific Implementations

Not all MRI manufacturers offer a parallelimaging pulse sequence, but those that douse a vendor-specific name. For example,Philips Healthcare offers two types of parallelimaging known as SENSE and SMASH,although SENSE is by far used more often.GE Healthcare offers two sequences known asASSET and ARC, and the Siemens sequence isknown as iPAT. Although each sequence hasslight differences in design and implementa-tion, all parallel imaging techniques abide bythe same basic principles and with the sameadvantages and pitfalls.


Figure 19 demonstrates appropriate body coverage using the different types of coils.

HEADHead coil or

neurovascular array.

CHEST/BODYBody coil, torso/body array, cardiac array, or

breast array.

SPINECTL array.

KNEE Knee coil orextremityarray.

FOOT/ANKLE Knee coil or

extremity array.

NECKNeck coil, CTL arrayor neuro vascular

array.

PELVISBody coil, torso/

body array, or endorectal coil.

NO

T FO

R D

ISTR

IBU

TIO

N

SUMMARY

Surface coils are an extremely integral part ofany complete MR system. The coils, as well asthe entire RF subsystem, are as important asany other major component and as importantas the gradients and even the magnet itself.Advances in coil technology and in the entireRF chain have contributed significantly toMRI's exponential growth in applications andreductions in scan time while improvingimage quality. Continued technologicaladvancements in the area of number ofreceiver channels seed the development ofnew high-channel phased array coils. Thecontinued evolution of clinical ultra high-fieldimagining, such as 3.0 T and higher, will alsofoster new developments in surface coil tech-nology.


NO

T FO

R D

ISTR

IBU

TIO


Figures

1 Courtesy of Hitachi, Ltd.

2a Courtesy of GE Healthcare

2b, 6 Courtesy of Siemens Medical Systems

2c Courtesy of Philips Medical Systems

4 Tom Schrack, BS, ARMRIT

5 Courtesy of Fairfax Radiology Consultants

6 Courtesy of Siemens Medical Systems

7 Dan Thedens, PhD and Alan Stolpen, MD, PhD; adapted from Hinshaw WS, Lent AH. An Introduction to NMR imaging: From the Bloch equation to the imaging equation. Proceedings of the IEEE 1983;71:338-350.

8, 9b, 10b, 11-12 Tom Schrack, BS, ARMRIT

9a, 10a, 13-17 Courtesy of Fairfax Radiology Consultants

18 Courtesy of Medrad, Inc.

19 Adapted from Siemens Medical Solutions and MRI for Technologists, Module 2: Technical Considerations of MRI, ©2002

Tables

1 Dan Thedens, PhD

2 Dan Thedens, PhD and Alan Stolpen, MD, PhD

Additional Reading

Pruessmann KP, Weiger M, Scheidegger MB, Boesiger P. SENSE: Sensitivity encoding for fast MRI. MagnReson Med. 1999:42(5):952-962.

Schmitt F, Dewdney A, Renz W. Hardware considerations for MRI imaging physics. MR Clinics of NA.1999 November;7(4):734-735.

Sodickson DK, Manning WJ. Simultaneous acquisition of spatial harmonics (SMASH): fast imaging withradiofrequency coil arrays. Magn Reson Med. 1997:38(4):591-603.

Triolo J. Low field and open MRI: Complement of competition for high field? Lecture presented at MRAdvocates Development meeting, January 15, 2000; Orlando, FL.

NO

T FO

R D

ISTR

IBU

TIO


ADC apparent diffusion coefficients

AUC area under the curve

B1 the RF transmission field

B0 the main magnetic field

bFFE balanced fast field echo

BOLD blood oxygen level dependent

cm centimeter

CNS central nervous system

CSF cerebrospinal fluid

CT computed tomography

CTL cervical-thoracic-lumbar

dB/dt time-varying B fields (gradient-altered B0)

dB change in the main magnetic field(B0) –OR– decibel

dBA decibel attenuation

DSC dynamic susceptibility contrast

DTI diffusion tensor imaging

eADC exponential ADC map

EKG electrocardiogram

EPI echo-planar imaging

FLAIR fluid attenuated inversion recovery

fMRI functional magnetic resonanceimaging

FNH focal nodular hyperplasia

FS fat suppressed

FSE fast spin-echo

FOV field of view

G Gauss

g-factor geometry factor

GRE gradient echo

Hz hertz

IR inversion recovery

IV intravenous

kHz kiloHertz (1,000 hertz)

KVO keep vein open

mHz megaHertz (1,000,000 hertz)

min minute(s)

