How MRI Shimming Keeps the Magnetic Field Uniform

How an MRI scanner corrects small imperfections in its main magnetic field, why different patients disturb that field differently, and how poor shimming can become distortion, signal loss, fat-suppression failure, or bad spectroscopy

An MRI magnet can be several tesla strong, but simply producing a powerful magnetic field is not enough.

Back to Biomed Basics

What This Page Explains

This page covers:

The Simple Version

The main MRI magnet creates the static field known as B0, but real magnetic fields are never perfectly uniform. The magnet itself, nearby ferromagnetic structures, room construction, scanner components, and even the patient can slightly disturb the field. Because hydrogen resonance frequency depends directly on local magnetic-field strength, these small field differences become frequency differences across the anatomy.

Shimming corrects those unwanted variations. Passive shimming uses strategically placed magnetic material to improve the scanner's baseline field homogeneity, usually as part of installation or major service. Active shimming uses dedicated shim coils that generate small corrective magnetic fields. During patient imaging, the scanner may measure the field over the anatomy and calculate shim currents that make the effective B0 field more uniform in the selected region.

A scan can therefore have a field-homogeneity problem even when the magnet is at the correct nominal field strength. Poor shimming may appear as failed fat suppression, geometric distortion, signal loss, spectral line broadening, or artifacts that become worse in certain anatomies or positions. Preserve whether the problem affects every patient or only certain exams, whether center frequency or shim values changed, whether it follows one room condition or one sequence, and whether a phantom reproduces the symptom. Those clues help separate a scanner-wide field problem from normal patient susceptibility or a local imaging challenge.

Start With B0

The main static magnetic field in MRI is commonly called:

B0.

A 1.5 T MRI system aims to produce a field near:

1.5 tesla

within its imaging region.

A 3 T system operates near:

3 tesla.

Nominal Field Strength Is Only Part of the Story

Two questions matter:

  1. Is the average field near the correct strength?
  2. Is that field uniform throughout the region we need to image?

A scanner can satisfy the first and struggle with the second.

Field Homogeneity

Homogeneity describes how consistent the magnetic-field strength is across a defined volume.

An ideally uniform field would have exactly the same magnetic-field strength everywhere.

Real systems approximate that extremely closely.

MRI Requires Extreme Uniformity

Small field variations can matter because resonance frequency depends directly on field strength.

Larmor Frequency

For hydrogen:

Higher local magnetic field means:

Higher resonance frequency.

Lower local field means:

Lower resonance frequency.

At 1.5 T

Hydrogen resonates near:

64 MHz.

At 3 T

It is near:

128 MHz.

Exact operating frequency is determined by the scanner.

Tiny Field Difference, Detectable Frequency Difference

MRI is sensitive enough that a very small change in magnetic field can shift resonance frequency noticeably.

That is what makes shimming important.

What Causes Field Inhomogeneity?

Many things.

Some are built into the physical magnet.

Others come from the environment.

Others come from the patient.

The Magnet Is Not Mathematically Perfect

Even an extremely well-built superconducting magnet has small imperfections.

Manufacturing tolerances and magnet geometry contribute.

The Cryostat and Scanner Structure

Materials around the magnet can influence the field.

The MRI Room

Nearby ferromagnetic material can distort B0.

Examples could include:

Moving Metal Can Change the Field

A large ferromagnetic object moved near the magnet may affect field homogeneity.

The significance depends on:

This Can Create a Strange Complaint

MRI quality changes after:

The scanner electronics may not have changed at all.

The Patient Changes the Magnetic Field

This is especially important.

Human tissue has different magnetic properties.

When the patient is placed inside B0:

The patient slightly changes the local field.

Magnetic Susceptibility

The degree to which a material becomes magnetized in an external field is related to:

Magnetic susceptibility.

Different materials have different susceptibility.

Tissue Interfaces

Field variation becomes especially noticeable near boundaries such as:

Common Challenging Areas

Examples include:

These areas contain abrupt susceptibility changes.

Metal

Metal can create much larger local disturbances.

Implants can cause:

depending on material, orientation, sequence, and field strength.

Shimming Cannot Magically Remove Every Distortion

There are limits.

A large local disturbance caused by metal may be impossible to completely correct with global shim fields.

The Goal of Shimming

Shimming adds carefully shaped magnetic fields that oppose the unwanted variations.

Think of it as:

Field correction.

If one region is slightly too high:

The shim system tries to reduce the effective field there.

If another is too low:

It tries to increase it.

Two Broad Types of Shimming

You will often hear:

Passive Shimming

Passive shimming uses small pieces of magnetic material placed at selected locations around the magnet.

Why Add Metal on Purpose?

Because those pieces predictably alter the field.

By choosing:

engineers can correct baseline field imperfections.

Installation Process

During MRI installation, specialized measurements can map the magnetic field.

