Expanding the Horizons of SSFP: A Deep Dive into Magnetization Dynamics

Steady-state free precession (SSFP) sequences have long been workhorses of fast magnetic resonance imaging (MRI). Their ability to deliver high signal-to-noise ratio (SNR) and distinct contrast in a fraction of the time required by conventional sequences makes them indispensable for everything from cardiac cine imaging to musculoskeletal studies. At the heart of SSFP’s efficiency lies a unique magnetization state—a dynamic equilibrium achieved by applying a rapid train of radiofrequency (RF) pulses. This article explores the physics, clinical applications, and practical considerations that make SSFP an essential tool in the modern MRI suite.

Foundations: Magnetization, RF Pulses, and Relaxation

Before diving into the steady state itself, it is crucial to revisit the behavior of nuclear magnetization. A sample placed in a static magnetic field (B₀) develops a net magnetization vector (M) aligned with that field, conventionally along the +z direction. When an RF pulse (B₁) is applied at the Larmor frequency, M is tipped away from equilibrium by an angle called the flip angle (α). After the pulse, the transverse components (Mx, My) precess about B₀ and decay due to transverse (T₂) relaxation, while the longitudinal component (Mz) recovers toward equilibrium with time constant T₁. These processes are described by the Bloch equations, which form the mathematical backbone of MRI signal generation.

In a typical gradient-echo (GRE) sequence, one waits a relatively long repetition time (TR) so that Mz recovers substantially before the next excitation. In contrast, SSFP sequences use TR values that are short compared to T₁ and T₂—often on the order of a few milliseconds. This rapid pulsing prevents the magnetization from ever returning to full equilibrium; instead, it evolves into a periodic steady state where the magnetization vector follows a closed loop over each TR interval.

Transverse and Longitudinal Components in the Steady State

To understand SSFP dynamics, consider a train of identical RF pulses separated by a constant TR. After a few initial pulses (the so-called “transient period”), the magnetization reaches a steady state in which the magnetization vector at the beginning of each TR is identical. In this regime, the transverse magnetization (Mt) and longitudinal magnetization (Mz) become coupled through the repeated application of the RF pulse and the relaxation that occurs during the TR interval.

Analytically, the steady-state signals can be derived from the Ernst equation for a simple spoiled gradient-echo sequence, but SSFP introduces additional complexity because the residual transverse magnetization is not spoiled; instead, it is refocused and contributes to the signal in subsequent TR intervals. This recycling of transverse coherence is what gives SSFP its high SNR efficiency.

Types of SSFP: Balanced, Coherent, and Spoiled Variants

The term “SSFP” encompasses several related sequence families. The most widely used is balanced SSFP (bSSFP), also known by proprietary names such as TrueFISP, FIESTA, and balanced FFE. In bSSFP, all three gradient axes are rewound so that the net gradient area over each TR is zero. This results in refocusing of both T₂* and off-resonance phase accumulation, producing a signal that depends primarily on the T₂/T₁ ratio of the tissue.

Coherent (Unbalanced) SSFP

In contrast, coherent SSFP (e.g., FISP, GRASS) does not fully compensate the readout gradient, leaving a dephasing moment that spoils or partially spoils the transverse magnetization over multiple TRs. This reduces the effective transverse coherence and yields a signal that depends on the flip angle, TR, and the specific dephasing per TR. While coherent SSFP is less sensitive to B₀ inhomogeneities than bSSFP, it provides lower SNR for the same imaging time.

Spoiled GRE (Spoiled SSFP-like)

Although often referred to as “spoiled SSFP,” sequences such as SPGR or FLASH intentionally spoil all residual transverse magnetization after each readout. These are not true steady-state free precession sequences because the transverse magnetization is not allowed to contribute to the steady state. The signal follows the Ernst formula and yields a T₁-weighted contrast. True SSFP, by definition, requires that some or all of the transverse magnetization be reused in subsequent TRs.

Magnetization Dynamics in Balanced SSFP: Oscillations and Banding

In bSSFP, the repeated RF pulses cause the magnetization vector to precess around both B₀ and the effective field (B₁). Between pulses, the transverse magnetization dephases due to local B₀ inhomogeneities (off-resonance). When the next RF pulse arrives, the phase of the transverse component relative to the RF phase determines how much of it will be tipped back into the longitudinal direction or become part of the next transverse signal.

This phase sensitivity leads to a characteristic signal oscillation as a function of off-resonance frequency, known as the bSSFP banding artifact. At frequencies where the accumulated phase over TR is an odd multiple of π, the transverse coherence is destroyed, and signal intensity drops dramatically. The distance between these nulls (the “band spacing”) is given by 1/TR. Short TR values widen the band spacing, reducing the number of bands across a given field of view, but also increase the sensitivity to B₀ inhomogeneities because the condition for dephasing becomes more stringent.

Mathematical Description of the Steady State

The steady-state magnetization in bSSFP can be expressed using the solutions to the Bloch equations for a repeated train of pulses. For a given off-resonance angle θ (where θ = 2πγΔB₀TR), the transverse and longitudinal components after the nth pulse can be written in closed form. The final steady-state signal magnitude is proportional to:

S_ss ∝ M₀ sin α · (T₂ / (E₂(E₁−1) + (E₁−E₂)(1−cos θ) + . . .))

where E₁ = exp(−TR/T₁) and E₂ = exp(−TR/T₂). This formula reveals that the steady-state signal depends on both T₁ and T₂ and is maximized when the flip angle is set to the Ernst angle for the effective T₁ of the steady state—which, in bSSFP, is approximately the average of T₁ and T₂.

For a rigorous treatment, the reader is referred to this classic paper by Ernst and Anderson and the comprehensive review by Scheffler and Lehnhardt.

