Hubbry Logo
Gradient echoGradient echoMain
Open search
Gradient echo
Community hub
Gradient echo
logo
7 pages, 0 posts
0 subscribers
Be the first to start a discussion here.
Be the first to start a discussion here.
Gradient echo
Gradient echo
Gradient echo
View on Wikipedia
from Wikipedia

Gradient echo is a magnetic resonance imaging (MRI) sequence that has wide variety of applications, from magnetic resonance angiography to perfusion MRI and diffusion MRI. Rapid imaging acquisition allows it to be applied to 2D and 3D MRI imaging. Gradient echo uses magnetic gradients to generate a signal, instead of using 180 degrees radiofrequency pulse like spin echo; thus leading to faster image acquisition time.[1]

Mechanism

[edit]

Unlike spin-echo sequence, a gradient echo sequence does not use a 180 degrees RF pulse to make the spins of particles coherent. Instead, the gradient echo uses magnetic gradients to manipulate the spins, allowing the spins to dephase and rephase when required. After an excitation pulse (usually less than 90 degrees), the spins are dephased after a period of time (due to free induction decay) and also by applying a reversed magnetic gradient to decay the spins.[2] No signal is produced because the spins are not coherent. When the spins are rephased via a magnetic gradient, they become coherent, and thus signal (or "echo") is generated to form images. Unlike spin echo, gradient echo does not need to wait for transverse magnetisation to decay completely before initiating another sequence, thus it requires very short repetition times (TR), and therefore to acquire images in a short time.[2]

After echo is formed, some transverse magnetisations remains because of short TR.[2] Manipulating gradients during this time will produce images with different contrast. There are three main methods of manipulating contrast at this stage, namely steady-state free-precession (SSFP) that does not spoil the remaining transverse magnetisation, but attempts to recover them in subsequent RF pulses (thus producing T2-weighted images); the sequence with spoiler gradient that averages the transverse magnetisations in subsequent RF pulses by rotating residual transverse magnetisation into longitudinal plane and longitudinal magnetisation into transverse planes (thus producing mixed T1 and T2-weighted images), and RF spoiler that vary the phases of RF pulse to eliminates the transverse magnetisation, thus producing pure T1-weighted images.[1][2]

Gradient echo uses a flip angle smaller than 90 degrees, thus longitudinal magnetisation is not eliminated while flipping the spins. The larger the flip angle, the higher the T1 weighing of the tissue because more longitudinal magnetisation most recover to produce a difference in signals between the tissues.[2]

Steady-state free precession

[edit]
Main article: Steady-state free precession imaging

Steady-state free precession imaging (SSFP) or balanced SSFP is an MRI technique which uses short repetition times (TR) and low flip angles (about 10 degrees) to achieve steady state of longitudinal magnetizations as the magnetizations does not decay completely nor achieving full T1 relaxation.[1] While spoiled gradient-echo sequences refer to a steady state of the longitudinal magnetization only, SSFP gradient-echo sequences include transverse coherences (magnetizations) from overlapping multi-order spin echoes and stimulated echoes. This is usually accomplished by refocusing the phase-encoding gradient in each repetition interval in order to keep the phase integral (or gradient moment) constant. Fully balanced SSFP MRI sequences achieve a phase of zero by refocusing all imaging gradients.

MP-RAGE (magnetization-prepared rapid acquisition with gradient echo) [3] improves images of multiple sclerosis cortical lesions.[4]

Spoiling

[edit]

At the end of the reading, the residual transverse magnetization can be terminated (through the application of suitable gradients and the excitation through pulses with a variable phase radiofrequency) or maintained.

In the first case there is a spoiled sequence, such as the fast low-angle shot MRI (FLASH MRI) sequence, while in the second case there are steady-state free precession imaging (SSFP) sequences.

In-phase and out-of-phase

[edit]

In-phase (IP) and out-of-phase (OOP) sequences correspond to paired gradient echo sequences using the same repetition time (TR) but with two different echo times (TE).[5] This can detect even microscopic amounts of fat, which has a drop in signal on OOP compared to IP. Among renal tumors that do not show macroscopic fat, such a signal drop is seen in 80% of the clear cell type of renal cell carcinoma as well as in minimal fat angiomyolipoma.[6]

Effective T2 (T2* or "T2-star")

[edit]
Main article: T2*-weighted imaging

T2*-weighted imaging can be created as a postexcitation refocused gradient echo sequence with small flip angle. The sequence of a GRE T2*WI requires high uniformity of the magnetic field.[7]

