Balanced Steady-State Free Precession (bSSFP) Sequence — Physics, Contrast, Off-Resonance Banding, Variants, Vendor Nomenclature (TrueFISP, FIESTA, Balanced FFE / b-TFE) and Clinical Use

Balanced Steady-State Free Precession (bSSFP) Sequence — Physics, Contrast, Off-Resonance Banding, Variants, Vendor Nomenclature (TrueFISP, FIESTA, Balanced FFE / b-TFE) and Clinical Use

MRIninja Knowledge Base | Sequence Child Page Builds on: MRI Sequences Overview · Gradient Echo (GRE / FLASH) · RF Pulses and Pulse-Sequence Physics Version 1.0 — October 2026

1. Introduction and Scope

Balanced steady-state free precession (bSSFP) is a gradient-echo sequence family in which every gradient applied within one repetition time is compensated by a gradient of opposite polarity, so that gradient-induced dephasing is zero at the end of each TR [1]. Because neither the longitudinal nor the transverse magnetisation is spoiled, both reach a steady state, and the sequence combines two properties that few other sequences offer together: a very high signal-to-noise ratio per unit time and a contrast that depends on the T2/T1 ratio rather than on T1 or T2 alone [1]. The same two properties explain its dominant clinical roles — cardiac cine imaging, heavily T2-weighted 3D imaging of fluid-filled structures and nerves (CISS / FIESTA-C), and non-contrast angiography — and its dominant weakness, sensitivity to off-resonance.

This page assumes the general gradient-echo principles of the Gradient Echo (GRE / FLASH) page and the RF-pulse physics of RF Pulses and Pulse-Sequence Physics as already known. It develops the physics specific to bSSFP, the parameters that control it, the main variants, the cross-vendor naming (including the Philips names balanced FFE and b-TFE), and the practical failure modes. The angiographic applications built on bSSFP — inflow inversion-recovery bSSFP (NATIVE TrueFISP, B-TRANCE, Inhance Inflow IR, Time-SLIP), flow-sensitive-dephasing bSSFP and the QISS technique — are developed on Triggered Non-Contrast-Enhanced MR Angiography and are not repeated here.

Information status. Statements are attributed to the source in which they were verified. Quantities stated as "derived" are arithmetic consequences of a quoted relationship, shown so that the reader can check them; they are not claims from the sources. Where a source could not be read in full, this is stated in the bibliography.

2. Pulse-Sequence Structure

2.1 Gradient Balancing — What "Balanced" Means

In an ordinary (unbalanced) SSFP sequence, the gradients used for slice selection, phase encoding and frequency encoding leave a residual net dephasing at the end of each TR. In bSSFP, each gradient lobe is followed by a lobe of opposite polarity and equal area — "each applied gradient pulse is compensated by a gradient pulse with opposite polarity" [1] — so the net gradient moment over one TR is zero in all three axes. The only remaining source of dephasing within a TR is therefore the off-resonance (static field inhomogeneity, susceptibility, chemical shift) of the spins, which is what Section 5 develops. The Siemens name TrueFISP is the vendor name for this balanced form of the sequence that was originally introduced as FISP by Oppelt and colleagues [1].

2.2 RF Train, Phase Alternation and Echo Time

The excitation train consists of an initial α/2 preparation pulse, separated by TR/2 from the first α pulse, followed by a continuous train of pulses of alternating sign (+α, −α, +α …) [1]. The echo time is TE = TR/2; at that time "the magnetization vector is exactly aligned along the x-axis" [1]. Because the echo forms midway between two pulses, the signal that reaches the receiver is the result of nearly complete refocusing, and the measured signal decays with exp(−TE/T2) rather than exp(−TE/T2*): the sequence behaves like a spin echo in its sensitivity to T2 [1].

2.3 Typical Sequence Geometry

bSSFP requires a short TR — Scheffler and Lehnhardt give "3–6 ms to minimize banding artefacts" at 1.5 T — and relatively high flip angles of about 50–80° (up to 70–90° for tissues with similar T1 and T2) [1]. These are the features that make the sequence fast (single-shot and segmented 2D acquisitions, 3D acquisitions with sub-millimetre voxels) and also energy-intensive (Section 6).

