RF Pulse Research Articles

A fast MOSFET rf switch for low-field NMR and MRI.

TRansmit Array Spatial Encoding (TRASE) MRI uses trains of rf pulses alternately produced by distinct transmit coils. Commonly used coil switching involving PIN diodes is too slow for low- and ultra-low-field MRI and would introduce wait times between pulses typically as long as each individual pulse in a few mT. A MOSFET-based rf switch is described and characterised. Up to hundreds of kHz, it allows for sub-μs switching of rf currents from a single amplifier to several coils with sufficient isolation ratio and negligible delay between pulses. Additionally, current switching at null current and maximum voltage can be used to abruptly stop or start pulses in series-tuned rf coils, therefore avoiding the rise and fall times associated with the Q-factors. RF energy can be efficiently stored in tuning capacitors for times as long as several seconds. Besides TRASE MRI, this energy storage approach may find applications in fast repeated spin-echo experiments. Here, a threefold acceleration of TRASE phase-encoding is demonstrated when MOSFET switches are used instead of fast reed relays.

Global head SAR assessment of MRI-induced temperature change

Purpose: Underestimation of SAR (Specific Absorption Rate) in MRI poses potential risks to patients and in particular individuals with compromised thermoregulation. EU standards imposes a head SAR limit of 3.2 W/kg as SAR effects are negligible at up to 3 W/kg. We developed a phantom and protocol where heating due the RF pulse is measured and verified against scanner displayed SAR.

Microwave thermionic electron gun for synchrotron light sources

Thermionic RF guns are the source of electrons used in many practical applications, such as drivers for synchrotron light sources, preferred for their compactness and efficiency. RadiaBeam Systems has developed a new thermionic RF gun for the Advanced Photon Source at Argonne National Laboratory, which would offer substantial improvements in reliable operations with robust interface between the thermionic cathode and the cavity, as well as better RF performance, compared to existing models. This improvement became possible by incorporating new pi-mode electromagnetic design, robust cavity back plate design, and a cooling system that will allow stable operation for up to 1 A of beam current and 100 Hz rep rate at 3.0 μs RF pulse length, and 70 MV/m peak on axis field in the cavity. In this paper we discuss the electromagnetic and engineering design of the cavity and provide the test results of the new gun.

Determining a Line Strength in the ν3 Band of the Silyl Radical Using Quantum Cascade Laser Absorption Spectroscopy.

Silane (SiH4) plasmas are widely used for the deposition of hydrogenated amorphous silicon (a-Si:H) films. Nevertheless, the chemical processes governing film deposition are still incompletely understood. Moreover, there is still no general method available to determine the absolute concentration of the silyl radical (SiH3), which is the accepted chemical precursor of a-Si:H films. In this study, a 10% silane in helium RF plasma was spectroscopically investigated between 2085 and 2175 cm-1 using an external cavity quantum cascade laser (EC-QCL) based spectrometer. This led to the identification of 4 distinct species from their absorption features: SiH4, disilane (Si2H6), SiH3, and an unassigned short-lived species. Furthermore, 17 absorption features of SiH3 were identified and unambiguously assigned. Fast spectral scanning of selected absorption features belonging to the four species in a 10 Hz pulsed RF plasma enabled the measurement and interpretation of their temporal behavior in terms of plausible chemical reactions involving silicon containing species. By quantitatively measuring the decay of the SiH3 a ← a pP4 (5) transition at 2151.3207 cm-1 after the discharge was stopped, its line strength (S) was determined to be (7.5 ± 5.5) × 10-20 cm2 cm-1 mol-1.

Efficient and sideband-free 1H-detected 14N magic-angle spinning NMR.

