Zero/Ultrashort Echo Time Techniques in MRI
MRI images are created by applying a static magnetic field to an object being imaged and manipulating the nuclear magnetization with radiofrequency (RF) pulses and magnetic field gradients. Different signal-generation and readout methods may be chosen depending upon the particular imaging application. These methods differ in the sequences of magnetic field gradients and RF pulses that are applied to the object. Two common types of signal-generation methods in MRI are spin-echo (SE) and gradient-echo (GRE).
Spin-echo (SE) sequences work by applying a 90° RF pulse to tip the net magnetization of the object into the transverse plane, allowing the spins to dephase, before applying a 180° RF pulse to reverse the dephasing produced by magnetic field (B0) inhomogeneities and produce an echo (the spin-echo). The resulting signal has a low level of sensitivity to magnetic field inhomogeneities and mostly reflects intrinsic T2-relaxation.
For this reason, SE sequences are commonly used for producing robust, high-quality images with minimal artefacts.
A gradient-echo (GRE) sequence similarly generates an echo, but by applying two gradients of opposite polarity sequentially. The second gradient reverses the dephasing of spins produced by the first gradient, but does not reverse the dephasing of spins due to B0 inhomogeneities, and so produces images in which T2* contrast is shown. Since GRE sequences produce images that capture magnetic susceptibility, they are widely used in functional MRI (fMRI) to measure changes in blood oxygenation in the brain, which is a proxy for neural activity. Changes in blood oxygenation impact the level of magnetic field distortion (and hence the level of dephasing) and so are measurable using GRE sequences.
The conventional SE and GRE approaches have a number of limitations. In particular, they are unsuitable for imaging tissue or signals with very short T2 times, since the transverse net magnetisation in such cases has largely decayed before the intended echo would be formed. When applied to fMRI, these techniques may be affected by image distortions as a result of susceptibility artefacts and be highly sensitive to body motion[1]. Additionally, the techniques can result in a large amount of acoustic noise due to rapid gradient switching.
As a result of these shortcomings, there is an interest in free induction decay (FID)-based approaches, in which the RF signal is measured immediately or very shortly after the RF excitation and before any refocusing occurs. The FID-based imaging can be subdivided into zero echo time (ZTE) imaging and ultrashort echo time (UTE) imaging. In ZTE imaging, a short excitation pulse is applied whilst the encoding magnetic field gradient is applied, and readout begins almost immediately after an intrinsic T/R switch delay. In UTE, the excitation happens in the absence of gradients, and acquisition starts after the excitation pulse, together with the application of the encoding gradient.
In the case of both ZTE and UTE imaging, there is a very short time gap between excitation and measurement, and so there is very little time for the signal to decay. For this reason, FID-based approaches to MR imaging are useful for imaging structures with very short T2-relaxation times, such as bones and tendons, which are typically difficult to image with conventional MRI[2].
Earlier this year, a review was published on the outlook on zero/ultrashort echo times in functional MRI[1]. ZTE and UTE have been explored as techniques for producing images tracking neural activity in the brain. Like the GRE sequences discussed above, the contrast observed in ZTE and UTE brain images results from changes in blood flow in the brain but is understood to result from the inflow of magnetically unsaturated blood into the brain, rather than changes in blood oxygenation. The zero/ultrashort TE used in FID-based sequences involves minimal dephasing of spins, meaning that the techniques are tolerant of magnetic field (B0) variations, such as those induced by body motion and susceptibility gradients. In the case of ZTE sequences, the use of incremental gradient switching between readouts largely resolves gradient-induced artefacts and reduces acoustic noise. Additionally, the use of incremental gradient switching in ZTE minimises eddy currents, supporting compatibility with the use of simultaneous electrophysiological recordings, e.g. using EEG[3].
As noted in the review, clinical translation of FID-based fMRI has been challenging, with only preliminary human applications being published so far. The obstacles to clinical translation include limited access to local T/R coils for delivering RF precisely to small regions and to hardware components that enable the ultrafast T/R switching required, as well as off-the-shelf sequences. The authors of the review at least expect FID-based fMRI to address research questions in the future, especially at ultrahigh magnetic fields.
Patents play an important role in protecting developments in MRI. In particular, patents may be granted for novel aspects of controlling a magnetic resonance imaging system to obtain MR data for reconstructing an image. For example, these may include aspects of the sequences applied—such as the timing, order, or parameters of RF pulses and gradients—or the manner in which different sequences are combined. Patents may also relate to new ways of processing MR data to reconstruct an image, for instance to generate a particular type of imaging contrast or to reduce or eliminate artefacts. In addition, patents may protect novel and advantageous aspects of the hardware of the imaging system. Where such inventions arise in the context of ZTE or UTE, patent protection may therefore be sought for developments relating to those techniques.
- S. Mangia, S. Michaeli, and O. Gröhn, “ Outlook on zero/ultrashort echo time techniques in functional MRI,” Magnetic Resonance in Medicine 95, no. 2 (2026): 714–723, https://doi.org/10.1002/mrm.70065.
- Gerta Halilaj, Nebi Cemeta. Beyond radiation: Emerging applications of MRI in dental diagnostics and clinical practice[J]. Journal of Dentistry and Multidisciplinary Sciences, 2025, 1(1): 31-46. doi: 10.3934/jdms.2025004
- Paasonen J, Laakso H, Pirttimaki T, et al. Multi-band SWIFTenables quiet and artefact-free EEG-fMRI and awake fMRI studies in rat. Neuroimage. 2020;206:116338
Other News
