Relaxation Phenomena

Version 1.0 — July 2026

1. Executive Summary

This page covers, in depth, the second of the eleven planned topic groups listed in Section 4.2 of the MRI Physics — Fundamentals and Principles master page: the two fundamental relaxation processes — T1 (spin-lattice) and T2 (spin-spin) — together with the practically dominant T2* phenomenon and the tissue- and field-strength-dependence of all three. Relaxation is, without exaggeration, the single physical phenomenon most directly responsible for MRI’s unique soft-tissue contrast: essentially every sequence, weighting, and contrast-agent strategy documented elsewhere on MRIninja is, at root, a strategy for exploiting differences in T1 and/or T2 between tissues.

2. Relationship to the Master Page

The master page introduces relaxation only briefly, at survey level (Section 3.3), and the companion child page Fundamentals of Nuclear Magnetic Resonance introduces T1/T2/T2* only to the minimum extent needed to explain the decay of the free induction decay (FID) signal, explicitly deferring full treatment to this page. This page assumes familiarity with nuclear spin, precession, the Larmor equation, net magnetization, and the basic existence of the FID (covered in that companion page) and does not repeat that material. What this page adds: the physical mechanisms that actually cause T1 and T2 relaxation, the mathematical form of the recovery/decay curves, a full decomposition of T2* into its true-T2 and field-inhomogeneity components, and a detailed, referenced treatment of how and why relaxation times vary between tissues and across field strengths.

3. T1 Relaxation (Spin-Lattice)

3.1 Physical Mechanism

T1 relaxation describes the recovery of longitudinal magnetization (Mz) toward its equilibrium value after being disturbed by an RF pulse. Physically, this recovery requires the excited spin system to give up the energy it absorbed from the RF pulse back to its molecular surroundings — historically termed the “lattice,” a term retained from the technique’s origins in solid-state physics even though, in clinical MRI, the relevant surroundings are the tumbling molecules of a liquid or semi-solid biological tissue rather than a literal crystal lattice. This energy exchange is only efficient when the natural molecular motion in the tissue (rotational and translational tumbling, driven by thermal energy) produces fluctuating local magnetic fields with a significant component oscillating near the Larmor frequency itself — the physical requirement first formalised in the classic 1948 theoretical treatment by Bloembergen, Purcell, and Pound (universally referred to as “BPP theory”), which related T1 quantitatively to the rotational correlation time (τc) of the molecules involved [1].

3.2 The T1 Recovery Curve

Following a 90° excitation pulse (which places all longitudinal magnetization into the transverse plane), Mz recovers toward its equilibrium value M0 according to an exponential curve:

Mz(t) = M0 · (1 − e^(−t/T1))

T1 is, by definition, the time constant at which approximately 63% of the equilibrium longitudinal magnetization has been recovered. This exponential recovery curve is the direct physical basis for TR (repetition time) selection in every pulse sequence documented on MRIninja: a short TR relative to a tissue’s T1 leaves that tissue’s magnetization only partially recovered before the next excitation, producing lower signal and — critically, when different tissues have different T1 values — producing T1-weighted contrast between them.

3.3 What Determines T1 — the BPP Correlation-Time Model

The BPP model predicts that T1 depends on how well the frequency spectrum of a molecule’s random tumbling motion overlaps with the Larmor frequency. Molecules that tumble very rapidly (as in low-viscosity fluids such as cerebrospinal fluid) or, at the opposite extreme, molecules that are essentially immobile (as in solid bone or dense fibrous tissue) both produce relatively inefficient energy exchange at the Larmor frequency, and therefore relatively long T1 values. Molecules with intermediate tumbling rates — a correlation time on the order of the inverse Larmor frequency — exchange energy most efficiently and produce the shortest T1 values. This is the physical explanation for why fat (with intermediate-mobility lipid molecules) has a characteristically short T1, while pure water-like fluids (very fast tumbling) and cortical bone (essentially no tumbling) both have comparatively long T1 values, despite being physically very different tissues.

4. T2 Relaxation (Spin-Spin)

4.1 Physical Mechanism

T2 relaxation describes the loss of coherent transverse magnetization (Mxy) — the gradual randomisation of the phase relationship between individual precessing nuclear moments that were initially excited in phase. Unlike T1 relaxation, T2 relaxation does not require any net loss of energy from the spin system to its surroundings; it is fundamentally an entropy-increasing dephasing process, driven by direct, local magnetic interactions between neighbouring nuclear spins (spin-spin coupling) and by the same molecular-motion-driven field fluctuations responsible for T1 relaxation, considered at zero and low frequency components rather than only near the Larmor frequency. Because T2 relaxation can occur through mechanisms that do not require energy exchange with the lattice, it is always at least as fast as — and in practice virtually always faster than — T1 relaxation for any given tissue.

4.2 The T2 Decay Curve

Following excitation, the transverse magnetization decays toward zero according to:

Mxy(t) = Mxy(0) · e^(−t/T2)

T2 is, by definition, the time constant at which the transverse magnetization has decayed to approximately 37% of its immediate post-excitation value. This decay curve is the direct physical basis for TE (echo time) selection: a longer TE allows more T2-driven decay to occur before signal is measured, producing lower signal and — when tissues differ in T2 — producing T2-weighted contrast.