The resulting data helps determine where passive shim material should be placed.

Passive Shim Is Mostly Static

Once installed:

It stays there.

It establishes part of the scanner's baseline field homogeneity.

Passive Shim Can Be Disturbed

If shim pieces are:

the magnetic field can change.

This is specialized work and not normal BMET adjustment.

Active Shimming

Active shimming uses electrical coils.

Current through those coils creates small magnetic fields.

Shim Coils

Shim coils are designed to produce particular field shapes.

The scanner changes current through them to cancel measured field errors.

Similar Principle to Gradient Coils?

They both use current to create magnetic fields.

But their purpose is different.

Gradient Coils

Create controlled spatial field changes for:

Image encoding.

Shim Coils

Create corrective field shapes for:

B0 uniformity.

Shim Fields Are Relatively Small

They are corrections applied on top of the much stronger main magnetic field.

First-Order Shims

Some shim terms correspond to simple spatial field variations.

First-order corrections resemble linear field changes along:

directions.

Higher-Order Shims

More complicated field distortions require more complex field shapes.

Higher-order shim coils can correct some of these.

You Do Not Need to Memorize Every Mathematical Shim Term

For routine biomed understanding, the important concept is:

Different shim channels correct different spatial patterns.

Shim Current

The scanner applies specific currents to each shim channel.

These currents create the corrective field.

Shim Amplifier

Active shim coils therefore need controlled current sources.

Depending on scanner design, faults can occur in:

Baseline Shim

The magnet may have a set of baseline shim conditions established during installation and service.

Patient-Specific Shim

The scanner may also calculate adjustments for the patient or selected imaging volume.

Why Patient-Specific?

Because every patient disturbs the magnetic field differently.

Larger Patient

Different body geometry.

Different Anatomy

Head and abdomen produce different susceptibility patterns.

Implants

Can create strong local field changes.

The Scanner Needs to Adapt

A shim that works beautifully for a spherical phantom may not be optimal for:

A human abdomen.

Field Mapping

Before some sequences, the MRI measures how B0 varies over the selected region.

This is:

Field mapping.

What Is Being Measured?

The scanner looks at resonance-frequency differences across space.

Those differences reveal magnetic-field variation.

Shim Calculation

Software determines which combination of shim currents should reduce those variations.

Current Is Applied

The active shim system creates the corrective fields.

Then the Sequence Runs

The imaging sequence operates on top of the improved magnetic environment.

Automatic Shimming

Modern MRI often handles much of this automatically.

The technologist may select:

and the system performs the calculation.

Manual Shim

Specialized applications or service procedures may allow more manual control.

This requires appropriate knowledge.

Do Not Randomly Change Shim Values

You can easily worsen image quality or disturb established calibration.

Center Frequency

Before scanning, the MRI may also determine the hydrogen center frequency for the patient.

Why Center Frequency Changes

The patient slightly changes the magnetic environment.

Therefore the exact resonant frequency may shift.

Center Frequency and Shim Are Related but Not Identical

Center frequency addresses:

The overall resonance location.

Shimming addresses:

How uniform that resonance is across the region.

Good Average, Poor Uniformity

Imagine the center of the frequency distribution is exactly correct.

But one side is high and the other side low.

Average frequency can look fine.

Homogeneity can still be poor.

Spectral Linewidth

One way to think about field homogeneity is through the width of the resonance spectrum.

Uniform Field

Hydrogen throughout the sample resonates at very similar frequencies.

The spectral peak is relatively narrow.

Inhomogeneous Field

Different regions resonate at different frequencies.

The peak becomes broader.

Linewidth

The width of that peak can be used as a measure of field homogeneity.

Narrower Is Generally Better

Within the proper test and specification.

This Matters Greatly in Spectroscopy

MR spectroscopy depends heavily on distinguishing small chemical frequency differences.

Poor field homogeneity can blur those spectral peaks together.

Spectroscopy Can Be a Shim Stress Test

A scanner may make acceptable routine anatomical images but struggle with spectroscopy because the field needs to be especially homogeneous.

Fat Suppression

Field homogeneity also matters for frequency-selective fat suppression.

Why?

Fat and water resonate at slightly different frequencies.

A frequency-selective pulse is designed to affect fat without strongly affecting water.

If B0 Varies Too Much

Fat in different parts of the image may not resonate where the scanner expects.

Result

Fat suppression may become:

Classic Symptom

One side suppresses well.

Another side does not.

Does That Prove Bad Shimming?

No.

Possible factors include:

But shimming belongs high on the conceptual list.

Frequency-Selective Techniques Depend on B0

Any technique that depends on precise frequency separation becomes more sensitive to field homogeneity.

Echo-Planar Imaging

EPI can be particularly sensitive to B0 inhomogeneity.

Why?

EPI collects data rapidly using long readout trains.

