The Basics of MRI

Chapter 9


Hardware Overview

MRI scanners have evolved considerably since the first commercial units were introduced in the 1980s. This chapter covers the basic hardware components on current imagers as well as presents some details of the first generation scanners.

The graphics window displays a schematic representation of the major systems on a magnetic resonance imager and a few of the major interconnections. This overview briefly states the function of each component. Some will be described in detail later in this chapter.

At the top of the schematic representation you will find the components of the imager located in the scan room of a magnetic resonance imager. The magnet produces the Bo field for the imaging procedure. Within the magnet are the gradient coils for producing a gradient in Bo in the X, Y, and Z directions. Within the gradient coils is the RF coil. The RF coil produces the B1 magnetic field necessary to rotate the spins by 90°, 180°, or any other value selected by the pulse sequence. The RF coil also detects the signal from the spins within the body. The patient is positioned within the magnet by a computer controlled patient table. The table has a positioning accuracy of 1 mm. The scan room is surrounded by an RF shield. The shield prevents the high power RF pulses from radiating out through the hospital. It also prevents the various RF signals from television and radio stations from being detected by the imager. Some scan rooms are also surrounded by a magnetic shield which contains the magnetic field from extending too far into the hospital. In newer magnets, the magnet shield is an integral part of the magnet.

The heart of the imager is the computer. It controls all components on the imager. The RF components under control of the computer are the radio frequency source and pulse programmer. The source produces a sine wave of the desired frequency. The Pulse programmer shapes the RF pulses into apodized sinc pulses. The RF amplifier increases the pulses power from milli Watts to killo Watts. The computer also controls the gradient pulse programmer which sets the shape and amplitude of each of the three gradient fields. The gradient amplifier increases the power of the gradient pulses to a level sufficient to drive the gradient coils.

The array processor, located on some imagers, is a device which is capable of performing a two-dimensional Fourier transform in fractions of a second. The computer off loads the Fourier transform to this faster device.

The operator of the imager gives input to the computer through a control console. An imaging sequence is selected and customized from the console. The operator can see the images on a video display located on the console or can make hard copies of the images on a film printer.

The next three sections of this chapter go into more detail on the magnet, gradient coils, RF coils, and RF detector on magnetic resonance imagers.


The imaging magnet is the most expensive component of the magnetic resonance imaging system. Most magnets are of the superconducting type. This is a picture of a first generation 1.5 Tesla superconducting magnet from a magnetic resonance imager. A superconducting magnet is an electromagnet made of superconducting wire. Superconducting wire has a resistance approximately equal to zero when it is cooled to a temperature close to absolute zero (-273.15° C or 0 K) by immersing it in liquid helium. Once current is caused to flow in the coil it will continue to flow as long as the coil is kept at liquid helium temperatures. (Some losses do occur over time due to infinitely small resistance of the coil. These losses can be on the order of a ppm of the main magnetic field per year.)

The length of superconducting wire in the magnet is typically several miles. The coil of wire is kept at a temperature of 4.2K by immersing it in liquid helium. The coil and liquid helium is kept in a large dewar. The typical volume of liquid Helium in an MRI magnet is 1700 liters. In early magnet designs, this dewar was typically surrounded by a liquid nitrogen (77.4K) dewar which acts as a thermal buffer between the room temperature (293K) and the liquid helium. See the animation window for a cross sectional view of a first generation superconducting imaging magnet.

In later magnet designs, the liquid nitrogen region was replaced by a dewar cooled by a cryocooler or refrigerator. There is a refrigerator outside the magnet with cooling lines going to a coldhead in the liquid helium. This design eliminates the need to add liquid nitrogen to the magnet, and increases the liquid helium hold time to 3 to 4 years. The animation window contains a cross sectional view of this type of magnet. Researchers are working on a magnet that requires no liquid helium.