Commercial names of gradient echo sequences

[edit]
Academic Classification Spoiled gradient echo Steady-State Free Precession (SSFP) Balanced Steady-State Free Precession (bSSFP)
Ordinary type Turbo type
(Magnetization preparation,
extremely low angle shot, short TR) 		FID-like Echo-like
Siemens FLASH
Fast Imaging using Low Angle Shot 		TurboFLASH
Turbo FLASH 		FISP
Fast Imaging with Steady-state Precession 		PSIF
Reversed FISP 		TrueFISP
True FISP
GE SPGR
Spoiled GRASS 		FastSPGR
Fast SPGR 		GRASS
Gradiant Recall Acquisition using Steady States 		SSFP
Steady State Free Precession 		FIESTA
Fast Imaging Employing Steady-state Acquisition
Philips T1 FFE
T1-weighted Fast Field Echo 		TFE
Turbo Field Echo 		FFE
Fast Field Echo 		T2-FFE
T2-weighted Fast Field Echo 		b-FFE
Balanced Fast Field Echo

VIBE (volumetric interpolated breath-hold examination) is an MRI sequence that produces T1-weighted gradient echo images in three-dimensions (3D). Apart from lower fluid signal intensity than a typical T1-weighted image, other appearances of VIBE images is similar to a typical T1-weighted image. Since its acquisition is only 30 seconds, suitable for breath-holding, it is used in breast and abdominal imaging to obtain high-resolution images minimising respiratory movement artifacts. VIBE images have low contrast in soft tissues and cartilage but have high contrast between the bony cortex and bone marrow. Bony lesions such as callus and fibrous tissue can also be readily distinguished from surrounding cortical bone because high contrast between the bone lesions and the bony cortex.[8]

References

[edit]
Revisions and contributorsEdit on WikipediaRead on Wikipedia

Gradient echo

View on Grokipedia
from Grokipedia
Gradient echo is a fundamental pulse sequence in magnetic resonance imaging (MRI) that generates an echo signal through the application and reversal of magnetic field gradients, rather than using a 180° refocusing radiofrequency (RF) pulse, enabling faster imaging with shorter repetition times (TR) and echo times (TE). This technique relies on the rephasing of transverse magnetization induced by gradient fields to produce a coherent signal at the echo time, making it particularly sensitive to magnetic field inhomogeneities and T2* relaxation effects. Unlike spin echo sequences, which refocus dephasing due to spin-spin interactions and are more robust against static field distortions, gradient echo sequences do not correct for these inhomogeneities, resulting in contrast primarily weighted by T1 and T2* relaxation times. The core principle of gradient echo involves a single excitation RF pulse—typically with a flip angle less than 90°—followed by gradient-induced dephasing and rephasing along the readout direction to form the echo, often combined with spoiling mechanisms to eliminate residual transverse magnetization and achieve steady-state conditions. Spoiling can be achieved through gradient methods, which apply unbalanced gradients to dephase spins, or RF spoiling, which increments the phase of successive RF pulses quadratically (e.g., by 117°) to suppress transverse steady-state signals and emphasize T1-weighted contrast. These approaches allow for high temporal resolution, with TR values as short as a few milliseconds, facilitating applications such as dynamic contrast-enhanced imaging and real-time monitoring. Gradient echo sequences, first developed in the late 1970s and refined in the 1980s for rapid imaging, offer advantages including reduced scan times, lower specific absorption rate (SAR) due to smaller flip angles, and versatility in 2D and 3D acquisitions, though they may exhibit banding artifacts in balanced steady-state free precession (SSFP) variants under field inhomogeneities. Clinically, they are widely used for T2*-weighted imaging to detect susceptibility effects like hemorrhage, iron deposition in tissues such as the liver or brain, and microstructural changes in neurodegenerative diseases, as well as in functional MRI (fMRI) for blood-oxygen-level-dependent (BOLD) contrast. In angiography and perfusion studies, their speed enables high-resolution vascular imaging without contrast agents in some cases.

