Protons and neutrons are the basic constituents of atomic nuclei. Charged spinning particles such as protons behave like tiny bar magnets: they have both a north and south pole and produce an external magnetic field. These spinning protons are also termed "magnetic dipoles" or simply just "spins". When more than one proton and neutron occur in a nucleus, the dipoles tend to pair up together, canceling each other’s magnetic effect. Therefore, in nuclei with an even number of protons and even number of neutrons there is no magnetic dipole moment. Nuclei with either (or both) an odd number of protons or an odd number of neutrons do have a net magnetic diploe moment, making the phenomenon of nuclear magnetic resonance possible.

Not only is hydrogen the most abundant nucleus in biological tissues, its single unpaired spinning proton results in the strongest "magnetic moment" of any element. Because of its greater concentration in tissues and greater strength per nucleus, the signal that can be elicited from hydrogen is more than 1000 times stronger than from any other element. Therefore, hydrogen is used as the signal source in most clinical MR scans.

In the absence of an external magnetic field, the magnetic dipoles of spinning hydrogen protons are randomly oriented in the body and there is no net magnetisation of a tissue. If placed in a very strong externally applied magnetic field B_o, the spinning protons align themselves with or against field.

It is also important to understand that individual protons do not align themselves precisely along the z-axis but "precess" around it, analogous to a spinning top. The rapidity or frequency of precession of hydrogen nuclei is solely dependent on magnetic field strength; the higher the external magnetic field strength, the faster the spins precess.

Net intrinsic tissue magnetisation M_o is only a tiny fraction of the main external magnetic field B_o and therefore when M_o is aligned with B_o in the longitudinal or "z" axis it is extremely difficult to measure. However, the tiny M_o can be measured if it is rotated or "tipped" away from the longitudinal direction into perpendicular plane, the transverse or "xy" plane. When the net intrinsic tissue magnetisation M_o is tipped into the transverse plane it is referred to as M_xy. Tipping M_o into the transverse plane is accomplished by irradiating the patient with a short burst or "pulse" of electromagnetic radiation oscillating at the Lamor frequency of hydrogen. This burst of radio waves at the Lamor frequency is called a "90⁰ RF pulse" when it is applied for just enough time to tip the net longitudinal magnetisation M_o by 90⁰, into the transverse plane where it can be measured and reconstruct to produce an image.

Immediately after a 90⁰ RF pulse has tipped the longitudinal magnetisation M_o from its baseline state along the z axis into the transverse plane, longitudinal magnetisation is zero. That is, the residual vector along the z direction (longitudinal axis) is zero. As time progresses, magnetisation begins to re-grow or "recover" along the z longitudinal direction. The rate of recovery of longitudinal magnetisation is described by the longitudinal (or "spin-lattice") relaxation time T1. T1 is defined as the time required for re-growth or recovery of 63% of the magnetisation along the longitudinal direction after a 90⁰ RF pulse.

After a 90⁰ RF pulse, transverse magnetisation or M_xy is maximal, i.e., it is the same strength as M_o was just prior to the Rf pulse. Just as individual dipoles precess around the z axis, the net bulk transverse magnetisation M_xy also precesses around the z axis. With the passage of time, precessing dipoles interact with surrounding macromolecules and tend to get out of phase with each other. Their magnetic dipole vectors come to point in different directions from each other. This is termed "de-phasing". As the dipoles de-phase, the net transverse magnetisation vector M_xy becomes progressively smaller. If sufficient time elapses, every dipole pointing in one direction has a corresponding dipole pointing in the opposite direction. When this happens, transverse magnetisation decay is caused by reorientation of magnetic dipoles from the transverse axis to the z axis; a much larger source of signal decay is due to loss of phase coherence among magnetic dipoles within the same tissue volume element or voxel.

Extrinsic causes of transverse de-phasing include (1) non-uniformity of the static magnetic field B_o and (2) non-uniformities caused by intentionally applied magnetic gradients (used to "read" pr encode frequencies). These two sources of de-phasing are reversible (achieved by applying appropriate RF pulses and properly reversing gradients). On the other hand, intrinsic sources of transverse de-phasing are caused by interactions among magnetic dipoles and their local magnetic environments. These sources of de-phasing are irreversible. The rate or irreversible signal loss due to inherent intrinsic interactions between dipoles and their surrounding magnetic tissue environment is described by a second magnetic relaxation time called T2. The T2 (also called "spin-spin") relaxation time is the time required by for transverse magnetisation to decrease to 37% of its original value (which it had immediately after 90⁰ Rf pulse).

It is important to understand that T1 and T2 relaxation are occurring simultaneously in the same voxel of excited tissue. The brightness or whiteness of each pixel in an image represents the RF signal strength emitted from the irradiated tissue in the corresponding voxel and is measured only in the transverse plane. Signal strength is determined by the amplitude of the precessing transverse magnetisation. Both T1 and T2 (orT2*) contribute to the signal output. T1 recovery contributes because it governs the rate of return to equilibrium magnetisation and therfore determines the amount of magnetisation that is then available for re-tipping into the transverse plane diminishes as spins de-phase.

A tissue with a short T1 (rapid re-growth of net magnetisation in the z axis) has a greater subsequent longitudinal magnetisation M_o (more recovery or re-growth) at any given time than a tissue with a long T1 has.

If tissue has a short T1, more longitudinal magnetisation re-grows in between 90⁰ pulse and so more magnetisation exists to be tipped into the xy plane. This results in more measured signal and a brighter or "whiter" or "hotter" appearance on MR images. Therefore, tissues with short T1 values appear bright on T1-weighted images.

After a 90⁰ RF pulse tips the longitudinal magnetisation into the transverse plane. T2 relaxation effects take over. Tissues with a short T2 value de-phase more quickly and lose signal intensity more quickly. Therefore, tissues with short T2 values appear dark on T2-weighted images.