Another advance in magnet technology is the shielded magnet. This magnet has a smaller fringe field. The fringe field drops to 0.5 mT by four meters from the magnet. This is important for safety reasons (discussed later in this chapter) and makes it easier to site a magnet. The shielding is achieved by a second set of superconducting windings, outside of the main ones and with opposite current, which reduce the fringe field.

Gradient Coils

The gradient coils produce the gradients in the Bo magnetic field. They are room temperature coils, which because of their configuration, create the desired gradient. Since the horizontal bore superconducting magnet is most common, the gradient coil system will be described for this magnet.

Assuming the standard magnetic resonance coordinate system, a gradient in Bo in the Z direction is achieved with an antihelmholtz type of coil. Current in the two coils flow in opposite directions creating a magnetic field gradient between the two coils. The B field at one coil adds to the Bo field while the B field at the center of the other coil subtracts from the Bo field.

The X and Y gradients in the Bo field are created by a pair of figure-8 coils. The X axis figure-8 coils create a gradient in Bo in the X direction due to the direction of the current through the coils. The Y axis figure-8 coils provides a similar gradient in Bo along the Y axis.

Gradient coil technology has also evolved considerably since the introduction of first generation systems. Early systems had maximum gradient strengths 10 mT/m and rather slow switching times. Current generation systems can have maximum gradient strengths of 100 mT/m and much faster switching times (slew rates) of 150 mT/m/ms. These values allow the system to achieve a 0.7 mm slice thickness for 2D acquisitions and 0.1 in 3D.

RF Coils

RF coils create the B1 field which rotates the net magnetization in a pulse sequence. They also detect the transverse magnetization as it precesses in the XY plane. RF coils can be divided into three general categories; 1) transmit and receive coils, 2) receive only coils, and 3) transmit only coils. Transmit and receive coils serve as the transmitter of the B1 fields and receiver of RF energy from the imaged object. A transmit only coil is used to create the B1 field and a receive only coil is used in conjunction with it to detect or receive the signal from the spins in the imaged object. There are several varieties of each. The RF coil on an imager can be likened unto the lens on a camera. A photographer will use one lens for a close up shot and a different one for a wide angle long distance shot. Just as a good photographer has several lenses, a good imaging site will have several imaging coils to handle the variety of imaging situations which might arise.

An imaging coil must resonate, or efficiently store energy, at the Larmor frequency. All imaging coils are composed of an inductor, or inductive elements, and a set of capacitive elements. The resonant frequency, ν, of an RF coil is determined by the inductance (L) and capacitance (C) of the inductor capacitor circuit.

Some types of imaging coils need to be tuned for each patient by physically varying a variable capacitor. The other requirement of an imaging coil is that the B1 field must be perpendicular to the Bo magnetic field.

There are many types of imaging coils. Volume coils surround the imaged object while surface coils are placed adjacent to the imaged object. An internal coil is one designed to record information from regions outside of the coil, such as a catheter coil designed to be inserted into a blood vessel. Some coils can operate as both the transmitter of the B1 field and the receiver of the RF signal. Other coils are designed as only the receiver of the RF signal. When a receive only coil is used, a larger coil on the imager is used as the transmitter of RF energy to producing the 90° and 180° pulses. The following table is a partial list of the more common imaging coils with the type classifier, the mode of operation (transmit/receive-T/R or receive only-R), a diagram, and a literature reference for the coil. The diagrams show the direction of the B1 field.

RF Coils for Magnetic Resonance
Coil NameTypeModeDiagramReference
Alderman-Grant CoilVolumeT/R
Bird CageVolume T/R
Butterfly CoilSurfaceR
Dome ResonatorSurfaceT/R See Ref.
ImplantableInternalT/R See Ref.
Inside Out
(Schlumberger Coil)
Internal/SurfaceT/R See Ref.
Intravascular CoilInternalR See Ref.
Loop-Gap Resonator Volume R See Ref.
Loop-StickSurfaceT/R See Ref.
Mouse CoilSurfaceR
Multi-Turn Solenoid Volume T/R
Phased ArraySurfaceR ,
Phased Array VolumeVolumeR
Saddle CoilVolumeT/R
Scroll Coil Volume T/R See Ref.
Single Turn Solenoid Volume T/R
Surface CoilSurfaceR
Spiral CoilSurfaceR
Superconducting CoilSurfaceR See Ref.
Transmission Line (TEM)
(Slotted Tube)
Truncated SpiralSurfaceR
3-Axis Volume T/R