Basic Principles

Pulse Sequence Overview

The gradient echo is a fundamental magnetic ( that generates echoes through the application of radiofrequency (RF) pulses and gradients, rather than relying on 180° refocusing RF pulses as in sequences. This approach exploits the (FID) signal, which is refocused by reversing the direction of a gradient to rephase spins that have dephased due to field inhomogeneities and gradient-induced effects. The sequence is particularly valued for its efficiency in producing images sensitive to T2* relaxation, where T2* accounts for both intrinsic T2 decay and additional dephasing from susceptibilities. Key components of the gradient echo include an excitation phase with a low-flip-angle RF , typically less than 90° to preserve longitudinal for rapid repetition, followed by a dephasing lobe (often a negative portion of the readout gradient) that intentionally spreads the spin phases across the imaging volume. Echo formation occurs when the readout gradient is reversed or applied positively, rephasing the spins at the echo time (TE) to produce a gradient-recalled echo during signal acquisition. The concludes with a spoiler gradient or rewinder to eliminate residual transverse magnetization if needed, though in steady-state variants, this may be omitted. Basic timing parameters governing the sequence are the repetition time (TR), which is kept short (often 5–50 ms) to enable fast imaging; the echo time (TE), minimized to reduce T2* effects (typically 1–10 ms); and the flip angle (α), optimized via the Ernst angle (α = cos⁻¹[e^{-TR/T1}]) for maximal signal at given TR and tissue T1 relaxation time. Under steady-state conditions assuming spoiling of transverse magnetization, the signal intensity S can be approximated by the equation: S≈M0sin⁡α1−e−TR/T11−cos⁡α e−TR/T1e−TE/T2∗S \approx M_0 \sin \alpha \frac{1 - e^{-TR/T_1}}{1 - \cos \alpha \, e^{-TR/T_1}} e^{-TE/T_2^*} where M_0 is the equilibrium magnetization, T_1 is the longitudinal relaxation time, and T_2^* is the effective transverse relaxation time; this derivation arises from solving the Bloch equations for repeated low-flip-angle excitations in steady state, assuming spoiling eliminates residual transverse magnetization and partial recovery of longitudinal magnetization between pulses. Gradient echo sequences were introduced in the early 1980s to overcome the time-intensive nature of spin echo methods, which required long TR to allow full T1 recovery and 180° refocusing pulses that doubled acquisition demands. A seminal practical implementation came in 1986 with the FLASH (fast low-angle shot) sequence by Haase et al., which demonstrated rapid 2D imaging using gradient-recalled echoes and low flip angles, marking a shift toward ultrafast MRI techniques. Relative to spin echo, gradient echo offers advantages such as shorter TR for accelerated scan times—often reducing acquisition from minutes to seconds—and enhanced suitability for volumetric 3D imaging by maintaining high signal efficiency in multi-slice or multi-echo formats.

Echo Formation and Gradients

In gradient echo sequences, spatial encoding and echo formation rely on controlled application of magnetic field gradients along the slice-select, phase-encoding, and frequency-encoding (readout) directions. These gradients impose position-dependent frequency shifts on the spins via the Larmor equation, causing dephasing of the transverse magnetization within each imaging voxel. Dephasing arises from both the intentionally applied gradients, which create linear phase dispersions across the voxel for spatial localization, and inherent magnetic field inhomogeneities, which introduce additional irreversible T2* decay. Unlike uniform RF excitation alone, which produces a free induction decay (FID) signal that decays rapidly due to these effects, the gradients enable refocusing of the coherent, gradient-induced phase shifts to form a measurable echo. The echo formation process begins immediately after the initial RF excitation pulse, which tips the longitudinal magnetization into the transverse plane. A negative (dephasing) lobe of the readout gradient is first applied along the frequency-encoding direction, intentionally spreading the phases of spins across the voxel (typically over half the duration or area of the subsequent positive lobe). This dephasing ensures that the signal starts from a low baseline rather than the immediate post-excitation peak, allowing sampling during the rephasing phase. The readout gradient is then switched to positive polarity with equal area (∫G(t) dt) but opposite sign, causing the spins to rephase progressively. The echo peaks at the echo time (TE), defined as the time from the RF pulse center to the rephasing center, where the net phase dispersion due to the readout gradient is zero, maximizing the coherent signal for data acquisition. Similar but momentary gradients are applied in the phase-encoding direction to encode spatial information in k-space, while the slice-select gradient operates during the RF pulse to define the imaging plane. Mathematically, the phase accumulation for a spin at position r\mathbf{r} due to a gradient waveform G(t)\mathbf{G}(t) is given by ϕ(r,t)=γ∫0tG(s)⋅r ds,\phi(\mathbf{r}, t) = \gamma \int_0^t \mathbf{G}(s) \cdot \mathbf{r} \, ds, where γ\gamma is the gyromagnetic ratio. This phase twist causes intravoxel dephasing, reducing the signal unless reversed. For echo formation in the readout direction (assuming 1D along x for simplicity), the dephasing lobe imparts ϕ(x)=−γxA\phi(x) = -\gamma x A (with area A>0A > 0), while the rephasing lobe accumulates +γxA+\gamma x A over time tt, such that at t=TEt = \mathrm{TE}, the net area ∫0TEG(s) ds=0\int_0^{\mathrm{TE}} G(s) \, ds = 0, yielding ϕ(x,TE)=0\phi(x, \mathrm{TE}) = 0 independent of xx. This condition refocuses the linear gradient-induced phases, forming the echo; the derivation follows from integrating the time-dependent frequency offset Δω(x,t)=γG(t)x\Delta \omega(x, t) = \gamma G(t) x over the waveform, ensuring symmetry around TE. A full phase evolution diagram illustrates this: phases fan out negatively during dephasing, then converge to zero at TE before fanning out positively, with the signal envelope peaking when the area under the gradient waveform balances. In contrast to spin echo sequences, gradient echoes do not employ a 180° refocusing RF pulse, which in spin echoes inverts the spin phases to compensate for both T2 and static field inhomogeneity-induced dephasing. Gradient reversal refocuses only the predictable, linear dephasing from the applied gradients, leaving static inhomogeneities (e.g., from susceptibility differences) uncompensated, resulting in T2* rather than pure T2 weighting and greater sensitivity to artifacts like signal voids near air-tissue interfaces.

