Why STIR suppresses fat

Tonight · ~15 min · read · energy: low · setup: none

STIR (Short TI Inversion Recovery) is the cleanest demonstration of a powerful MRI idea: by choosing a timing, you can deliberately null a tissue’s signal. This short lesson explains inversion recovery and the inversion time TI, conceptually rather than with derivation.

The anchor: fat is bright, and that hides pathology

On many MRI weightings, fat is bright. That is useful for anatomy but a problem when you want to see pathology in or near fatty tissue (bone marrow oedema, soft-tissue oedema): the bright fat swamps the signal you care about. Fat suppression makes fat dark so pathology becomes visible. STIR is one — conceptually elegant — way to do it.

Inversion recovery, in one idea

An inversion recovery sequence adds a 180° inversion pulse before the usual excitation. That pulse flips the longitudinal magnetisation to the opposite direction; it then recovers (T1 recovery) and passes through zero on its way back to equilibrium. Different tissues cross zero at different times (because they have different T1). If you excite and read out exactly when a chosen tissue’s magnetisation is crossing zero, that tissue produces no signal — it is nulled.

flowchart LR
    I["180° inversion pulse<br/>magnetisation flipped"] --> R["recovers through zero<br/>(T1 recovery)"]
    R -->|"excite at tissue's<br/>zero-crossing (TI)"| N["that tissue nulled"]
    N --> O["other tissues still visible"]

The timing between the inversion pulse and the excitation is the inversion time (TI).

Why STIR nulls fat

STIR (Short TI Inversion Recovery) uses a short TI chosen so that fat’s magnetisation is crossing zero at excitation. Fat is nulled; fluid and most pathology (oedema, which has long T1/T2) still produce signal and appear bright against the suppressed fat background. The result is a fat-suppressed, oedema-sensitive image — which is why STIR is a workhorse in musculoskeletal and spinal imaging for showing bone marrow and soft-tissue oedema.

The transferable idea is bigger than STIR: TI is a dial that nulls a chosen tissue. A long TI nulls fluid instead (that is FLAIR — Fluid-Attenuated Inversion Recovery — which suppresses bright CSF so periventricular pathology is visible). Same mechanism, different tissue, different TI.

A note on “fat suppression” more broadly

STIR is one of several fat-suppression techniques; others (spectral fat sat, DIXON) exploit the frequency difference between fat and water rather than the T1 zero-crossing. The choice matters for image quality (STIR is robust at low field and with inhomogeneity, but mixes T1/T2 contrast; spectral sat is cleaner but field-sensitive) — but for the mental model, STIR is the example that makes “null a tissue by timing” click.

Stop and think — then reveal

You want an image where CSF is dark but brain oedema stays bright. Which timing dial do you reach for, and roughly how?

You want a FLAIR sequence — an inversion recovery with a long TI chosen to null fluid (CSF). CSF’s magnetisation crosses zero at excitation, so CSF goes dark; oedema (with different T1/T2) still produces signal and stays bright. It is the same mechanism as STIR (null a tissue by choosing TI at its zero-crossing), just aimed at fluid instead of fat. One dial (TI), two clinically crucial sequences (STIR nulls fat, FLAIR nulls fluid).

What to retain

Inversion recovery adds a 180° pulse before excitation; magnetisation recovers through zero, and exciting at a tissue’s zero-crossing nulls that tissue.
TI is the dial: short TI nulls fat (STIR); long TI nulls fluid (FLAIR).
STIR is fat-suppressed and oedema-sensitive — a musculoskeletal/spinal workhorse; robust to field inhomogeneity but mixes T1/T2 contrast.
Other fat-suppression methods (spectral sat, DIXON) use frequency, not T1 — STIR is the example that makes “null a tissue by timing” click.

Next: the most quantitative thing routine MRI produces — what diffusion and ADC actually measure.