Surface coils are very popular because they are a receive only coil and have a good signal-to-noise ratio for tissues adjacent to the coil. In general, the sensitivity of a surface coil drops off as the distance from the coil increases. Here is an example of an image of the lower human spine obtained with a surface coil.

Here is a picture of a flat circular surface coil with its connecting cable. The cable will connect the coil to the imager. This is a picture of a surface coil molded to conform to the back of the knee.

The bird cage coil is the most routinely used volume coil. It is the coil of choice for imaging the head and brain. Here is a picture of the human head in a bird cage coil. All of the head images in this hypertext book were obtained using a bird cage coil.

The single turn solenoid is useful for imaging extremities, such as the breasts and the wrist. This animation window entry shows a single turn solenoid imaging coil around the human wrist. The detail icon will provide you with more information on the construction of a single turn solenoid.

RF Detector

RF detectors on MRI systems have evolved considerably since the 1980s. Initially, linear analog detectors and single channel digitizers were used. These were replaced with quadrature analog detectors with two channel digitizers. With the more recent availability of fast digitizers, single channel digitizers followed by digital quadrature detection is more common.

Linear analog detectors produce only the Mx' or the My' magnetization as the signal as a function of time (S(t)). The signal is then digitized. Linear detection requires that the digitization rate for the signal be at least two times the highest frequency in the signal. The factor of two is because half of the signal must be discarded to allow discrimination between positive and negative frequencies in the signal. The reader is encouraged to review the chapter on Fourier Transforms to prove this statement.

Quadrature analog detectors separates out the Mx' and My' signals coming from the RF coil. For this reason it can be thought of as a laboratory to rotating frame of reference converter. Both the Mx' and My' signals are digitized producing a complex signal as a function of time. For this reason, the digitization rate need only be equal to the largest frequency in the signal. Again, the reader is encouraged to review the chapter on Fourier Transforms to prove this statement.

The heart of an analog linear or quadrature detector is a device called a doubly balanced mixer (DBM). The doubly balanced mixer has two inputs and one output. If the input signals are Cos(A) and Cos(B), the output will be 1/2 Cos(A+B) and 1/2 Cos(A-B). For this reason the device is often called a product detector since the product of Cos(A) and Cos(B) is the output.

The analog linear detector consists of one DBM, a filter, and an amplifier. The reference frequency is νo, the isocenter resonance frequency. Frequencies ν and νo are put in and the MX or MY component of the transverse magnetization comes out.

The quadrature detector typically contains two doubly balanced mixers, two filters, two amplifiers, and a 90° phase shifter. There are two inputs and two outputs on the device. Frequencies ν and νo are put in and the MX and MY components of the transverse magnetization come out. There are some potential problems which can occur with this device which will cause artifacts in the image. These will be addressed in the chapter on Image Artifacts.

It is more common to find the following detection scheme on state-of-the-art detectors. The RF signals at frequency ν from the RF coil are mixed with (νo + νo') to produce an intermediate frequency νi using a DBM. This frequency is digitized or oversampled using a high speed digitizer. Once digitized, the rotating frame signals (Mx'(t) and My'(t)) are created using a digital quadrature detector and a digital filter. This is done entirely in software using the equations for a product detector

Cos(2π νi t) Cos(2π νo' t) = 1/2 Cos(2π νi t - 2π νo' t) + 1/2 Cos(2π νi' t + 2π νo' t)

Sin(2π νi t) Cos(2π νo' t) = 1/2 Sin(2π νi t + 2π νo' t) + 1/2 Sin(2π νi' t - 2π νo' t)

and a time domain digital filter shaped like a sinc function in the time domain. The use of digital quadrature detection removes the possibility of a quadrature ghost artifact, to be discussed in the Artifact Chapter.