Steady-State Variants

Steady-State Free Precession

Steady-state free precession (SSFP) is a variant of the gradient echo sequence in which transverse magnetization is preserved across successive excitations to achieve a dynamic steady state, enabling rapid imaging with high signal-to-noise ratio (SNR) and intrinsic T2/T1 contrast. This is accomplished by employing very short repetition times (TR much less than T2), balanced gradient waveforms that result in zero net gradient moment over each TR, and flip angles typically near the Ernst angle to optimize signal. The balanced gradients refocus phase dispersion caused by the imaging gradients, allowing the transverse magnetization to remain coherent and continuously available for readout rather than decaying fully between pulses. As a result, SSFP provides exceptional SNR efficiency, making it ideal for applications demanding fast acquisition and high temporal resolution. The steady-state transverse MssM_{ss} in balanced SSFP is given by Mss=M0(1−E1)sin⁡α1−E1cos⁡α−E2(cos⁡α−E1),M_{ss} = M_0 \frac{(1 - E_1) \sin \alpha}{1 - E_1 \cos \alpha - E_2 (\cos \alpha - E_1)}, where E1=e−TR/T1E_1 = e^{-TR/T_1} is the longitudinal relaxation factor, E2=e−TR/T2E_2 = e^{-TR/T_2} is the transverse relaxation factor, M0M_0 is the equilibrium , and α\alpha is the excitation flip . This arises from solving the Bloch equations under steady-state assumptions, where the configuration at the end of one TR matches the start of the next. The derivation begins with the transformation of the magnetization vector through the sequence elements: an RF pulse rotates the magnetization by α\alpha, followed by free precession (incorporating relaxation via the E1 and E2 factors and any off-resonance phase), and gradient refocusing that nulls dephasing. Representing these as rotation matrices, the steady-state condition m(0)=Am(0)+b\mathbf{m}(0) = A \mathbf{m}(0) + \mathbf{b} is solved for the fixed point, yielding the denominator as the determinant-related term balancing recovery and decay, with the numerator capturing the tipped transverse component. For short TR, the signal approximates Mss≈M0sin⁡α(1−E1)(1−E1cos⁡α)⋅T2T1+T2M_{ss} \approx M_0 \frac{\sin \alpha (1 - E_1)}{(1 - E_1 \cos \alpha)} \cdot \frac{T_2}{T_1 + T_2}, emphasizing the T2/T1 weighting, but the full form is essential for understanding off-resonance sensitivity. Balanced SSFP, often implemented as True FISP, represents the fully refocused form where gradients are precisely balanced to maintain coherence. In contrast, variants like FIESTA incorporate minor modifications in gradient timing or RF phasing but retain the core steady-state principles. These sequences excel in T2/T1-weighted imaging, providing bright signal from tissues with high T2/T1 ratios, such as myocardium or fluid-filled structures, and are widely used in cardiac MRI for real-time cine imaging of ventricular function and in abdominal MRI for non-contrast evaluation of organs like the liver. The high SNR efficiency stems from refocusing both longitudinal and transverse components, allowing up to twice the signal of conventional gradient echo at equivalent speed. A primary limitation of SSFP is its vulnerability to off-resonance effects, which cause banding artifacts due to accumulated phase from magnetic field inhomogeneities (B0 offsets) or susceptibility differences. In the steady state, the signal amplitude is modulated by a factor involving sin⁡(ϕ/2)/(ϕ/2)\sin(\phi/2) / (\phi/2), where ϕ=2πΔf⋅TR\phi = 2\pi \Delta f \cdot TR is the phase accrued from off-resonance frequency Δf\Delta f over TR; signal nulls occur when ϕ=(2k+1)π\phi = (2k+1)\pi, producing dark bands spaced by Δf=1/(2⋅TR)\Delta f = 1/(2 \cdot TR). These artifacts are exacerbated in regions of poor field homogeneity, such as near air-tissue interfaces, and necessitate precise B0 shimming to keep Δf\Delta f within the passband (typically < 1/(2 TR)). Unlike spoiled gradient echo sequences, which discard transverse coherence to avoid such issues, SSFP's preservation enhances SNR but demands careful field management. Recent advances in the 2020s have incorporated multi-band excitation into SSFP to accelerate imaging, enabling simultaneous multi-slice acquisition for faster coverage of large volumes, including whole-body protocols, while mitigating aliasing through parallel imaging techniques like CAIPIRINHA. More recent advances as of 2024 include network-based reconstructions for joint suppression of banding and flow artifacts, and accelerated submillimeter 3D acquisitions for BOLD fMRI at ultrahigh fields. This builds on the foundational echo formation in gradient echo by extending steady-state efficiency to multidimensional acceleration.