4.3 The T1–T2 Relationship

Because T2 relaxation processes are, physically, a superset of the mechanisms that also drive T1 relaxation (both arise from the same underlying molecular field fluctuations, just sampled at different frequency components), a fundamental physical constraint applies universally: T2 ≤ T1, always, for any tissue at any field strength. This is not an empirical coincidence but a direct mathematical consequence of BPP-type relaxation theory, and it explains why, across essentially every tissue in the body, measured T2 values are substantially shorter than T1 values measured in the same tissue (see Section 6.3 for representative figures).

5. T2* and the Role of Field Inhomogeneity

5.1 Decomposing T2* into Its Components

The transverse-magnetization decay actually observed in any real MRI experiment — the envelope of the raw FID signal, introduced in the companion Fundamentals of Nuclear Magnetic Resonance child page — decays faster than the true molecular T2 predicts. This observed, faster decay is characterised by T2*, and it combines two physically distinct contributions, related by:

1/T2* = 1/T2 + 1/T2′

Here, T2 is the true, irreversible, molecular spin-spin relaxation process described in Section 4, and T2′ (sometimes written 1/T2′ = γ·ΔB0, where ΔB0 is the local field inhomogeneity) represents an additional, in-principle-reversible dephasing contribution caused by spins at different spatial positions experiencing slightly different local magnetic fields, and therefore precessing at slightly different frequencies.

5.2 Sources of Field Inhomogeneity

Two broad categories of field inhomogeneity contribute to T2′, and therefore to T2*:

• Macroscopic magnet imperfection: no real magnet produces a perfectly uniform B0 field across the entire imaging volume; residual inhomogeneity remaining after shimming (discussed in the planned Hardware, Field Homogeneity, and Magnet Physics child page group, master page Section 4.7) contributes to T2′.
• Local tissue magnetic susceptibility variation: different tissues (and, especially, interfaces between tissue and air, tissue and bone, or tissue and metal/haemorrhage products) locally distort the magnetic field due to differences in magnetic susceptibility — the physical basis of susceptibility artefact (discussed in the planned Physics of MRI Artefacts child page group, master page Section 4.8) and of susceptibility-based contrast mechanisms such as T2*-weighted and susceptibility-weighted imaging.

5.3 Recovering True T2 — the Spin Echo

Because the field-inhomogeneity contribution to T2* decay (T2′) arises from static, non-random spatial variations in field strength, it is — unlike true T2 dephasing — reversible. Erwin Hahn’s 1950 description of the spin echo, already introduced in the companion Fundamentals of Nuclear Magnetic Resonance child page in the context of the FID, is the direct practical solution: a 180° refocusing pulse applied partway through the dephasing period exactly reverses each spin’s accumulated phase error due to static field inhomogeneity, causing the dephased spins to rephase (forming an “echo”) at a predictable time, with the resulting echo amplitude governed by true T2 decay alone rather than the faster T2* decay [2]. This is the physical reason spin-echo-family sequences are considered to provide “true” T2-weighted contrast, while gradient-echo sequences — which have no refocusing pulse — are inherently T2*-weighted rather than T2-weighted.

6. Tissue- and Field-Strength-Dependence of T1/T2

6.1 Why T1 Increases with Field Strength

BPP theory (Section 3.3) predicts, and extensive experimental measurement confirms, that T1 relaxation times increase with increasing static field strength for essentially all biological tissues. Stanisz and colleagues’ widely cited 3T relaxometry study found that T1 values for every tissue measured increased significantly when moving from 1.5T to 3T, while T2 values remained comparatively similar between the two field strengths [3]. This field dependence is a direct, quantitative consequence of the Larmor-frequency-dependence built into BPP theory: as field strength (and therefore Larmor frequency) increases, the molecular tumbling motions responsible for T1 relaxation become progressively less well matched to the (now higher) Larmor frequency for most tissue types, reducing relaxation efficiency and lengthening T1.

6.2 Why T2 Is Comparatively Field-Independent

T2 relaxation, because it is driven substantially by low-frequency and static field-fluctuation components (Section 4.1) rather than specifically by components at the Larmor frequency, shows markedly less field-strength dependence than T1 — the practical reason that T2-weighted image contrast between tissues changes comparatively little when moving between clinical field strengths, even though absolute signal-to-noise and many other sequence parameters do change substantially.

6.3 Representative Tissue Relaxation Values

The published literature shows considerable variability in reported absolute relaxation times for the same nominal tissue — a comprehensive literature review at 3T found reported white matter T1 values ranging from 699 to 1735 ms and fat T2 values ranging from 41 to 371 ms across different studies, reflecting genuine differences in measurement technique, sample handling (in vivo vs ex vivo), and population variability, not measurement error alone [5]. The figures below are representative, illustrative values compiled from widely cited sources, not a single authoritative ground truth:

Sources: Stanisz et al. (3T tissue relaxometry, with 1.5T comparison) [3]; Bojorquez et al. (comprehensive literature review of reported 3T values, explicitly discussing the wide reported ranges) [5]; O’Reilly et al. (comparative grey/white matter T1 measurements across 50 mT, 1.5T, and 3T illustrating the field-strength trend directly) [4].