Contrast Mechanisms: Why T₂/T₁ Wins

One of the most clinically appealing properties of bSSFP is its intrinsic dependence on the ratio T₂/T₁. Tissues with long T₂ relative to T₁, such as fluid, fat, and blood, appear bright, while tissues with shorter T₂/T₁, such as muscle, become darker. This provides a unique contrast that is distinct from conventional T₁- or T₂-weighted images. For example, in cardiac cine imaging, bSSFP produces bright blood (high T₂/T₁) against the dark myocardium, without the need for contrast agents.

The contrast can be modulated by adjusting the flip angle. At small flip angles (<30°), the sequence behaves similarly to spoiled GRE with T₁-weighting. At larger flip angles (40°–70°), the T₂/T₁ weighting becomes dominant. In extreme cases (e.g., flip angle close to 90°), the sequence becomes very sensitive to off-resonance and banding artifacts.

Clinical Applications: Where SSFP Excels

SSFP sequences have become standard in several clinical areas due to their speed and high SNR efficiency.

Cardiac Imaging

Cardiac cine MRI almost exclusively uses bSSFP (TrueFISP) because it provides excellent blood-myocardium contrast at very high temporal resolution (TR ~ 3 ms). The sequence is used for ventricular function assessment, wall motion studies, and valvular flow evaluation. The bright blood effectively outlines the endocardial boundaries, making automated segmentation more reliable.

Musculoskeletal Imaging

In the knee and shoulder, bSSFP offers high-resolution isotropic 3D imaging of cartilage and labrum. The fat suppression inherent in the sequence (fat has high T₂/T₁) can be advantageous, but banding artifacts from bone-tissue interfaces require careful shimming. Several recent studies have demonstrated the utility of bSSFP for cartilage T₂ mapping.

Abdominal Imaging

While bSSFP is sensitive to respiratory motion, faster 2D sequences (e.g., single-shot bSSFP) can freeze motion for free-breathing liver or kidney imaging. The bright blood and high SNR help delineate vessels and cysts.

Functional and Contrast-Enhanced Studies

SSFP has been used for non-contrast MR angiography (e.g., fresh blood imaging, inflow techniques) and for dynamic contrast-enhanced studies where combined T₁ and T₂* sensitivity is desired.

Overcoming Artifacts: Shimming, TR, and Frequency Scouting

Banding artifacts remain the primary limitation of bSSFP. Strategies to mitigate them include:

  • Shimming: Localized shimming (e.g., volume shim) reduces off-resonance variations across the anatomy.
  • Short TR: Decreasing TR widens the band spacing, shifting nulls farther apart. This often requires stronger gradients and higher receiver bandwidths.
  • Flip angle optimization: Large flip angles exacerbate banding; sometimes a small reduction in α can improve homogeneity.
  • Frequency scouting (phase-cycled bSSFP): Acquiring multiple images with different RF phase increments and then combining them (e.g., maximal intensity projection) removes banding entirely. This technique is now commercially available on many platforms.
  • Linear shimming coils: For 3T imaging, B₀ inhomogeneities are larger, making banding more problematic. 3T bSSFP benefits from high-order shim and short TR.

Fat-Water Separation in SSFP

Fat and water precess at frequencies separated by ~440 Hz (at 1.5T). Because the bSSFP steady state is periodic with off-resonance, fat and water signals can appear at different brightness levels depending on the phase accumulation over TR. This effect is called the “chemical shift artifact of the second kind.” In practice, one can suppress fat by choosing a TR that places water at the center of a high-signal band and fat at a null (or by using a frequency-selective fat-saturation pulse).

Alternatively, Dixon techniques have been adapted for SSFP, acquiring images at multiple echo times to separate fat and water without the SNR penalty of fat saturation. This approach is particularly useful in the abdomen and spine.

Advanced Topics and Future Directions

The flexibility of SSFP continues to inspire new pulse sequence developments.

Hyperecho and Variable Flip Angle

By gradually increasing the flip angle during the pulse train, one can generate a “hyperecho”—a refocusing effect that boosts SNR in later echoes without increasing RF power. This approach, known as variable flip angle (VFA) SSFP, has been applied to relaxometry and to reduce SAR at ultra-high fields.

Magnetization Transfer SSFP

Combining SSFP with off-resonance saturation pulses enables measurement of magnetization transfer ratios (MTR) in just a few seconds, useful for characterizing white matter and cartilage composition.

High-Resolution 3D SSFP

Parallel imaging and compressed sensing allow 3D isotropic bSSFP sequences with submillimeter voxels for structural imaging of the brain (e.g., 3D FIESTA for cranial nerves) and for assessment of inner ear fluid.

SSFP at Ultra-High Field (7T)

At 7T, the B₀ inhomogeneities are severe, but the short T₂ of tissues can be exploited for high-resolution microscopy. New RF coils and advanced shimming strategies are making 7T bSSFP feasible for the brain and knee.

Conclusion

Steady-state free precession sequences are far more than a simple “fast imaging” alternative. Their unique magnetization dynamics—characterized by recycling of transverse coherence and a periodic steady state—yield a powerful combination of SNR efficiency and tissue contrast that is unmatched by conventional GRE or SE sequences. Understanding the Bloch-equation-level behavior, the origins of banding artifacts, and the methods to mitigate them is essential for any MRI practitioner who wishes to harness the full potential of SSFP. With continued innovation in pulse design, shimming, and reconstruction, SSFP is poised to remain a central technique in clinical MRI for years to come.

For a deeper technical reference, see the review by Scheffler and Hennig. For clinical protocols, the Radiopaedia article on balanced SSFP provides an accessible overview.