Spoiled Gradient Echo

The spoiled gradient echo sequence modifies the basic gradient echo to suppress residual transverse magnetization at the end of each repetition time (TR), ensuring that only longitudinal magnetization contributes to the subsequent excitation and yielding predominantly T1-weighted contrast. This suppression is achieved through spoiling techniques that disrupt coherent transverse components, preventing the buildup of a steady-state transverse magnetization that would otherwise introduce T2-dependent effects. Two primary methods are employed: gradient spoiling, which applies strong, unbalanced gradients (typically in the slice-select or readout direction) to dephase spins across the voxel such that the net transverse magnetization averages to zero, and RF spoiling, which introduces phase cycling to the excitation pulses. In RF spoiling, the phase of the nth RF pulse is incremented quadratically, typically as \phi_n = 117^\circ \cdot \frac{n(n-1)}{2}, causing refocused echoes to become incoherent over multiple cycles and effectively nulling the transverse steady state. Under the full spoiling condition, where transverse magnetization Mxy=0M_{xy} = 0 immediately before each excitation, the steady-state signal intensity simplifies to a form dominated by T1 relaxation, with a minor T2* modulation due to the echo time (TE). The signal equation is derived by considering the longitudinal recovery during TR and the projection by the flip angle α\alpha, assuming negligible transverse carryover: S=M0sin⁡α1−e−TR/T11−cos⁡α e−TR/T1e−TE/T2∗S = M_0 \sin\alpha \frac{1 - e^{-TR/T_1}}{1 - \cos\alpha \, e^{-TR/T_1}} e^{-TE/T_2^*} This expression, known as the Ernst equation in the spoiled regime, highlights how short TR values enhance T1 contrast while the exponential T2* term accounts for dephasing during TE. To optimize signal for a given tissue T1 and TR, the flip angle is selected as the Ernst angle αE=arccos⁡(e−TR/T1)\alpha_E = \arccos(e^{-TR/T_1}), which balances excitation efficiency and recovery to maximize steady-state longitudinal magnetization. Implementation often combines RF and gradient spoiling for robustness, with gradient moment nulling (zeroing the first-order moment of the spoiler gradient) applied to ensure uniform dephasing without introducing net phase shifts that could cause artifacts. This sequence excels in applications requiring high-resolution T1-weighted imaging, particularly in three-dimensional acquisitions where rapid coverage is essential. A prominent example is the magnetization-prepared rapid acquisition of gradient echoes (MP-RAGE) sequence, which segments spoiled gradient echoes following an inversion pulse to achieve excellent gray-white matter differentiation in brain imaging. Post-2015 developments have incorporated hybrid approaches combining RF phase cycling with optimized gradient spoilers to maintain effective spoiling while reducing specific absorption rate (SAR) in high-field MRI (≥3 T), allowing higher flip angles or faster repetitions without exceeding safety limits. These advancements build on steady-state principles to support quantitative T1 mapping and contrast-enhanced studies.

Contrast and Relaxation Properties

In-Phase and Out-of-Phase Imaging

In gradient echo imaging, the chemical shift between fat and water protons arises from their differing resonance frequencies, primarily due to the lower electron shielding around fat protons, resulting in a frequency difference Δf of approximately 3.5 ppm or 220 Hz at 1.5 T. This chemical shift causes the fat and water magnetization vectors to evolve at different rates, accumulating a relative phase φ = 2π Δf TE by the echo time TE. Consequently, in-phase images are acquired at TE values where φ is an integer multiple of 2π (TE = n / Δf, with n = 0, 1, 2, ...), yielding additive signal from fat and water, while out-of-phase images use TE where φ = π + 2π n (TE = (2n + 1) / (2 Δf)), leading to subtractive interference. At 1.5 T, the first out-of-phase echo typically occurs at ~2.3 ms and the first in-phase at ~4.6 ms, with the cycle repeating every ~4.5 ms (1/Δf). The Dixon method exploits these phase differences for fat-water separation, originally using a pair of echoes to generate water-only and fat-only images. In a voxel containing both fat (fraction f) and water (fraction 1 - f), the observed signal is S(TE) = (1 - f) S_water + f S_fat e^{i φ}, assuming negligible T2* differences or corrections; here, S_water and S_fat represent the magnitudes modulated by relaxation. For the two-point Dixon technique with in-phase (S_in, φ = 0) and out-of-phase (S_out, φ = π) echoes, separation yields water and fat signals as: W=Sin+Sout2,F=Sin−Sout2W = \frac{S_\text{in} + S_\text{out}}{2}, \quad F = \frac{S_\text{in} - S_\text{out}}{2} This derivation assumes equal T2* decay across echoes; when f ≈ 0.5, complete phase cancellation occurs in out-of-phase images (S_out ≈ 0), highlighting boundaries. Extensions to multiple echoes enhance accuracy by sampling more phase points, mitigating noise and field inhomogeneities through least-squares fitting. The iterative decomposition of water and fat with echo asymmetry and least-squares estimation (IDEAL) algorithm, introduced in 2005, optimizes echo spacing for robust decomposition even with asymmetric acquisitions. Advancements in the 2000s–2020s incorporated multi-peak fat spectral modeling in IDEAL to account for fat's complex spectrum (six main peaks shifted relative to water), improving quantification in tissues with heterogeneous fat composition. More recent developments as of 2023–2025 include deep learning approaches for artifact-free fat-water separation in Dixon MRI, such as style transfer models that predict swap-free images from acquired data, and accelerated techniques combining IDEAL with phase-contrast MRI for simultaneous water-fat and velocity mapping, enhancing efficiency and robustness. Clinically, in-phase and out-of-phase imaging enables detection of intracellular lipid in lesions, such as adrenal adenomas, where lipid-rich adenomas exhibit signal intensity loss (>20%) on out-of-phase images relative to in-phase due to intravoxel phase cancellation. This is particularly valuable for characterizing indeterminate adrenal masses, distinguishing benign adenomas from metastases or pheochromocytomas, which lack sufficient microscopic fat. Dual-echo sequences are parameterized with short TEs (e.g., first echo out-of-phase, second in-phase) to minimize T2* decay, as longer TEs amplify signal loss from field inhomogeneities. A prominent artifact in out-of-phase images is the India ink border, manifesting as a thin black rim at fat-water interfaces from partial volume phase cancellation, which can aid in confirming fat presence but may obscure subtle boundaries if severe. Short echo times are selected to reduce such T2*-related effects during TE accrual.