Indirect detection via sensitive spin-1/2 nuclei like protons under magic-angle spinning (MAS) has been developed to overcome the low spectral sensitivity and resolution of 14N NMR. The 14N quadrupolar couplings cause inefficient encoding of the 14N frequency due to large frequency offsets and make the rotor-synchronization of the evolution time necessary. It is shown that 14N rf pulses longer than the rotor period can efficiently encode 14N frequencies and generate spinning sideband free spectra along the indirect dimension. Average Hamiltonian and Floquet theories in the quadrupolar jolting frame (QJF) are used to treat the spin dynamics of the spin-1 quadrupolar nucleus under long 14N rf pulses and MAS. The results show that the rf action can be described by a scaled and phase-shifted effective rf field. The large quadrupolar frequency offset becomes absent in the QJF and therefore leads to sideband-free spectra along the indirect dimension. More importantly, when a pair of long 14N rf pulses are used, the distribution of the phase shift of the effective rf field does not affect the 14N encoding for powder samples; thus, high efficiencies can be obtained. The efficient and sideband-free features are demonstrated for three 1H/14N indirectly detected experiments using long 14N pulses under fast MAS.

Low SAR 31 P (multi-echo) spectroscopic imaging using an integrated whole-body transmit coil at 7T.

Phosphorus (31P) MRSI provides opportunities to monitor potential biomarkers. However, current applications of 31P MRS are generally restricted to relatively small volumes as small coils are used. Conventional surface coils require high energy adiabatic RF pulses to achieve flip angle homogeneity, leading to high specific absorption rates (SARs), and occupy space within the MRI bore. A birdcage coil behind the bore cover can potentially reduce the SAR constraints massively by use of conventional amplitude modulated pulses without sacrificing patient space. Here, we demonstrate that the integrated 31P birdcage coil setup with a high power RF amplifier at 7 T allows for low flip angle excitations with short repetition time (T R) for fast 3D chemical shift imaging (CSI) and 3D T 1‐weighted CSI as well as high flip angle multi‐refocusing pulses, enabling multi‐echo CSI that can measure metabolite T 2, over a large field of view in the body. B 1 + calibration showed a variation of only 30% in maximum B 1 in four volunteers. High signal‐to‐noise ratio (SNR) MRSI was obtained in the gluteal muscle using two fast in vivo 3D spectroscopic imaging protocols, with low and high flip angles, and with multi‐echo MRSI without exceeding SAR levels. In addition, full liver MRSI was achieved within SAR constraints. The integrated 31P body coil allowed for fast spectroscopic imaging and successful implementation of the multi‐echo method in the body at 7 T. Moreover, no additional enclosing hardware was needed for 31P excitation, paving the way to include larger subjects and more space for receiver arrays. The increase in possible number of RF excitations per scan time, due to the improved B 1 + homogeneity and low SAR, allows SNR to be exchanged for spatial resolution in CSI and/or T 1 weighting by simply manipulating T R and/or flip angle to detect and quantify ratios from different molecular species.

Single-spin vector analysis of strongly coupled nuclei in TOCSY NMR experiments.

This paper presents a new way to represent the effect of complex radio-frequency (rf) pulse sequences on J-coupled nuclear spin systems. The model uses a vector representation of the single-spin interactions (chemical-shift and rf interactions) and provides a simple route to gain analytical insight into multipulse NMR experiments. The single-spin Hamiltonian is expressed in an interaction representation as Fourier components. These Fourier components are combined for the two spins to establish the averaged coupling term of the Hamiltonian. This effective Hamiltonian is fast to calculate as only single-spin rotations are used and followed by simple summation of numbers for reconstruction of given coupling interactions. The present method is used to gain analytical insight into the performance of the J-coupling transfer sequence DIPSI-2 through two figures of merit (FOM) providing useful information for optimization of such pulse sequences. The first FOM (ΞAB) reports the efficiency of the desired total correlation spectroscopy transfer and should be as large as possible, while the second (ΞHet) reports the potential leakage of coherence to a heteronuclear spin and should be as small as possible.

Development and analysis of synoptic evaluation strategy for research platforms for Cochlear Implants and hearing aids