The intermediate frequency and digitization is typically 1 MHz, resulting in oversampling and the generation of too much data to be conveniently stored. Digital filtering eliminates the high frequency components from the data, and decimation reduces the size of the data set. The following flowchart summarizes the effects of the three steps by showing the result of performing an FT after each step.

A closer examination of oversampling, digital filtering, and decimation can reveal how a combination of steps can be used to reduce the wraparound problem.

Oversampling is the digitization of a time domain signal at a frequency much greater than necessary to record the desired field of view. For example, if the sampling frequency, fs, is increased by a factor of 10, the field of view will be 10 times greater, thus eliminating wrap around. Unfortunately digitizing at 10 times the speed also increases the amount of raw data by a factor of 10, thus increasing storage requirements and processing time.

Digital Filtering
Filtering is the removal of a select band of frequencies from a signal. For an example of filtering, consider the following frequency domain signal. Frequencies above fo could be removed from this frequency domain signal by multiplying the signal by this rectangular function. In MRI, this step would be equivalent to taking a large FOV image and setting to zero intensity those pixels greater than some distance from the isocenter.

Digital filtering is the removal of these frequencies using the time domain signal. Recall from the chapter on Fourier Transforms that if two functions are multiplied in one domain (i.e. frequency), we must convolve the FT of the two functions together in the other domain (i.e. time). To filter out frequencies above fo from the time domain signal it must be convolved with the Fourier transform of the rectangular function, a sinc function. (See the chapter on Fourier Transforms.) This process eliminates frequencies greater than fo from the time domain signal. Fourier transforming the resultant time domain signal yields a frequency domain signal without the higher frequencies. In MRI, this step will remove image components fo / 2 γ Gf away from the center of the image.

Decimation is the elimination of data points from a data set. A decimation ratio of 4/5 means that 4 out of every 5 data points are deleted, or every fifth data point is saved. Decimating the digitally filtered data above, followed by a Fourier transform, will reduce the data set by a factor of five.

High speed digitizers, capable of digitizing at 2 MHz, and dedicated high speed integrated circuits, capable of performing the convolution on the time domain data as it is being recorded, are used to realize this procedure.


I am often asked, how safe is MRI? As with every piece of technology, there are risks and benefits. Those pieces of technology in wide use generally have a high benefit to risk ratio, while those with a low benefit to risk ratio are generally used more sparingly. Although MRI does not use ionizing radiation to produce images there are still some important safety considerations which one should be familiar with. These concern the use of strong magnetic fields, radio frequency energy, time varying magnetic fields, cryogenic liquids, and magnetic field gradients.

In 1982 the US FDA set guidelines for MRI exams that covered the maximum Bo field, change in magnetic field with respect to time (dB/dt), the absorption of radio frequency energy, and acoustic noise levels. These are sumarized in the animation window table. In 1997, the US FDA revised these guidelines due to the accumulated data on MRI. The limits were revised again in 2003 and are summarized in the following table.

FDA MRI Guidelines (2003)
BoAdults, Children, and Infants age > 1 month 8 T
neonates (infants age < 1 month)4 T
dB/dtNo discomfort, pain, or nerve stimulation
SARwhole body, average, over >15 min4 W/Kg
head, average, over >10 min 3 W/Kg
head or torso, per g of tissue, in >5 min 8 W/Kg
extremities, per g of tissue, in >5 min12 W/Kg
Peak unweighted140 dB
A-weighted rms with hearing protection99 dBA

MRI personnel often forget about the dangers associated with ferromagnetic objects near the MRI magnet. Magnetic fields from large bore magnets can literally pick up and pull large ferromagnetic items into the bore of the magnet. Caution must be taken to keep ALL ferromagnetic items away from the magnet for two main reasons. The first reason is they can injure or kill an individual in the magnet. The second reason is they can seriously damage the magnet and imaging coils. The force exerted on a large metal object, such as a mop wringer can damage the concentric cryogenic dewars within a magnet. The kinetic energy of such an object being sucked into a magnet can smash an RF imaging coil.

