Magnetic resonance is a physical phenomenon in which atomic nuclei absorb and release energy when placed in a magnetic field and hit with a pulse of radio waves at exactly the right frequency. It’s the principle behind MRI scanners, chemical analysis techniques, and functional brain imaging. The “resonance” part is key: just as a tuning fork vibrates only at its natural pitch, atomic nuclei respond only to a very specific frequency of energy, determined by the strength of the surrounding magnetic field.
How Atoms Behave in a Magnetic Field
Every hydrogen nucleus (a single proton) acts like a tiny spinning magnet. Under normal conditions these miniature magnets point in random directions, so their effects cancel out. But when you place them inside a strong external magnetic field, they snap into alignment. Most align with the field, which is the lower-energy position. A smaller number align against it, which requires slightly more energy. The difference between those two groups creates a measurable net magnetization, a collective magnetic signal pointing in the direction of the external field.
This net magnetization is what makes magnetic resonance possible. It’s tiny, which is why the external magnets need to be extraordinarily powerful. Clinical MRI machines typically operate at 1.5 or 3 Tesla, thousands of times stronger than the Earth’s magnetic field. Research systems approved for clinical use since 2017 run at 7 Tesla, and experimental magnets reach 10.5, 11.7, and even 14 Tesla.
The Resonance Frequency
Once the nuclei are aligned, they don’t just sit still. Each proton wobbles (precesses) around the axis of the magnetic field at a very precise rate called the Larmor frequency. For hydrogen, that frequency equals 42.58 megahertz for every Tesla of field strength. So in a 1.5 Tesla scanner, hydrogen nuclei wobble about 63.87 million times per second, firmly in the radio-wave range of the electromagnetic spectrum.
This relationship is the heart of resonance. If you broadcast a radio-frequency (RF) pulse at exactly the Larmor frequency, the protons absorb that energy. It’s the same principle that lets you shatter a wine glass with a note at its natural frequency: energy transfers efficiently only when the frequencies match. Any other frequency and nothing happens.
What the RF Pulse Does
The RF pulse is delivered by coils positioned around whatever is being scanned. It does two things simultaneously. First, it tips the net magnetization away from its resting position along the main magnetic field and into the perpendicular (transverse) plane. A pulse calibrated to tip the magnetization exactly 90 degrees is the workhorse of most imaging sequences. Second, it pulls the individual protons into synchrony so they all precess together, in phase, like a line of marching soldiers swinging their arms in unison.
This synchronized, rotating magnetization generates a small but detectable electrical signal in receiver coils, the same way a spinning magnet inside a generator produces current. That signal is the raw material of every magnetic resonance measurement.
Relaxation: How the Signal Fades
The moment the RF pulse switches off, the system begins returning to its original state. This recovery happens through two independent processes that occur at different speeds depending on the tissue or material being examined.
The first, called T1 relaxation, describes how quickly the tipped magnetization regrows along the main field direction. Different tissues recover at different rates because of differences in molecular motion and environment. Fat recovers quickly; fluid-filled spaces recover slowly.
The second, T2 relaxation, describes how quickly the synchronized protons fall out of step with each other. Small, random fluctuations in the local magnetic field cause each proton to precess at a slightly different speed, and the collective signal fans out and weakens. This isn’t a loss of energy so much as a loss of order. Dense tissues with tightly packed molecules cause faster dephasing than watery environments where molecules tumble freely.
Because T1 and T2 times vary between tissue types, they provide natural contrast. By adjusting when you listen for the returning signal relative to the RF pulse, you can highlight different structures. This is why an MRI of the brain can distinguish gray matter from white matter, or why a tumor appears brighter than surrounding tissue, without any radiation or injected dye.
Building an Image With Gradient Fields
A uniform magnetic field would make every hydrogen atom in your body resonate at the same frequency, producing one undifferentiated signal. To create an image, MRI scanners add smaller, variable magnetic fields called gradients on top of the main field. These gradients make the field strength slightly different at every point in space, which means the resonance frequency also differs from point to point.
By switching gradients on and off in carefully timed sequences along three axes, the scanner encodes spatial information into the signal’s frequency and phase. Sophisticated mathematical processing (primarily Fourier transforms) then reconstructs a detailed three-dimensional map of where each signal originated. The result is cross-sectional images with sub-millimeter resolution, all derived from the basic physics of magnetic resonance.
Magnetic Resonance Beyond Medical Imaging
The phenomenon was first harnessed not for medicine but for chemistry. Felix Bloch and Edward Mills Purcell independently demonstrated nuclear magnetic resonance (NMR) in 1946 and shared the 1952 Nobel Prize in Physics for the work. Their discovery became the foundation of NMR spectroscopy, a technique chemists use daily to determine molecular structures.
In NMR spectroscopy, the exact resonance frequency of a hydrogen atom shifts slightly depending on its chemical neighbors. Electrons surrounding a nucleus generate a small secondary magnetic field that partially shields the nucleus from the external field, effectively changing the frequency needed to achieve resonance. Atoms bonded to elements that pull electrons away (like oxygen or fluorine) lose that shielding and resonate at a different frequency than atoms in electron-rich environments. These shifts create a fingerprint-like spectrum that reveals how atoms are connected within a molecule. Drug development, forensic analysis, and quality control in food and petroleum industries all rely on this technique.
Functional Brain Imaging
A specialized application called functional MRI (fMRI) extends magnetic resonance into neuroscience. It works by detecting changes in blood oxygenation. When a brain region becomes active, blood flow to that area increases, delivering more oxygen than the tissue actually consumes. Oxygenated and deoxygenated hemoglobin have different magnetic properties: deoxygenated hemoglobin is weakly magnetic and disrupts the local field, while oxygenated hemoglobin does not. As active regions receive a surplus of oxygenated blood, the local signal increases. This blood-oxygen-level-dependent (BOLD) signal can be mapped to produce images of which brain areas are active during a task, a thought, or a sensation.
Safety Considerations
Magnetic resonance imaging uses no ionizing radiation, which is one of its major advantages over CT scans and X-rays. The primary safety concern is the magnetic field itself. At clinical field strengths, ferromagnetic objects, anything containing iron or certain other metals, can be pulled violently toward the scanner bore. Documented projectile incidents have involved items as varied as oxygen tanks, wheelchairs, IV poles, scissors, stethoscopes, and even wooden chairs with hidden metal strips in the cushion.
Implanted medical devices pose a separate concern. Devices that are magnetically, electrically, or mechanically activated may malfunction in the MR environment. Cardiac pacemakers, certain cochlear implants, and some types of aneurysm clips carry specific restrictions. Modern implants are increasingly designed to be MR-compatible, but compatibility depends on the specific field strength and scanning conditions. An implant tested as safe at 1.5 Tesla is not necessarily safe at 3 Tesla.
MRI facilities enforce strict screening protocols and designate controlled zones around the scanner. The boundary where the fringe field drops below 5 gauss (0.0005 Tesla) is considered the safety line for the general public and for people with cardiac devices. Everything closer requires screening.