Frequency offsets can accumulate into spatial distortion.

Result

Areas near:

may appear warped.

Diffusion Imaging

Diffusion commonly uses EPI.

That is why susceptibility-related distortion can be prominent in diffusion images.

This Is Not Automatically a Gradient Failure

The gradients may be performing perfectly.

The issue can arise because local resonance frequency differs from the assumed value.

Geometric Distortion

Poor B0 homogeneity can cause positions to be encoded incorrectly in frequency-sensitive directions.

Signal Loss

Field variations can also cause spins within one voxel to lose phase coherence.

Dephasing

If different spins in one voxel precess at different frequencies:

Their signals can cancel one another.

Result

Local signal loss.

Gradient-Echo Sequences

Gradient-echo imaging can be especially sensitive to magnetic susceptibility and B0 inhomogeneity because it lacks the same 180-degree refocusing behavior used in spin-echo sequences.

Sequence Dependence Is a Clue

If an artifact is:

Severe on gradient echo

but mild on spin echo,

susceptibility and field effects become more interesting.

Spin Echo

A 180-degree refocusing pulse can correct some static dephasing effects.

It cannot correct every type of field-related distortion.

Shimming Region

Automatic shim performance depends partly on what region the scanner is told to optimize.

Too Large a Region

The system may have difficulty optimizing a field over anatomy containing very different susceptibility regions.

Too Small a Region

The shim may work well locally but less well elsewhere.

Positioning Matters

If the anatomy of interest is far from isocenter:

Field uniformity may be worse.

Magnet Is Designed Around a Particular Homogeneous Volume

The best field uniformity is generally near:

Isocenter.

Off-Center Imaging

Can experience greater field variation.

That does not necessarily indicate failure.

Bore Geometry Matters

The homogeneous field specification applies over defined volumes.

Do not expect identical performance everywhere inside the bore.

Scanner Installation

MRI site planning considers nearby magnetic materials for a reason.

Large Ferromagnetic Structure

Can distort the magnet's field.

Moving Structure

Examples might include:

depending on site layout.

Temporal Field Disturbance

If a magnetic disturbance moves:

The field may change over time.

Intermittent MRI Artifact

This creates a fascinating possibility.

Artifact appears only when:

A nearby elevator passes.

That would not be fixed by replacing an RF coil.

Magnetic Shielding

Some MRI installations use passive magnetic shielding.

Others use actively shielded magnet designs.

This helps control the magnetic fringe field.

Magnetic Shielding and RF Shielding Are Different

Important distinction.

Magnetic shielding

manages static magnetic-field behavior.

RF shielding

keeps external radiofrequency signals from contaminating receive data.

Do Not Confuse Shim With RF Noise

Poor shimming changes:

Resonance frequency distribution.

RF interference adds:

Unwanted external signal.

They can create very different artifacts.

Magnet Drift

Superconducting magnets can exhibit slow changes in field over time.

Modern systems account for expected behavior.

Frequency Tracking

The scanner can adjust operating frequency to account for small field changes.

Sudden Change Is Different

A sudden major change in center frequency or shim requirement may indicate something significant changed.

Ask What Changed

Magnet Quench

A quench is a completely different and much more serious event involving loss of superconductivity and rapid helium boil-off.

Shimming is not a treatment for a quench.

Post-Service Shim

Certain major magnet or cryogenic work may require specialized field evaluation afterward.

This is typically OEM or specialist work.

Shim Calibration

The scanner needs to know how each shim current changes the field.

Calibration establishes these relationships.

Shim-Coil Fault

If an active shim channel does not produce the expected field:

Automatic shimming may fail or produce poor results.

Possible Symptoms

Amplifier Fault

If the commanded shim current is not produced:

The corrective field will not match what the software expected.

Coil Open or Connection Problem

Again, the result depends on architecture.

This is specialized service territory.

Shim Values Can Be Evidence

Suppose the same phantom historically needs modest shim correction.

Suddenly one shim channel is driven near its maximum.

Something changed.

But Do Not Diagnose From the Number Alone

The patient or phantom setup matters.

Use:

Phantom Testing

A known homogeneous phantom is extremely useful for evaluating field behavior.

Why?

It removes many unpredictable patient susceptibility effects.

If Phantom Fails Too

Scanner or environment becomes more interesting.

If Phantom Passes but One Patient Fails

Patient anatomy, implant, positioning, or selected shim region may be more likely.

This Is a Powerful Split

Phantom problem

vs:

Patient-specific problem.

Center Frequency Trend

If center frequency changes unexpectedly over time:

That may indicate field or calibration changes.

Shim Current Trend

If shim values gradually move:

That can also provide evidence.

Do Not Overinterpret Normal Variation

Different patients naturally require different shim settings.

Real-World Example: Failed Fat Suppression

Technologists report uneven fat suppression in abdominal MRI.