Effective T2 Relaxation (T2*)

In magnetic resonance imaging (MRI), effective transverse relaxation time, denoted as T2*, characterizes the decay of transverse magnetization in the presence of both intrinsic spin-spin interactions and extrinsic magnetic field perturbations. This decay is faster than the intrinsic T2 relaxation, as quantified by the relation 1T2∗=1T2+1T2′\frac{1}{T_2^*} = \frac{1}{T_2} + \frac{1}{T_2'}, where T2 represents the irreversible dephasing due to molecular-level interactions, and T2' accounts for reversible dephasing arising from static magnetic field inhomogeneities and applied gradients. The causes of T2* shortening encompass intrinsic factors, such as spin-spin relaxation (T2), which result from fluctuating generated by nearby during molecular motion. Extrinsic contributions include differences (Δχ) between tissues, leading to variations (ΔB₀); effects from differing frequencies of molecular like and ; and of through applied gradients or inherent field gradients, which induces additional phase dispersion. Susceptibility-induced inhomogeneities are particularly prominent in gradient echo sequences, where dephasing is not refocused by 180° pulses, amplifying sensitivity to these effects. Measurement of T2* in gradient echo imaging relies on acquiring signals at multiple echo times (TE) following excitation, assuming a mono-exponential decay model for the transverse magnetization magnitude. The signal intensity S(TE) is given by S(TE)=S0e−TE/T2∗,S(TE) = S_0 e^{-TE / T_2^*}, where S₀ is the initial signal at TE = 0. This form derives from the Bloch equations, where the transverse magnetization evolves as Mxy(t)=M0e−t/T2∗eiϕ(t)M_{xy}(t) = M_0 e^{-t / T_2^*} e^{i \phi(t)}, with the phase ϕ(t)\phi(t) accumulating due to off-resonance effects; the observed signal magnitude thus decays exponentially with time constant T2* after accounting for the rephasing gradient. For susceptibility-induced dephasing, the local field perturbation inside a spherical inclusion is approximated by the Lorentz sphere model as ΔB=Δχ3B0\Delta B = \frac{\Delta \chi}{3} B_0, where Δχ is the susceptibility difference and B₀ is the main field strength; this contributes to the off-resonance frequency shift Δω=γΔB\Delta \omega = \gamma \Delta B, accelerating the phase accrual and shortening T2*. Quantitative T2* mapping involves fitting multi-echo gradient echo data to this model, often with corrections for noise and confounder effects like fat-water interference. In gradient echo sequences, T2* relaxation builds on the dephasing mechanisms during echo formation by incorporating these relaxation effects into the signal decay post-rephasing, enabling susceptibility weighting that highlights field-sensitive contrasts. Advancements in quantitative T2* mapping techniques since the 2010s, including multi-echo acquisitions with R2* relaxometry (where R2* = 1/T2*), have improved accuracy for applications such as tissue iron assessment, achieving strong correlations (r ≈ 0.98) with biopsy-validated concentrations up to 40 mg/g in liver tissue at 1.5 T. These methods, refined through multicenter validations in the 2020s, incorporate fat separation and noise corrections to enhance reproducibility. As of 2023–2025, further advancements include AI-accelerated T2* mapping for reduced acquisition times and high-resolution techniques using echo merging and undersampling, improving quantification in brain, cardiac, and low-field MRI while maintaining accuracy.