7. Mathematical Formalism — Relaxation Terms in the Bloch Equations

The relaxation terms in the Bloch equations, introduced in the companion Fundamentals of Nuclear Magnetic Resonance child page (Section 7 there), are the formal mathematical expression of everything described qualitatively above:

dMz/dt = … − (Mz − M0)/T1 dMx/dt = … − Mx/T2, dMy/dt = … − My/T2

These terms describe, respectively, the exponential recovery of Mz toward M0 with time constant T1 (Section 3.2) and the exponential decay of the transverse components toward zero with time constant T2 (Section 4.2). BPP theory (Section 3.1) supplies the microscopic physical justification for why T1 and T2 take the particular values they do in a given tissue at a given field strength, while the Bloch equations describe the resulting macroscopic magnetization behaviour without needing to reference the underlying molecular mechanism directly.

8. MRI Technologist and Radiologist Pearls — Common Misconceptions

• “T1 and T2 are properties of the scanner.” T1 and T2 are intrinsic properties of the tissue (and, per Section 6, of the field strength), not of the scanner hardware — the scanner’s TR/TE choices determine how much T1 or T2 weighting is expressed in a given image, but do not change the tissue’s underlying relaxation times.
• “T2 can be longer than T1.” Physically impossible for any given tissue at a given field strength (Section 4.3) — if an apparent measurement suggests otherwise, this points to a methodological or fitting error, not a genuine tissue property.
• “T2* and T2 are interchangeable terms.” As detailed in Section 5.1, they are related but numerically distinct quantities; using them interchangeably in a report or protocol description is a common but avoidable imprecision, particularly relevant when distinguishing genuinely T2-weighted spin-echo-family sequences from T2*-weighted gradient-echo-family sequences.
• “Relaxation times don’t matter for gradient-echo sequences.” Gradient-echo sequences are governed by T2* rather than T2 (Section 5.3) — the underlying relaxation physics still applies, just via the faster, field-inhomogeneity-inclusive T2* pathway rather than true T2.
• “Field strength doesn’t really change tissue contrast.” T1-weighted contrast can change meaningfully between 1.5T and 3T because T1 values themselves shift with field strength (Section 6.1) at different rates for different tissues — a protocol optimised for one field strength is not automatically optimal, unchanged, at another.

9. Practical and Clinical Relevance

Relaxation phenomena are the direct physical basis of essentially every contrast mechanism documented across MRIninja’s anatomical protocol pages. T1-weighted sequences exploit tissue T1 differences (Section 3); T2- and T2*-weighted sequences exploit T2/T2* differences (Sections 4–5); STIR and other inversion-recovery fat-suppression techniques exploit fat’s characteristically short T1 (Section 3.3) directly; and gadolinium-based contrast agents, documented in depth on the Contrast Media in MRI master page, work by paramagnetically shortening the local T1 (and, at sufficient concentration, T2) of tissue in their immediate vicinity — a direct practical application of the same BPP relaxation-efficiency principle (Section 3.1) that governs intrinsic tissue relaxation.

11. Evidence Gaps and Ongoing Debate

• True population reference ranges for tissue relaxation times remain unsettled. As Bojorquez and colleagues’ systematic literature review explicitly documents, reported relaxation times for nominally the same tissue vary substantially across studies (e.g. white matter T1 ranging from 699 to 1735 ms at 3T in the studies they reviewed), reflecting real differences in measurement methodology, sample condition (in vivo vs ex vivo), and possibly genuine population variability — no single authoritative reference-value table has achieved universal consensus adoption, which is why this page presents representative rather than definitive figures (Section 6.3) [5].
• Quantitative relaxometry standardisation. Efforts to make T1/T2 mapping a routine, standardised, vendor-independent quantitative clinical tool (rather than a research technique) continue, with ongoing debate about optimal acquisition and fitting methodology — outside the scope of this foundational physics page, but relevant to the future planned Reconstruction and Post-Processing Physics child page group (master page Section 4.11).
• Ultra-high-field relaxation behaviour. Most of the tissue relaxation data cited here (Section 6.3) comes from 1.5T and 3T studies; behaviour at 7T and above, while broadly following the same BPP-predicted trends, is less thoroughly characterised across the full range of clinically relevant tissues and is more directly addressed in the planned High-Field Physics child page group (master page Section 4.10).

12. Evidence-Based References

A. Guidelines / Consensus / Society Recommendations

No dedicated society guideline exists for foundational relaxation physics as such — this is textbook/landmark-paper physics and quantitative literature review material rather than a clinical practice area subject to society guidance. Category A is therefore not populated for this child page.

C. Important Prospective / Original Studies

D. Technical MRI Papers

End of document — Relaxation Phenomena — Child Page under the MRIninja MRI Physics — Fundamentals and Principles master page — v1.0 — July 2026 Parent page: MRI Physics — Fundamentals and Principles