Cochlear Implants—Hearing Aids(CI/HA) offer opportunities for researchers wishing to advance algorithm development, since the existing commercial CIs/HAs are closed/sealed for customization, due to intellectual property and safety requirements. In general, Research Platforms (RPs), such as CCi-MOBILE (developed at CRSS-CILab, UTDallas), emulate the functionality of CIs/HAs, and allow researchers to insert their customized sub-algorithms for either benchtop (soundbooth) or field testing. To date, little formalized studies have explored RP performance effectiveness or a formal safety test paradigm. Here, a two phase comprehensive evaluation strategy is proposed that involves acoustical signal processing and user specified electrical stimulation configuration for CCi-MOBILE. While the full evaluation strategy is tedious (+407 h of audio testing: Acoustical test {380 h}, Stimulation parameter test {27 h}), an agile reduced safety test and researcher performance evaluation paradigm is developed by sub-sampling the comprehensive test and system evaluation parameter space. The synoptic safety and performance evaluation paradigm is evaluated to selectively prioritize test conditions based on: (i) acoustical diversity and electrical current level characteristics and (ii) CI stimulation configuration test conditions using physical limits of RF pulses. The proposed synoptic safety and performance evaluation paradigm is shown to be efficient, effective, and ensures reliability of each potentially deployed RP in the field.

A time domain signal equation for multi-echo spin-echo sequences with arbitrary excitation and refocusing angle and phase.

Accurate T2 mapping using multi-echo spin-echo is usually impaired by non-ideal refocusing due to B1+ inhomogeneities and slice profile effects. Incomplete refocusing gives rise to stimulated echo and so called "T1-mixing" and consequently a non-exponential signal decay. Here we present a time domain formula that incorporates all relaxation and pulse parameters and enables accurate and realistic modelling of the magnetization decay curve. By pulse parameters here we specifically mean the actual refocusing angle and axis, and phase angle of both the excitation and refocusing pulse. The method used for derivation comprises the so called Generating functions approach with subsequent back-transformation to the time domain. The proposed approach was validated by simulations using realistic RF pulse shapes as well as by comparison to phantom measurements. Excellent agreement between simulations and measurements underpin the validity of the presented approach. Conclusively, we here present a complete time domain formula ready to use for accurate T2 mapping with multi-echo spin-echo sequences.

Loss Path Influence on the MRI Radio Frequency Pulse Sequence: A Theoretical Evidence.

The RF pulse is initiated from either the loop or loopless MRI antenna. It has shown an increased advancement in recent times. Somehow, the concept has proven successful in the MR imaging procedure. Using the fundamental theories of the MRI concept, mathematical experimentation was carried out analytically to investigate the Loss Path Concept (LPC). The LPC was proposed to be one of the defects responsible for poor/blurred medical imaging of certain parts of the body. The LPC results obtain in this mathematical experimentation was found to be -56 dB and 6.7 dB. Theoretically, the LPC can be resolved mathematically by incorporating the molecular boundaries of the tissues. Practically, LPC can be resolved by introducing a detachable RF strip detector to synchronise-particles across different molecular boundaries and prevent patients from excess exposure to RF radiation.

Iterative correction of RF envelope distortion with GRATER-measured waveforms.

To develop and evaluate a method for RF envelope correction without extra hardware or synchronization. Transmitted RF waveforms are measured through a simple pulse sequence called the gradient reversal approach to evaluate RF (GRATER). The measured RF waveforms are used to compute predistorted RF waveforms. This process is repeated until a stopping criterion is met, for example, based on pulse performance or a maximum number of iterations. Excitation profiles and simultaneous multi-slice (SMS) image quality are compared before and after RF predistortion. The proposed method improved the accuracy of multiband RF pulses, reducing normalized root mean squared error (NRMSE) by >12-fold, and reducing spurious side-lobe excitation by >6-fold. The reduction in unwanted side-lobe signal is demonstrated using SMS bSSFP imaging at 3T in phantoms and in the heart. Iterative GRATER-based predistortion is a practical, hardware-free way to boost performance of short duration, low flip angle RF pulses, such as those used in SMS bSSFP imaging. Because of its efficiency, this technique could be included as part of an initial scan setup or for use with subsequent scans.