Head imaging remains excellent.

System phantom passes baseline homogeneity tests.

Issue is much worse in large patients.

This points more toward:

Patient-specific B0/B1 conditions

than an immediately obvious scanner-wide hardware failure.

Real-World Example: Phantom Linewidth Worsens

Same spectroscopy phantom historically produces a narrow water peak.

Now linewidth is substantially broader.

Automatic shim cannot restore prior performance.

No patient is involved.

A scanner or environmental field issue now deserves deeper investigation.

Real-World Example: Problem After Construction

Image distortion begins shortly after nearby structural work.

Phantom reproduces the problem.

Center frequency and shim requirements have shifted from baseline.

A change in the magnetic environment becomes an important clue.

Real-World Example: Metal Implant

Patient with metallic implant shows severe local distortion and signal loss.

Phantom and other patients image normally.

The scanner is not necessarily malfunctioning.

Local magnetic susceptibility is disrupting B0 around the implant.

Real-World Example: Active Shim Fault

Automatic shim repeatedly fails.

One shim channel reports current-regulation error.

Phantom shows reproducible field nonuniformity.

Now the issue is much more specific than:

MRI images look distorted.

Common Mistakes

Thinking a strong magnet is automatically a uniform magnet. Field strength and field homogeneity are different properties.

Confusing gradients with shims. Gradients intentionally encode spatial location. Shims correct unwanted B0 variation.

Assuming every poor fat-suppression exam is a hardware failure. Patient anatomy, air/tissue interfaces, implants, B1 variation, and positioning can contribute.

Randomly adjusting shim settings. Baseline magnet shimming is specialized work.

Ignoring environmental changes. Large nearby ferromagnetic structures can influence field homogeneity.

Ignoring phantom results. A phantom helps separate scanner-wide problems from patient-specific susceptibility.

Confusing RF shielding and magnetic shimming. One addresses unwanted RF energy; the other addresses static-field uniformity.

A Useful MRI Shimming Framework

Think:

Main Magnet Produces B0

↓

Magnet + Environment + Patient Disturb B0

↓

Field Mapping Measures Variation

↓

Shim Algorithm Calculates Correction

↓

Active Shim Coils Generate Corrective Fields

↓

B0 Becomes More Uniform

↓

Center Frequency / Sequence Operate on Corrected Field

↓

MRI Signal Is Acquired

That gives you a complete conceptual shim loop.

Another Useful Troubleshooting Split

Ask:

Is the inhomogeneity scanner-wide or patient-specific?

Then ask:

Does a phantom reproduce it?

Then:

Did center frequency, linewidth, or shim demand change from baseline?

That is far more useful than simply:

Fat sat looks bad.

What Did You Actually Prove?

If the scanner's nominal field is:

3 T,

you proved:

The system is designed to operate around that field strength.

You did not prove:

The field is perfectly homogeneous throughout the imaging volume.

If automatic shimming completes:

You proved:

The system completed its shim procedure according to its internal criteria.

You did not necessarily prove:

If a qualified homogeneity or spectroscopy phantom test passes:

You have stronger evidence that the magnetic field meets the tested uniformity requirements within that defined volume and condition.

If one patient's fat suppression is poor while the phantom and other patients are normal:

You have not proven a scanner failure.

You have evidence that the failure depends on the imaging condition.

Final Thoughts for Biomeds

An MRI magnet does not only need to be strong.

It needs to be predictably strong in the right way throughout the region being imaged.

A few tiny differences in magnetic-field strength can become:

Shimming is how the scanner cleans up that magnetic landscape.

Some of that correction is built into the installation.

Some is produced electrically.

Some is recalculated for every patient.

That is why:

Magnet is at field

is not the end of the conversation.

A healthy main magnet can still be operating in a field that is too nonuniform for a particular imaging task.

And a terrible-looking image does not automatically mean the magnet needs service.

Sometimes the source is:

The useful skill is learning to separate those possibilities.

Does the problem occur on a phantom?

Does it affect every sequence?

Does it affect every patient?

Did the room or surrounding environment change?

Are shim values or center frequency moving away from historical behavior?

Those questions turn:

MRI looks weird

into something much more technical.

And once again, the most important question is not whether the automatic shim screen says:

Complete.

It is:

What did you actually prove?

— Jake

Important Note

MRI field-homogeneity specifications, passive-shim methods, active-shim coil architecture, shim orders, field-mapping procedures, center-frequency calibration, linewidth limits, phantom tests, and service boundaries vary significantly by manufacturer, field strength, magnet design, and application. MRI systems present serious magnetic, RF, electrical, and cryogenic hazards. Follow current OEM documentation, MRI safety procedures, facility policy, and authorized service scope when evaluating magnetic-field or shimming problems.

Related Biomed Basics

Back to Biomed Basics