MinIP minimum intensity projection

MIP maximum intensity projection

mm millimeter

mmol/kg millimole per kilogram

mOsm/kg milliosmoles per kilogram

mOsm/L milliosmoles per Liter

MPR multiplanar reconstruction

MRI magnetic resonance imaging

MRA magnetic resonance angiography

MRCP magnetic resonance cholangiopan-creatagraphy

MRS magnetic resonance spectroscopy

MRV magnetic resonance venography

mT milliTesla

mT/m milliTesla per meter

mT/m/s milliTesla per meter per second

MTT mean transit time

NAA N-acetyl aspartate

NEX number of excitations (also NSA)

nm nanometer

NSA number of signal averages (alsoNEX)

OCP o-cresolpthalein complexone technique

Osm/kg osmoles per kilogram

PACS picture archiving communicationsystem

PNS peripheral nerve stimulation

r1 T1 recovery

r2 T2 decay

RBW receive bandwidth

rCBV relative cerebral blood volume

A B B R E V I AT I O N S O F T E R M S

NO

T FO

R D

ISTR

IBU

TIO


RES reticuloendothelial system

RF radiofrequency

ROI region of interest

SAR specific absorption rate

SE spin-echo

SNR signal-to-noise ratio

SPGR spoiled gradient echo

SPIO superparamagnetic iron oxide

SSFP steady state free precession

STIR short tau inversion recovery

T Tesla

T1 time for 63% of a tissue’s longitu-dinal magnetization to recovery

T2 time for 63% of a tissue’s trans-verse magnetization to decay

TE echo time

T/m/s tesla per meter per second

TOF time of flight

TR time to recovery –OR– repetitiontime

TSE turbo spin-echo (Siemens MedicalSystems term for fast spin-echo)

UAE uterine artery embolization

USPIO ultrasmall superparamagnetic ironoxide

W/kg watts/kilogram

NO

T FO

R D

ISTR

IBU

TIO


absolute zerothe theoretical point at which all molecular motionstops, measured as 00 Kelvin

aliasinga common artifact caused when the FOV selectedis smaller than the area of tissue excited; alsoknown as “wrap-around”

algorithm1

a step-by-step method of solving a problem ormaking decisions, as in making a diagnosis. Anestablished mechanical procedure for solving certain mathematical problems

artifact1in radiology, a substance or structure not naturallypresent in living tissue, but of which an authenticimage appears in a radiograph

BOLD (blood oxygen level dependent) imagingan imaging technique based on the EPI pulsesequence in which the patient is given mental tasksto perform while undergoing MR examination.Increases in blood flow to affected areas of thebrain are postprocessed into functional maps indicating brain activity.

borethe tubular portion of the magnet in which thepatient is placed

contraindicationan absolute reason or cause not to proceed with adiagnostic examination or procedure. In MRI it istypically a patient condition, such as an embeddedcardiac non-MR compatible pacemaker, thatprohibits the patient from undergoing MRexamination.

contrastdifferences in signal intensity between two adjacentareas on an MR image

cryogen super-cooled liquid gas used to cool a given material to near absolute zero, thus becomingsuperconductive; in MR, liquid helium

cryostatthe vessel in which the magnet is immersed in liquid helium

diffusion imagingan imaging technique usually based on the EPIpulse sequence; it is used to indicate the amount ofwater absorption across a brain cell membrane

duty cyclethe amount of time (stated in percentage) that anMR system gradients can be “active” before theymust be turned off in order to cool; most MR man-ufacturers report a duty cycle of 100%, indicatingthe gradients can run at full power continuously

echo planar imaging (EPI)a very fast pulse sequence characterized by rapidoscillation of the frequency encoding gradient tocreate an echo train; EPI fills k-space quickly

echo spacingthe time from the peak of one echo in an echotrain to the next

echo trainthe series of echoes created by a FSE or EPI pulsesequence

electromagnetic spectrumcontinuous series of different types of electromag-netic radiation, ordered according to wave-lengthof frequency

fast spin-echo (FSE)a rapid pulse sequence characterized by a series of 1800 RF echo pulses used to create numerousechoes within a single TR, thus filling k-space morequickly

fat suppressionany of the methods used to reduce signal from faton an MR image

ferromagneticmaterials that react to a magnetic force; all ironand some stainless steel are ferromagnetic

field of view (FOV)an area of tissue or anatomy to be imaged in anMRI scan

fluid attenuated inversion recovery (FLAIR)inversion recovery-based pulse sequence that utilizes a relative long TI in order to suppress signalfrom a long T2 tissue while maintaining heavy T2-weighting throughout the rest of the image

flip anglethe rotation of the amount of RF energy used toexcite some portion of protons in the longitudinalplane into the transverse plane

