How an MRI scanner sees inside you without radiation

An X-ray works by firing radiation through you and recording what gets absorbed. A CT scan does the same thing many times over from different angles. An MRI scanner does nothing of the sort. It fires no ionising radiation at all. Instead it places you in a powerful magnetic field, nudges the hydrogen nuclei in your own body with a pulse of radio waves, and listens to the faint signal they give off as they settle back down. Everything about the resulting picture — its extraordinary soft-tissue detail, its cost, its noise, its slowness — follows from that single choice.

Your body is mostly hydrogen, and hydrogen is magnetic

You are largely water and fat, which means you are full of hydrogen. A hydrogen nucleus is a single proton, and a proton behaves like a tiny spinning magnet. Ordinarily these point in random directions and cancel out.

Put someone inside a strong magnetic field — typically 1.5 or 3 tesla in a clinical scanner, tens of thousands of times the Earth's field — and a small majority of those protons line up with it. That slight alignment is all the scanner has to work with, and it is why the magnets must be so strong.

Now send in a radio-frequency pulse tuned to exactly the right frequency, and the aligned protons absorb the energy and tip out of alignment. This is resonance, and the frequency depends precisely on the magnetic field strength — a relationship called the Larmor equation, which turns out to be the key to the whole technique.

Relaxation is where the picture comes from

Switch the pulse off and the protons relax back toward alignment, emitting a weak radio signal as they do. Receiver coils pick that signal up. Two independent things are happening during relaxation, and they happen at different rates in different tissues:

• T1 relaxation is how quickly the protons realign with the main magnetic field.
• T2 relaxation is how quickly they lose synchronisation with one another.

Fat, muscle, grey matter, white matter, cerebrospinal fluid and tumour tissue all have their own characteristic T1 and T2 times, because relaxation depends on the local molecular environment each proton sits in. By timing the pulses and the measurement differently — the pulse sequence — a radiographer can make the image emphasise T1 differences, T2 differences, or something else entirely. The same patient, unmoved, yields images that look completely different and answer different clinical questions.

This is why MRI is unmatched for soft tissue. An X-ray essentially measures density, so brain, muscle and ligament look much alike. MRI measures the molecular environment, which is exactly where those tissues differ.

How the machine knows where the signal came from

A signal from the whole body at once would be useless. The trick is that resonant frequency depends on field strength, so the scanner deliberately makes the field slightly uneven using gradient coils — electromagnets that add a controlled ramp across the body.

With a gradient applied, protons at one end of the body sit in a slightly stronger field than those at the other, and so resonate at slightly different frequencies. Position is therefore encoded into frequency and phase. Gradients along all three axes let the scanner select a slice and map position within it, and a Fourier transform converts the collected frequency data into an image.

Those gradient coils also explain the noise. They are switched on and off rapidly, hundreds of times a second, inside an intense magnetic field — and the resulting forces make the coils physically knock and vibrate. The banging is not a fault; it is the sound of the spatial encoding being performed.

An MRI does not photograph you. It asks every hydrogen nucleus in your body a question, and reconstructs the picture from how fast each one answers.

Strengths, limits and safety

MRI's advantages are soft-tissue contrast, the absence of ionising radiation, and the ability to image in any plane. It is the default for brain, spinal cord, joints, ligaments and many tumours, and specialised sequences do more still — diffusion-weighted imaging detects stroke within minutes, functional MRI tracks blood-oxygen changes as a proxy for brain activity, and MR angiography images vessels without contrast dye.

Its limits are practical. Scans take minutes rather than seconds, so motion blurs them and CT remains the better choice in trauma. Machines are expensive to buy and to run, since the superconducting magnet must be kept near absolute zero with liquid helium. Bone and lung image poorly, having little mobile hydrogen.

The safety picture is unusual: there is no radiation risk, but the magnet is never switched off, so ferromagnetic objects become projectiles. Metallic implants, pacemakers and shrapnel must be screened for, and gadolinium contrast agents need care in patients with poor kidney function.

Why it matters for students and researchers

MRI is a rare case of pure physics — nuclear magnetic resonance, developed to study molecular structure — becoming an everyday clinical instrument, and it rewards study from either direction. Current research includes faster acquisition through compressed sensing and machine-learning reconstruction, ultra-high-field 7 tesla imaging, low-field portable scanners aimed squarely at settings where a conventional machine is unaffordable, and quantitative MRI that reports tissue properties as numbers rather than relative brightness. That last strand matters for India, where access rather than capability is the binding constraint. Following the peer-reviewed literature is how physics, biomedical engineering and medical students and professionals keep pace with a field where imaging advances arrive through algorithms as often as through hardware.

Frequently asked questions

How does an MRI scanner work?

A strong magnetic field aligns hydrogen nuclei in the body. A radio-frequency pulse tips them out of alignment, and as they relax back they emit a weak radio signal that receiver coils detect. Gradient coils vary the field across the body so that signal position can be decoded, and the data is reconstructed into an image.

Does an MRI use radiation?

No. MRI uses magnetic fields and radio waves, not ionising radiation, so there is no radiation dose and no associated cancer risk. This is a major reason it is preferred for repeated scans and for imaging children, though safety screening for metal implants remains essential.

Why is an MRI scan so noisy?

The loud knocking comes from the gradient coils, which are switched on and off rapidly inside the strong magnetic field. The changing currents produce forces that make the coils vibrate against their mountings. The noise is an unavoidable by-product of encoding spatial position.

What is the difference between MRI and CT?

CT uses X-rays and is fast, widely available and excellent for bone, bleeding and trauma. MRI uses magnetic fields and radio waves, takes longer and costs more, but gives far better soft-tissue contrast for brain, spinal cord, ligaments and many tumours, with no ionising radiation.