Despite numerous safety warnings issued by the manufacturers, professional societies, and the government, I have heard numerous stories of ferromagnetic objects being pulled into imaging magnets. The most common one is similar to this story. A metal pail on wheels was filled with water and had a mop wringer in it. The pail was located approximately 10 feet from the bore of a 1.5 T magnet. The magnet pulled it across the floor and lifted it up off the ground three feet into the magnet. The wringer caused serious damage to the magnet in that the cryogen boil off rate increased and the magnetic field homogeneity decreased. The head coil located in the bore of the magnet was destroyed.

The most tragic story was the death of a six year old boy in an MRI magnet. This story should serve as an example and reminder of the responsibility that MRI personnel and administrators have to maintain a safe facility. In July 2001 a boy was injured in an MRI and later died when a ferromagnetic oxygen tank was brought into the magnet room and pulled into the magnet where the boy was being imaged.

Another frightening story was of a law enforcement officer being allowed to go near a magnet with a loaded firearm. The handgun was pulled out of its holster, and into the magnet. The force of the impact with the magnet caused the gun to discharge. Luckily, no one was injured in this incident. In addition to the damage to the MRI and the bullet lodged in the scan room wall, the gun was magnetized. Mechanical objects, in general, do not function properly when magnetized. Please, respect the physical laws of nature that cause ferromagnetic objects to be attracted by magnets!

Similar forces are at work on ferromagnetic metal implants or foreign matter in those being imaged. These forces can pull on these objects cutting and compressing healthy tissue. For these reasons individuals with foreign metal objects such as shrapnel or older ferromagnetic implants are not imaged. There are additional concerns regarding the effect of magnetic fields on electronic circuitry, specifically pacemakers. An individual with a pacemaker walking through a strong magnetic field can induce currents in the pacemaker circuitry which will cause it to fail and possibly cause death. Magnetic fields will also erase credit cards and magnetic storage media.

The United States Food and Drug Administration (USFDA) safety guidelines state that field strengths not exceeding 2.0 Tesla may be routinely used. People with pacemakers must not be exposed to magnetic fields greater than 5 gauss. A 50 Gauss magnetic field will erase magnetic storage media.

The radio frequency energy from an imaging sequence can cause heating of the tissues of the body. The USFDA recommends that the exposure to RF energy be limited. The specific absorption rate (SAR) is the limiting measure.

SAR = Joules of RF / Second / kg of body weight = Watts/kg

The recommended SAR limitations depend on the anatomy being imaged. The SAR for the whole body must be less than 4 W/kg. It must be less than 3.2 W/kg averaged over the head. Any pulse sequence must not rise the temperature by more than 1° Celsius and no greater than 38° C in the head, 39° C in the trunk, and 40° C in the extremities.

Body tattoos can be a contraindication for an MRI scan. Tattoo inks can contain paramagnetic or ferro/ferrimagnetic pigments. These pigments can absorb electromagnetic energy associated with the scan, resulting in localized heating. There have been reports of slight tingling or first-degree burns associated with tattoos on people having an MRI scan. These symptoms are not felt by all persons with tattoos indicating that it may be pigment or physiology dependent. Often times persons with body tattoos that will be in an imaging coil will be given an ice pack to place on the tattoo during a scan. Some cosmetic tattoos can cause image artifacts.

Some RF coils, such as surface coils, have failure modes which can cause burns to the patient. The animation window contains a picture of an RF burn to the elbow of a man's arm. The patient's arm was against the wall of a body coil being operated in a transmit mode with a surface coil as the receiver. A malfunction in the body coil caused the third degree RF burn. The burn first appeared as a simple blister and progressed to a charring that had to be surgically removed. The surgeon excised a volume approximately 3 cm in diameter and 2.5 cm deep. Therefore, if you are operating an imager and your patient or volunteer tells you he or she is experiencing a burning sensation, stop the scan. Additionally, care should be taken to keep RF imaging coils in proper operating order.