G LO S S A R Y

NO

T FO

R D

ISTR

IBU

TIO


frequencycycles per unit time; usually measured as cycles persecond, or hertz (Hz)

frequency encodinggeneration of frequency differences along a partic-ular direction of a tissue slice for use in spatial local-ization of MR signal

fringe fieldthe extended magnetic field generated by an MRImagnet system that is outside the magnet bore, orscan area. The fringe field decreases in strength asthe distance from the magnet increases

functional MRI (fMRI)any of the imaging techniques that demonstratefunction of anatomical structure; examples includespectroscopy, dynamic liver imaging, cardiac imag-ing, and BOLD imaging

Gauss1[Karl Friedrich Gauss, Geman mathematician andphysicist, 1777-1855] the cgs unit of magnetic fluxdensity, equal to 10-4 tesla. Symbol, G

gradient (magnetic field)a magnetic field that changes in strength along agiven direction

gradient amplitudethe degree to which a gradient field can vary fromzero to the peak maximum point; typically meas-ured in milliTesla/meter (mT/m)

gradient coilone of six coils (3 pair) placed in the three orthogo-nal planes (denoted X,Y, and Z) that generate smallmagnetic fields along the plane of the main mag-netic field; it is used for slice selection, phase andfrequency encoding

gradient echopulse sequence characterized usually by a <900 RFexcitation pulse and an echo generated by gradientreversal instead of a 1800 RF echo pulse

gradient moment nullinga method to reduce flow artifact in which one, orall three, pair of gradients are pulsed to dephaseand rephase spins that flow along the axis of thatgradient

homogeneity the uniformity of any field; in MRI it is the unifor-mity of the B0 field

inhomogeneityabsence of homogeneity or uniformity; inhomo-geneity in a magnetic field may occur as one areaof the field deviates from the average magneticfield strength

inversion recoverypulse sequence characterized by an initial 1800 RFinversion pulse

k-space the domain in which the information from each phase-encoding step is placed during a pulsesequence. Each “filled in” line of k-space corre-sponds to each phase-encoding step; once therequired amount of k-space is filled, image recon-struction can begin

Larmor frequencythe frequency at which magnetic resonance is produced in a sample of hydrogen nuclei, or othertypes of nuclei used in MRI

ligand1

a molecule that donates or accepts a pair of elec-trons to form a coordinate covalent bond with thecentral metal atom of a coordination complex

magnetic momentthe net magnetic properties of an object or particle(such as a magnetic dipole)

magnetic shielding metal surrounding an MR magnet used to containthe main magnetic field fringe field within accept-able limits; magnetic shielding can be passive (steellined walls) or active (built into the system)

magnetic susceptibilitythe degree to which a tissue can become magnetized

maximum intensity projection (MIP)a ray tracing algorithm where a ray goes through adesignated imaging block or volume. Signal intensi-ty is designated based on nearness to the observer

motion artifactan artifact or signal not naturally present in livingtissue, but which appears on MRI film due tomovement of muscle or fluid or motion of anybody part

multiplanar reconstruction (MPR)two-dimensional views of a single voxel thicknessof vascular structures reconstructed from three-dimensional or multi-slice images

NO

T FO

R D

ISTR

IBU

TIO


magnetic resonance angiography (MRA)1a form of magnetic resonance imaging used tostudy blood vessels and blood flow, particularly fordetection of abnormalities in the arteries and veinsthroughout the body

magnetic resonance imaging (MRI)a method of visualizing soft tissues of the body byapplying an external magnetic field that makes itpossible to distinguish between hydrogen atoms indifferent environments

magnetic resonance spectroscopy (MRS)MR examination in which the data collected are notreconstructed into an image, but into a spectrum of signals based on metabolite presence within thetissue being examined; a type of fMRI

nanometer (nm)1a unit of linear measure equal to one-billionth of a meter, 109 meter

number of excitations (NEX)the number of cycles of completed k-space filling;also known as number of signal averages (NSA)

number of signal averages (NSA)see number of excitations

paramagnetica substance with magnetic properties that may sig-nificantly reduce T1 and T2 relaxation times in MRI

perfusion imagingan imaging technique based on the EPI pulsesequence; gadolinium contrast is used to indicatethe amount of blood perfusion across a brain cellmembrane

peripheral nerve stimulation (PNS)activation of a peripheral nerve fiber(s) caused byrapidly switching gradient fields; PNS is not apatient safety concern but a potential patient com-fort concern

phaseparticular stage or point of advancement in a cycle

phased array coila type of surface coil composed of several coils andreceivers that are linked together. The signals fromeach of the coils and receivers are subsequentlyunited to form an image with good SNR.