3. Signal and Contrast

3.1 The T2/T1 Contrast

For TR much shorter than T1 and T2 — "true for most tissues at 1.5 T with TR of 3–5 ms" [1] — and on-resonance, the standard simplified steady-state signal for TE = TR/2 is

S = M₀ · sin α / [ (T1/T2 + 1) − (T1/T2 − 1)·cos α ] · exp(−TE/T2)

and the flip angle that maximises it satisfies cos α_opt = (T1/T2 − 1)/(T1/T2 + 1). The equation is written here from standard theory; in the retrieved text of reference [1] the equation itself did not render, and it was checked only for consistency with the values that source quotes: an optimum flip angle "around 70–90°" for tissues of similar T1 and T2, and a maximum signal that "approaches 50% of M0" [1]. The consequence is that contrast depends on the ratio T2/T1: "liquids and fat appear bright" because their T2/T1 ratio is high [1].

The T1/T2 values in the table are illustrative ratios chosen to show the trend, not measured tissue values. The table also shows why bSSFP images look "bright for water, bright for fat, dark for solid tissue" and why a single flip angle is a compromise between tissues.

3.2 Behaviour After Gadolinium

Scheffler and Lehnhardt note that gadolinium reduces T1 and T2 together, so that the T2/T1 ratio of enhanced tissue remains nearly constant [1]. Consequently bSSFP is intrinsically a poor sequence for depicting enhancement, and it is not used as a post-contrast T1-weighted sequence; this is a consequence of the mechanism just quoted rather than a statement taken directly from the source.

3.3 Fat

Fat has a very high signal on bSSFP because its T2/T1 ratio resembles that of liquids [1]. For angiography the same source calls the fat signal "very intense" and states that efficient fat saturation is "mandatory" [1]; fat suppression, where used, is an additional preparation or a water-selective excitation, not an intrinsic property of the sequence.

3.4 Flow

Flow and inflow effects add to the signal during the transient phase, and 3D acquisitions lose the inflow enhancement that 2D acquisitions show [1]. Scheffler and Lehnhardt report higher blood-to-myocardium contrast than with FLASH in cardiac imaging [1], which follows from the T2/T1 contrast of Section 3.1. Reasoning, not a quoted statement: blood is therefore bright on bSSFP largely because of its T2/T1 ratio rather than because of inflow. The inflow-based angiographic variants (inversion-recovery and quiescent-interval) deliberately re-introduce flow dependence by adding a preparation, as developed on the angiography page.

4. The Transient Phase and Approach to Steady State

The steady state is not reached instantaneously. Scheffler and Lehnhardt state that "the duration of the transient phase is of the order of 50–500 excitation pulses", and that without a preparation "the initial signal fluctuations are severe" [1]. Four strategies are described [1]:

• α/2 preparation pulse (Section 2.2), which places the magnetisation on the α/2 cone for a smoother approach;
• Linearly increasing flip angles over the first 5–15 pulses;
• TIDE: a 90° pulse, 180° refocusing pulses, then decreasing flip angles toward 50–70°;
• "Catalyst" preparation schemes that establish the steady-state magnetisation.

Practical consequence, general reasoning not specific to one source: in a segmented or inversion-prepared acquisition, the images are acquired during or just after the transient, so k-space segmentation, view ordering and the number of start-up excitations determine the contrast, and a change in these parameters can change image contrast, not only artefact. The finite duration of the RF pulses also modifies the transient and the steady state; the signal discrepancy of the ideal (instantaneous-pulse) theory has been quantified at 10–20% in common protocols [3].

5. Off-Resonance and the Banding Artefact

5.1 Mechanism

The phase accumulated by spins of off-resonance frequency ν over one TR is θ = 2πνTR. The steady-state signal profile is 2π-periodic in θ and drops sharply at θ = ±π [1]. In practice this produces dark bands in regions where the local field differs from the nominal value by about ±1/(2·TR). The source quotes the example "θ = ±π, or ν = ±167 Hz, for TR = 3 ms" [1].

The same source quotes typical field variation of "approximately 50 Hz for the brain, and 100–200 Hz for the abdomen at 1.5 T" [1]. Comparing the two shows (derived) that at TR = 3–6 ms most of the brain lies inside the first band, whereas abdominal field variation can reach it, and that a slow bSSFP with TR 15–20 ms — the regime of CISS, which the source places at "TR of about 15–20 ms" — cannot tolerate even brain-level field variation (±25–33 Hz). This is why slow, high-resolution bSSFP is used only in the phase-cycled forms of Section 7.

5.2 Field Strength

The SCMR consensus states that bSSFP is the default method of choice for cine imaging at 1.5 T, but that "at 3 T, the increased sensitivity of bSSFP to off-resonance effects tends to worsen dark banding and flow artifacts" [2]. Field inhomogeneity and susceptibility effects, expressed in Hz, scale with B₀, so the same TR leaves a smaller safety margin at higher field (general reasoning consistent with [2]).