The USFDA recommendations for the rate of change of magnetic field state that the dB/dt for the system must be less than that required to produce peripheral nerve stimulation.

Imaging gradients do produce high acoustic noise levels. The American OSHA limits the peak acoustic noise to 200 pascals or 140 dB references to 20 micropascals. Here are some examples of the sounds made by the turning on and off of the magnetic field gradients in various imaging sequences.

Sequence TR (ms)TE (ms)SlicesSound
Spin-Echo 500 35 1
200 1
15 10
Echo-Planar 120 54 10
Gradient Echo 16.7 4 19

Every MRI magnet room with a superconducting magnet should have an oxygen monitor. These devices measure the percentage of O2 in the air and sound an alarm when the level falls below a set threshold. These devices are needed because leaks in the venting system that handle the boil off of cryogens could create a situation where excess N2 or He in the room air would deplete the percentage of O2 to a dangerous level.


An MRI phantom is an anthropogenic object that can be imaged to test the performance of the magnetic resonance imaging system. Phantoms are used instead of a standard human because it is much easier to locate a phantom standard at each of the many MRI systems in the world then it is to send the standard human from site to site to be imaged. Phantoms are composed of materials that have a magnetic resonance signal. Many materials have been used as the signal bearing substance in MRI phantoms. Some of these are aqueous paramagnetic solutions; pure gels of gelatin, agar, polyvinyl alcohol, silicone, polyacrylamide, or agarose; organic dopped gels; paramagnetically dopped gels; and reverse micelle solutions.

Water is most frequently used as the signal bearing substance in an MRI phantom. It is usually necessary to adjust the spin-lattice (T1) and spin-spin (T2) relaxation times of aqueous solutions so images may be acquired in reasonable time periods (i.e. short TR). Paramagnetic metal ions are typically used to adjust the relaxation times of the water hydrogens. The approximate functional form of the T1 and T2 values of aqueous solutions of various paramagnetic species at 1.5 T are listed below.

Aqueous Nickel
T1(s) = 1/(632 [Ni (mole/L)] +0.337)
T2(s) = 1/(691 [Ni (mole/L)] + 1.133)
Nickel in 10 wt % gelatin
T1(s) = 1/(732 [Ni (mole/L)] +0.817)
T2(s) = 1/(892 [Ni (mole/L)] + 4.635)
Aqueous Oxygen
T1(s) = 1/(0.013465 [O2 (mg/L)] + 0.232357)
Aqueous Manganese
T1(s) = 1/(5722 [Mn (mole/L)] +0.0846)
T2(s) = 1/(60386 [Mn (mole/L)] + 3.644)
Aqueous Copper
T1(s) = 1/(606 [Cu (mole/L)] +0.349)
T2(s) = 1/(850 [Cu (mole/L)] + 0.0357)

There are four basic types of MRI phantoms: resolution, linearity, homogeneity, and signal. The latter is used to affirm the signal or some measurable property from the signal that results from a pulse sequence. Examples of this type are a T1 phantom and a diffusion coefficient phantom. Homogeneity phantoms can be used to measure both the RF and Bo homogeneity. Resolution, linearity, and homogeneity can be used to measure the Bo homogeneity, but they allow the measurement of homogeneity to be made differently then in the homogeneity phantom. The following paragraphs describe the differences in more detail.

Resolution and Linearity Phantoms
A resolution and linearity phantom can be used to test several spatial properties of an imager. These spatial properties include in-plane resolution, slice thickness, linearity, and the signal-to-noise ratio as a function of position. Resolution phantoms are typically constructed from plastic. Portions of the inside of the phantom are removed to create a test pattern. The phantom is filled with an aqueous solution. When imaged, the image displays the signal from the water in the removed portions of the plastic. Some resolution phantoms also have signal standards with known T1, T2, and ρ values that allow the phantom to be used to test contrast-to-noise ratios.