Clinical and Commercial Aspects

Applications in MRI

Gradient echo sequences play a pivotal role in functional magnetic resonance imaging (fMRI), where they enable blood oxygenation level-dependent (BOLD) contrast by exploiting T2* susceptibility effects from deoxyhemoglobin variations during neural activation, facilitating high-resolution brain mapping. This approach offers superior temporal resolution compared to spin-echo methods, allowing rapid acquisition of multiple echoes to capture dynamic hemodynamic responses with volumes updated every 1-2 seconds. In structural imaging, gradient echo techniques are extensively used for acquiring high-resolution 3D T1-weighted volumes, particularly in neuroimaging and musculoskeletal applications, where they provide excellent soft-tissue contrast for volumetric analysis of brain anatomy and joint structures. For angiography, time-of-flight (TOF) methods leverage flow-related enhancement in gradient echo sequences to visualize arterial structures without contrast, suppressing stationary tissue signals through repeated radiofrequency excitations while preserving inflowing blood signals. Specialized applications include susceptibility-weighted imaging (SWI), a gradient echo-based technique that enhances visualization of venous structures and hemorrhage by amplifying phase differences from magnetic susceptibility variations in blood products and iron deposits. In cardiac imaging, steady-state free precession variants of gradient echo enable cine sequences for assessing wall motion and function, offering high contrast between myocardium and blood with short repetition times for breath-hold acquisitions. Quantitative perfusion mapping via dynamic contrast enhancement (DCE) employs spoiled gradient echo sequences to track gadolinium uptake kinetics, yielding parameters like Ktrans for evaluating tumor vascularity and tissue permeability in oncology. Overall, gradient echo sequences excel in fast acquisition times, enabling whole-brain coverage in under 10 minutes for many protocols, but they are prone to motion artifacts and susceptibility distortions, particularly in regions near air-tissue interfaces. Emerging applications at ultra-high fields like 7T leverage enhanced resolution and contrast for finer detection of subtle epileptogenic lesions in drug-resistant epilepsy patients, with advancements as of 2025 providing up to a 20–40% increase in lesion-detection rates over 3T.

Commercial Sequence Names

Major MRI vendors employ proprietary names for their gradient echo (GRE) sequences, which often correspond to spoiled or steady-state variants but are optimized for specific hardware and clinical workflows. These names facilitate clinical use but can complicate cross-vendor comparisons, as equivalents share underlying principles like low flip angles for rapid imaging. Siemens Healthineers designates its spoiled GRE sequence as FLASH (Fast Low Angle Shot), a fast T1-weighted method using RF spoiling to minimize transverse magnetization. For balanced steady-state free precession (SSFP), it uses TrueFISP (True Fast Imaging with Steady-state Precession), which balances gradients for high signal in tissues with short T2. VIBE (Volumetric Interpolated Breath-hold Examination) is Siemens's 3D spoiled GRE variant tailored for abdominal imaging, incorporating fat suppression and parallel imaging acceleration. GE Healthcare's spoiled GRE is known as SPGR (Spoiled Gradient Recalled echo), emphasizing incoherent readout for T1 contrast in dynamic studies. Its SSFP implementation is FIESTA (Fast Imaging Employing Steady-state Acquisition), providing bright blood and fluid signals with minimal banding artifacts. LAVA (Liver Acquisition with Volume Acceleration) represents an advanced 3D spoiled GRE sequence optimized for liver imaging, integrating parallel imaging techniques like SENSE for faster volumetric coverage. Philips Healthcare employs T1-FFE (T1-weighted Fast Field Echo) for its spoiled GRE, supporting high-resolution angiography and perfusion. The balanced FFE (Fast Field Echo) serves as its SSFP equivalent, enabling cine cardiovascular applications with strong T2/T1 weighting. For fat-water separation in GRE imaging, mDIXON uses multi-echo acquisitions to generate in-phase and out-of-phase images via Dixon methods. Canon Medical Systems (formerly Toshiba) offers equivalents such as Fast FE (Fast Field Echo) for spoiled GRE and True SSFP for balanced SSFP, aligning with VIBE-style volumetric acquisitions under names like 3D Fast. Historically, GE's early steady-state gradient echo sequence was branded GRASS (Gradient Recalled Acquisition in the Steady State), a coherent GRE from the 1980s similar to FISP. Modern balanced SSFP is branded FIESTA. Commercial names have evolved since the 1990s to incorporate advancements like parallel imaging (e.g., GRAPPA in Siemens VIBE or SENSE in GE LAVA), reducing scan times by factors of 2-4 without sacrificing resolution, particularly post-2000. These optimizations reflect hardware improvements, with names varying by software versions but maintaining core GRE physics.
VendorSpoiled GREBalanced SSFPSpecialized Variant
SiemensFLASHTrueFISPVIBE (abdominal 3D)
GE HealthcareSPGRFIESTALAVA (liver 3D)
PhilipsT1-FFEbalanced FFEmDIXON (fat-water separation)
Canon (Toshiba)Fast FETrue SSFP3D Fast (volumetric)