5.3 Mitigation

6. SNR, Speed and SAR

The principal advantages stated in the literature are "a very high signal-to-noise ratio" and fast acquisition per unit time [1]. The cost is RF power: with high flip angles and a very short TR, "this can easily exceed the SAR limits, especially in ultra-high field applications beyond 1.5 T" [1]. Reasoning, not a quoted statement: the relevant practical consequence is that, when SAR limits the sequence at 3 T, the operator is forced to lower the flip angle or lengthen the TR — and the latter, as Section 5.1 shows, lowers the banding tolerance. Sequence settings that look interchangeable on the console therefore interact through the physics, which is why a protocol validated at 1.5 T should not be transferred unchanged to 3 T.

7. Variants and Related Sequences

7.1 CISS / FIESTA-C in Detail

FIESTA-C is described as a modified FIESTA that uses "two balanced-SSFP acquisitions performed with different RF phase-cycling schemes"; Siemens CISS is its stated equivalent ("Constructive Interference in Steady State") [4]. Because the two acquisitions have different off-resonance profiles, the dark bands of one are filled by the signal of the other; the combination is performed automatically after acquisition, may modestly increase reconstruction time, and may use a maximum-intensity or a sum-of-squares-type combination, the latter possibly giving smoother signal [4]. The same slow bSSFP regime (TR of about 15–20 ms) is what gives the high T2-weighted contrast between CSF and nerves exploited in the CPA, cranial-nerve and brachial-plexus protocols of this platform (Section 9).

7.2 A Frequent Misgrouping: DRIVE Is a Turbo Spin-Echo Technique

Platform protocol pages sometimes list "CISS / DRIVE / FIESTA-C" together as the 3D heavily T2-weighted sequence family. They are used for the same clinical purpose, but they are not the same physical sequence. DRIVE is a "driven equilibrium radio frequency reset pulse" applied with a T2-weighted 3D turbo spin-echo sequence [5]; CISS and FIESTA-C are bSSFP-based [1,4]. DRIVE is a TSE option (it appears as a selectable setting in the TSE parameter list of the Philips console screenshots supplied by the site owner), not a bSSFP variant, and therefore does not share the banding behaviour of Section 5.

8. Vendor Nomenclature

Philips: balanced FFE versus balanced TFE. Both names appear in the literature retrieved for this page — "balanced FFE" in the reviews [1,4] and "balanced turbo field echo (B-TFE)" in a renal-artery imaging study from a group with Philips affiliation [7]. The retrieved sources do not state how the two names relate, so this page does not claim a console-level distinction between them; the Philips console labels, and the fields that change between releases, should be checked on the scanner. Entries for Canon and Fujifilm come from a single secondary educational source [6] and should be checked against the vendor's own documentation.

9. Clinical Applications on This Platform and in the Literature

The brachial-plexus and CPA protocols are described in their own pages; this page only supplies the physical reasoning behind the sequence choice.

10. Artefacts and Pitfalls

11. Choosing bSSFP Against Other Sequences

12. Evidence Gaps and Ongoing Debate

• Physics is well established; clinical comparisons are not uniform. The signal, banding and transient descriptions rest on a technical review [1] and are not disputed; accuracy claims for specific applications (for example renal artery imaging [8]) come from single studies.
• Consensus coverage is narrow. The one consensus statement verified for this page (SCMR protocols [2]) concerns cardiac imaging; no consensus or guideline specific to the neuroradiological or angiographic uses of bSSFP was verified.
• Philips naming. The relationship between "balanced FFE" and "balanced TFE" is not established in the retrieved sources (Section 8).
• Not covered by retrieved sources. Magnetisation-transfer effects, intrinsic diffusion sensitivity and the chemical-shift (fat–water) profile of bSSFP are the subject of published work (titles were seen in search results) but were not read for this page and are not described here.
• Second-hand vendor names. The Canon and Fujifilm names rest on a single secondary source [6]; BASG is listed on the platform overview page only.

13. Evidence-Based References

References are listed by evidence tier and recency. The only consensus statement included is cardiac; the remaining references are technical reviews and studies.

B. Consensus Statements

C. Important Prospective / Original Studies

D. Technical MRI Papers and Reviews

End of document — Balanced Steady-State Free Precession (bSSFP) Sequence — MRIninja v1.0 — October 2026