Here is an example of a resolution phantom. A 24 cm field-of-view image of an axial slice through this phantom displays the following features. The series of identical size squares are used to test the linearity. In-plane resolution is determined by a group of thin signal-bearing regions. Three signal standards contain a liquid with a known T1, T2, and ρ values. The slice thickness (Thk) metric is a wedge shaped cut-away in the plastic. The width of the image of this wedge increases as the slice thickness is increases. The following schematic diagrams of the phantom imaged with a thin , and thick slice thickness will help you see how this shape an help measure slice thickness. Here are images of the resolution phantom imaged with a 3 , 5 , and 10 mm slice thickness. Note the change in the slice width metric.

A problem with the previous design resolution phantom is that it does not allow determination of resolution throughout the imageable volume without being repositioned. The following resolution phantom allows determination of both linearity throughout a 10 by 10 by 10 cm volume, and resolution at multiple points throughout this volume. It consists of a three sets of parallel tubes, each set being orthogonal to the others. Depending on the slice thickness, any imaged plane gives either a matrix of dots or dots and lines. Linearity across the field of view can be determined from the straightness of the lines or linearity of the points. Resolution can be determined from the size of the dots or width of the lines.

Homogeneity Phantoms
Homogeneity phantoms are used to test the spatial uniformity of the static magnetic field, as well as the transmit and receive radio frequency magnetic fields. Let us first address their use for monitoring the homogeneity of the Bo magnetic field.

The NMR spectral linewidth (Γ) of a single spin packet equals

Γ = (π T2)-1.

because there by definition a uniform applied magnetic field. As the volume of the signal bearing material increases, the linewidth becomes

Γ = (π T2*)-1.

because the magnetic field varies from location to location. The smaller this variation, the smaller the distortion in the image. A high resolution NMR spectrometer can have a linewidth of 0.5 Hz for sample in a 5 mm NMR tube. In a clinical imager, the variation in Bo across a 27 cm diameter spherical phantom filled with water causes the NMR linewidth from the phantom to be 30 to 40 Hz. This is because there is a broad distribution of resonance frequencies for all the spin packets in the sample. The width of absorption line for a large volume phantom is therefore a measure of the distribution of magnetic field values across the volume of the phantom.

Homogeneity phantoms are also used to test the spatial uniformity of the transmit and receive radio frequency magnetic fields. The transmit RF field (B1T) is the B1 field is that used to rotate magnetization. The receive RF field (B1R) is the sensitivity of the RF coil to signals from precessing spin packets. The ideal situation for most transmit/receive coils is a spatially uniform B1T to assure uniform rotation of the spins, and a spatially uniform B1R to assure uniform sensitivity across the imaged object. Here is a picture of a 27 cm diameter homogeneity phantom. A series of spheres can be used to measure the homogeneity over a larger volume. Here is an array of homogeneity phantoms which can be used to measure the homogeneity of B1R field from a surface coil used for imaging the spine.

The challenge in making a homogeneity phantom is to fill it with a low dielectric constant material to minimize the standing wave artifact, and moderate resistivity so as to load the RF coil like the body would but not create a conductivity artifact. The standing wave artifact is seen when the wavelength of the RF in the phantom is approximately twice the diameter of the phantom. RF reflects off the phantom-air boundary causing a standing wave. The conductivity artifact is seen when the conductivity of the phantom is high and causes the RF to concentrate on the surface of the phantom.

Several images from an RF homogeneity phantom must be used to calculate B1T and B1R. Please click on the detail icon to receive more information on these calculations.

Diffusion Phantoms
Diffusion phantoms are phantoms which can be used to test the performance of a diffusion imaging sequence. Some consist of vials of liquids with different diffusion coefficients.

