Dr. Evelyn Cartwright, PhD
Magnetic Resonance Imaging (MRI) is a premium diagnostic imaging modality that provides exceptional soft-tissue contrast without using ionizing radiation. It is based on the principles of nuclear magnetic resonance (NMR), where atomic nuclei with an odd number of protons or neutrons interact with static magnetic fields and radiofrequency (RF) pulses. Because the human body consists primarily of water and fat, MRI focuses almost exclusively on the hydrogen nucleus (^1H or single proton).
This article explores nuclear spin angular momentum, the Larmor precession equation, T1 and T2 relaxation mechanisms, spatial encoding, and safety principles.
A hydrogen proton possesses an intrinsic property called spin angular momentum. Because the proton is a rotating positive charge, its spin generates a tiny magnetic field called a magnetic dipole moment (represented as a vector vector \vec{μ}).
In the absence of an external magnetic field, these magnetic dipoles point in random directions, resulting in a net magnetization of zero. However, when the patient is placed in a strong, static magnetic field (B_0, measured in Tesla):
* The dipoles align either parallel (low-energy state) or anti-parallel (high-energy state) to the field.
* A slight excess of protons align in the parallel direction, creating a macroscopic Net Magnetization Vector (M_0 or \vec{M}) pointing along the direction of the static field (z-axis, or longitudinal axis).
Protons do not sit still. The torque exerted by the magnetic field B_0 on the rotating magnetic dipole causes the protons to precess (wobble) around the axis of B_0, similar to a spinning top precessing in a gravitational field.
The frequency of this precession, known as the Larmor frequency (\omega_0 in radians/sec, or f_0 in MHz), is directly proportional to the strength of the static magnetic field:
Where \gamma is the gyromagnetic ratio, a constant unique to each atomic nucleus. For hydrogen (^1H):
Thus, in a standard clinical 1.5 Tesla scanner, hydrogen protons precess at:
3.0 Tesla scanner, the Larmor frequency is 127.74 MHz.To obtain a signal from the precessing protons, we must perturb them from their equilibrium state. This is achieved by applying a radiofrequency (RF) magnetic pulse (B_1) perpendicular to the z-axis:
* Resonance Condition: The frequency of the RF pulse must match the Larmor precession frequency of the protons.
* When the RF pulse is turned on, the protons absorb energy, causing the net magnetization vector \vec{M} to tip away from the z-axis into the x-y plane (transverse plane).
* A 90^\circ RF pulse tips the entire vector \vec{M} into the transverse plane, creating transverse magnetization (M_{xy}) while reducing longitudinal magnetization (M_z) to zero.
* This rotating transverse magnetization induces an electrical current in an external receiver coil, producing the raw Free Induction Decay (FID) signal.
Once the RF pulse is turned off, the protons return to their original equilibrium alignment through two simultaneous, independent physical processes:
T1 is the time required for M_z to recover to 63\% of its original value. Solid structures and large molecules (like fat) transfer energy quickly, resulting in short T1 times. Free fluids (like cerebrospinal fluid or water) transfer energy slowly, resulting in long T1 times.
T2 is the time required for M_{xy} to decay to 37\% of its original value. T2 is always shorter than T1.
T2 Relaxation: In clinical scanners, dephasing occurs much faster than T2 due to local magnetic field inhomogeneities (caused by tissue interfaces, bone, or iron deposits). The combined relaxation rate is 1/T2^* = 1/T2 + 1/T2_{inhom}.
To form a 3D image, we must localize the origin of the MR signal. This is done by using three orthogonal gradient coils (G_x, G_y, G_z) to linearly vary the magnetic field across the patient:
1. Slice Selection (G_z): A gradient is applied along the z-axis during the RF pulse. This varies the Larmor frequency along the patient. The RF pulse is designed to contain a narrow range of frequencies, exciting only the slice of protons that match those frequencies.
2. Phase Encoding (G_y): A gradient is turned on briefly along the y-axis, causing protons along that axis to precess at different rates temporarily. When the gradient is turned off, they return to the same Larmor frequency but retain a phase offset corresponding to their vertical position.
3. Frequency Encoding / Readout (G_x): A gradient is turned on along the x-axis during signal readout. This maps horizontal positions to specific precession frequencies.
The raw signals are filled into a 2D data matrix called K-space. A 2D Inverse Fast Fourier Transform (2D-IFFT) converts K-space data into the final anatomical image.
Because MRI utilizes exceptionally strong magnets, safety is paramount:
* Projectiles: Ferromagnetic objects (scissors, oxygen tanks, implants) are pulled toward the magnet bore with immense velocity.
* SAR (Specific Absorption Rate): RF pulses deposit thermal energy in the patient. The SAR (measured in W/kg) must be monitored to prevent tissue heating.
* Gradient Induced Fields: Rapidly changing magnetic fields can induce electrical currents in nerves (peripheral nerve stimulation) and produce intense acoustic noise (up to 120 dB), requiring hearing protection.
MRI physics represents a highly complex interplay of quantum mechanics, electromagnetism, and signal processing. By manipulating static fields, RF pulses, and spatial gradients, medical physicists can probe the molecular environment of tissues. The ability to tune T1, T2, and proton density contrasts enables MRI to serve as the gold standard for soft-tissue visualization, assisting clinical diagnosis and high-precision radiotherapy planning.