phase encodinggeneration of phase differences along a particulardirection of a tissue slice for use in spatial localiza-tion of MR signal

pixelpicture element; the smallest discrete part of a digital image display

pulse sequencea series of events for exciting protons and detectingsignals during the MR examination; every pulsesequence includes slice excitation, echo generation,and phase and frequency encoding

quenchsudden and massive expansion of liquid helium intogaseous helium due to an increase of heat from aloss of superconductivity of the magnet

R-R intervalthe period of time between each R-wave; one cardiac cycle

R-wave1

the initial upward deflection of the QRS complex,following the Q-wave in the normal electrocardio-gram

radiofrequency (RF)1the range of frequencies of electromagnetic radia-tion between 10 kilohertz and 100 gigahertz that isused for radio communication

ramp timethe minimum time required for a gradient field togo from its peak maximum point to its peak mini-mum point; measured in microseconds

ramp up/ ramp downthe controlled process of bringing the magnet tomaximum field or reducing the magnet from itsmaximum field

receive bandwidth (RBW)the range of frequencies collected during the frequency encoding portion of the pulse sequence

region of interest (ROI)a specific defined area, ie, fluid or a portion of anorgan or tumor, where a relative signal intensitymeasurement can be obtained

RF shieldingmetal used to prevent stray RF frequencies fromentering the magnet room during an MR exam,typically made of copper

rise timethe minimum time required for a gradient field togo from zero to its required maximum; measured in microseconds

NO

T FO

R D

ISTR

IBU

TIO


resonancestate of a system through which energy may betransferred to another system with the same preferred or resonant frequency; characterized by absorption and dissipation of energy throughresonant oscillation

saturationlack of signal due to overexcitation of protons

signal-to-noise ratio (SNR)amount of true signal relative to the amount ofrandom background signal (noise) on an image

slew ratedescribes overall gradient performance as a func-tion of gradient amplitude and gradient rise time.Slew rate is derived by dividing the amplitude bythe rise time and typically is described in units of T/m/sec.

specific absorption rate (SAR)the FDA-regulated amount of RF heat energy that a patient can absorb during an MRI exam,measured in Watts/kilogram

spin-echobasic pulse sequence of MR imaging using a 900 RFexcitation pulse and a 1800 RF echo pulse

static magnetic fieldthe large main magnet field generated by the magnet to place the protons into the longitudinalplane prior to the MRI pulse sequence; also knownas the B0 field

short tau inversion recovery (STIR)inversion recovery-based pulse sequence that utilizes a relatively short TI time in order to obtainheavy fat suppression; the TI is selected based onT1 recovery time of fat

superconductivitythe state where all molecular motion becomes sorestricted, due to a lack of heat, that electricity can flow through a conductor without resistance.Materials such as a magnet coil become supercon-ductive after being immersed in liquid helium andreaching a temperature of 40 Kelvin

surface coilspecialized antenna for transmitting and/or receiv-ing RF energy to or from the patient during a pulsesequence

T1the amount of time for 63% of a tissue’s protonsto recover to longitudinal magnetization

T1-weightinggeneration of MR images under conditions thathighlight differences in T1 between tissues

T2the amount of time for 63% of a tissue’s protonsto decay in the transverse plane

T2-weightinggeneration of MR images under conditions thathighlight differences in T2 between tissues

Tesla1

the SI unit of magnetic flux density, calculated aswebers per square meter. It replaces the gauss.

time to echo (TE)the amount of time selected by the technologist to allow for T2 decay of excited protons; also theamount of time from the beginning of initial sliceexcitation pulse and the generated echo signal

time-of-flight imaging (TOF)gradient echo-based pulse sequence that uses flow-related enhancement to greatly increase thecontrast between flowing blood and stationary tissue; performed in either a 2D or 3D acquisition, the images are postprocessed into a maximumintensity pixel projection (MIP) to display like anMR angiogram

time to inversion (TI)the time allowed from the initial 180 degree excita-tion pulse until the 90 degree pulse in an inversionrecovery pulse sequence; TI determines the amountof T1 recovery time for a given tissue

torque1

a rotatory force causing a part of a structure totwist about an axis

time to recovery (TR)the amount of time selected by the technologist toallow for T1 recovery of the excited protons; alsothe time from the beginning of one pulse sequenceto the beginning of the next

trigger windowthe time delay before each R-wave

vectormathematical quantity representing both magni-tude and direction, symbolized by an arrow

voxela three-dimensional volume of tissue correspondingto a pixel on an MR image, a "volume element"

mri for technologists mri systems and coil … mri systems and coil...mri for technologists . mri...

Documents