Since the solution is isotropic, the phantom will yield the same diffusion coefficient in x, y, and z. Geometrical diffusion phantoms rely on special geometries to create anisotropic diffusion. For example, a set of parallel tubes will have unrestricted diffusion along the axis of the tube, and restricted diffusion along a diameter of the tube. Parallel plates will have unrestricted diffusion parallel to the plates, and restricted diffusion perpendicular to the plates. One diffusion phantom contains hydrated parallel fibers.

Fat-Saturation Phantoms
Fat saturation phantoms consist of liquids with two chemical shifts, one for water and one for fat. These are usually oil-water emulsions. Since body fat contains multiple peaks, better fat saturation phantoms contain oils with multiple peaks similar to body fat.

T1 and T2 Phantoms
T1 and T2 phantoms contains liquids with specific T1 and T2 values. These liquids are usually aqueous solutions of paramagnetic materials.

fMRI Phantoms
Functional MRI (fMRI) is the use of a fast imaging sequence, such as an echo-planar imaging sequence, to measure very small, rapid signal changes associated with oxygenated and deoxygenated blood in brain tissues. These differences are associated with changes in activity within regions of the brain. The basis of fMRI will be presented in Chapter 13. Assessing the proper performance of an fMRI procedure requires a phantom which can change its signal as rapidly as the blood oxygen level dependent (BOLD) signal. One fMRI phantom utilizes a liquid filled cell through which a voltage is applied to cause small ion currents in the liquid.   The ion currents create small magnetic fields which distort Bo, thus changing the magnetic resonance signal from the liquid. This plot displays the change in phantom signal in response to the application of an electric field (red line) to the phantom for a series of images repeated every 205 ms. Each data point represents the signal intensity for a different image.


  1. A particular imaging coil has a resonance frequency of 63.85 MHz. You observe that it has 50 picofarad capacitor across the inductor. What is the size of the inductor in units of Henrys?

  2. The two anti-Helmholtz coils of a Z gradient coil are separated by 1 meter. If each coil produces a 50 Gauss magnetic field at its center, what is the magnitude of the gradient?

  3. Will the following pulse sequence adhere to the USFDA SAR guidelines for the head?
    Sequence = multi-slice, spin echo
    TR = 500 ms
    TE = 25 ms
    Thk = 5 mm
    Number of slices = 10
    Energy per 90° pulse = 5 J
    Width of 90° pulse = 50 µs
    Signal acquisition time = 10 ms
    Patient weight = 72 kg

  4. The quadrature detector on your imager has a reference signal of 63.85 MHz coming into one of the channels. The signal channel has a signal of 63.86 MHz coming into it. What frequency or frequencies will be coming out of the detector?

  5. What is the resonance frequency of an imaging coil which has a 1x10-7 Henry inductor and a 100 picofarad capacitor across the inductor? How can this coil be made to resonate at 63.85 MHz?

  6. The two anit-Helmholtz coils of a Z gradient coil are separated by 120 cm. If this coil configuration can produce a magnetic field gradient of 2 G/cm, what is the magneitude of the magnetic field at the center of each coil?

  7. Will the following pulse sequence adhere to the USFDA SAR guidelines for the head?
    Sequence = multi-slice, spin echo
    TR = 600 ms
    TE = 25 ms
    Thk = 3 mm
    Number of slices = 17
    Energy per 90° pulse = 10 J
    Width of 90° pulse = 50 µs
    Signal acquisition time = 10 ms
    Patient weight = 80 kg

  8. The quadrature detector on your imager has a reference signal of 63.85 MHz coming into one of the channels. The signal channel has a signal of 63.84 MHz coming into it. What frequency or frequencies will be coming out of the detector?

  9. Safety Questions:
    Is it safe for someone with a tattoo to have an MRI? Why?
    Is it safe for a pregnant woman to have an MRI? Why?
    How many MRI scans can a person have? Why?

Go to the: [next chapter | chapter top | previous chapter | cover ]

Copyright © 1996-2020 J.P. Hornak.
All Rights Reserved.