References

  1. https://pmc.ncbi.nlm.nih.gov/articles/PMC3502662/
  2. https://web.stanford.edu/class/rad229/Notes/Lecture-09/Rad229_2020_Lecture09D-RFSpoiling.pdf
  3. https://pmc.ncbi.nlm.nih.gov/articles/PMC4009216/
  4. https://radiopaedia.org/articles/gradient-echo-sequences-1?lang=us
  5. https://mriquestions.com/gradient-echo.html
  6. https://www.magnetic-resonance.org/ch/08-03.html
  7. https://mriquestions.com/types-of-gre-sequences.html
  8. https://pubs.rsna.org/doi/abs/10.1148/radiology.186.1.8416546
  9. https://www.sciencedirect.com/science/article/pii/0022236486904336
  10. https://www.imaios.com/en/e-mri/sequences/gradient-echo
  11. https://doi.org/10.1002/jmri.24163
  12. https://web.stanford.edu/class/rad229/Notes/C1-BalancedSSFP.pdf
  13. https://pubmed.ncbi.nlm.nih.gov/31746055/
  14. https://www.journalofcmr.com/article/S1097-6647(24)01150-5/fulltext
  15. https://onlinelibrary.wiley.com/doi/full/10.1002/mrm.30064
  16. https://onlinelibrary.wiley.com/doi/10.1002/mrm.1910210210
  17. https://mriquestions.com/spoiled-gre-parameters.html
  18. https://pmc.ncbi.nlm.nih.gov/articles/PMC5324613/
  19. https://www.thieme-connect.com/products/ejournals/pdf/10.1055/a-1460-8566.pdf
  20. https://labs.dgsom.ucla.edu/file/59469/M229_Lecture12_WaterFat.pdf
  21. https://mriquestions.com/in-phaseout-of-phase.html
  22. https://pubs.rsna.org/doi/10.1148/radiology.153.1.6089263
  23. https://onlinelibrary.wiley.com/doi/abs/10.1002/mrm.20624
  24. https://www.semanticscholar.org/paper/IDEAL-Water-Fat-Decomposition-with-Multipeak-Fat-Yu-Shimakawa/475d687068b1b928d7c5a5d4383fadeacd695ec3
  25. https://pubmed.ncbi.nlm.nih.gov/36686622/
  26. https://pmc.ncbi.nlm.nih.gov/articles/PMC12034513/
  27. https://www.mdpi.com/2306-5354/12/5/502
  28. https://pubs.rsna.org/doi/abs/10.1148/rg.2016150139
  29. https://radiopaedia.org/articles/black-boundary-artifact?lang=us
  30. https://radiopaedia.org/articles/t2-relaxation-1?lang=us
  31. https://pubmed.ncbi.nlm.nih.gov/3406413/
  32. https://pubs.rsna.org/doi/10.1148/radiol.221856
  33. https://www.sciencedirect.com/science/article/pii/S0720048X25002098
  34. https://eurradiolexp.springeropen.com/articles/10.1186/s41747-024-00476-8
  35. https://www.frontiersin.org/journals/physiology/articles/10.3389/fphys.2023.1281147/full
  36. https://pmc.ncbi.nlm.nih.gov/articles/PMC5552052/
  37. https://pubs.rsna.org/doi/10.1148/radiology.182.3.1535892
  38. https://mriquestions.com/time-of-flight-tof-mra.html
  39. https://onlinelibrary.wiley.com/doi/10.1002/mrm.20198
  40. https://pubs.rsna.org/doi/abs/10.1148/rg.284075031
  41. https://ajronline.org/doi/10.2214/AJR.12.8510
  42. https://onlinelibrary.wiley.com/doi/10.1111/epi.18682
  43. https://mriquestions.com/commercial-acronyms.html
  44. https://www.imaios.com/en/e-mri/sequences/sequences-acronyms
  45. https://mriquestions.com/true-fispfiesta.html
  46. https://asset.fujifilm.com/www/uk/files/2024-05/ec8914c80b6b1fec4c826319e700a9f8/Fujifilm-MRI-Acronym-Doc.pdf
  47. https://www.mrgschedule.com/schedule_static/MRI_Acronyms.pdf
  48. https://www.imv-imaging.com/media/9258/signa-explorer-air-iq-edition-data-sheet-mr291.pdf
  49. https://www.mriclinicalcasemap.philips.com/content/downloads/Brochure_Clinical-applications.pdf
  50. https://sandrofenelon.com.br/wp-content/uploads/2022/04/MRI-Acronyms-Canon-Hitachi-v2019.pdf
  51. https://pmc.ncbi.nlm.nih.gov/articles/PMC9199974/
Add your contribution
Related Hubs
User Avatar
No comments yet.