Eliminating susceptibility induced hyperintensities in T1w MPRAGE brain images at 7 T

IntroductionAt ultrahigh field, local susceptibility induced hyperintensities are pronounced in brain areas close to air-tissue boundaries in the inferior frontal lobe and temporal lobes on T1w MPRAGE images. Resulting from incomplete inversion, these artefacts can introduce biases in brain volumetry and erroneously suggest the existence of local tissular anomaly. We propose a straightforward approach to eliminate these artefacts by applying a shift (ΔfIR) to the center frequency of the adiabatic inversion pulse while widening the bandwidth of the latter by shortening the pulse duration (ΔtIR). MethodsAn MPRAGE sequence was customized allowing to change the duration (standard: 10,240 μs) and center frequency of the hyperbolic secant inversion RF pulse (IR). All measurements were performed on a 7 T whole body scanner (Siemens, Erlangen, Germany). 13 healthy volunteers (7 female and 6 male, average age (SD) = 38 ± 15 yrs) were recruited for the study, 3 of which were scanned for protocol optimization and the rest for performance evaluation. ΔB0 was mapped through the brain with a gradient echo sequence. The effects of ΔfIR and ΔtIR were studied separately and jointly to determine optimal parameter combinations to achieve the largest spatial extent of complete inversion throughout the brain. ResultsApplying a positive ΔfIR restored inversion efficiency in the inferior frontal and temporal lobes, but also introduced undesired hyperintensities in the anterior temporal lobes. Widening the bandwidth alone could also partially reduce hyperintensities in the frontal area but with a limited efficiency. By simultaneously applying a positive ΔfIR of 300 Hz and shortening ΔtIR by 40%, these artefacts were eliminated across the whole cerebrum. ConclusionA robust elimination of susceptibility induced hyperintensities near air-tissue boundaries in T1w MPRAGE 7 T brain images is demonstrated. This technique only requires limited MR sequence modifications.

Fat navigators and Moiré phase tracking comparison for motion estimation and retrospective correction.

To compare motion tracking by two modern methods (fat navigators [FatNavs] and Moiré phase tracking [MPT]) as well as their performance for retrospective correction of very high resolution acquisitions. A direct comparison of FatNavs and MPT motion parameters was performed for several deliberate motion patterns to estimate the agreement between methods. In addition, two different navigator resolution were applied. 0.5 mm isotropic MP2RAGE images with simultaneous MPT and FatNavs tracking were acquired in 9 cooperative subjects with no intentional motion. Retrospective motion corrections based on both tracking modalities were compared qualitatively and quantitatively. The FatNavs impact on quantitative T1 maps was also investigated. Both methods showed good agreement within a 0.3 mm/° margin in subjects that moved very little. Higher resolution FatNavs (2 mm) showed overall better agreement with MPT than 4 mm resolution ones, except for fast and large motion. The retrospective motion corrections based on MPT or FatNavs were at par in 33 cases out of 36, and visibly improved image quality compared to the uncorrected images. In separate fringe cases, both methods suffered from their respective potential shortcomings: unreliable marker attachment for MPT and poor temporal resolution for FatNavs. The magnetization transfer induced by the navigator RF pulses had a visible impact on the T1 values distribution, with a shift of the gray and white matter peaks of 12 ms at most. This work confirms both FatNavs and MPT as excellent retrospective motion correction methods for very high resolution imaging of cooperative subjects.

S15-09 SESSION 15: PLANNING/IMAGING - PART I BONE-SELECTIVE MRI AS A NONRADIATIVE ALTERNATIVE TO CT FOR CRANIOFACIAL IMAGING

Introduction: CT, in contrast to conventional MRI, allows evaluation of craniofacial skeletal pathologies with high-resolution 3D visualization of cortical bone structures. However, ionizing radiation is of concern, particularly for pediatric patients. This study evaluates the feasibility of producing 3D human skull renderings using a novel bone-selective MRI technique, with comparison to CT. Methods: Dual-radiofrequency (RF) pulse and ultrashort echo time (UTE) sequence: two RF pulses are alternately applied and two echoes are acquired for each RF. The first echo captures signals with very short lifetimes (bone), while the second starts after a longer delay. Four echoes are obtained in total: ECHO11 (RFshortTEshort), ECHO12 (RFshortTElong), ECHO21 (RFlongTEshort), and ECHO22 (RFlongTElong). Two images are constructed: one combining the bone-selective echoes (ECHO11, ECHO21), and one combining the longer echoes (ECHO12, ECHO22). Bone proton magnetization exhibits a substantial level of signal decay during RFlong, while soft tissue retains nearly the same level of signal intensities over all echoes. Thus, subtraction of the two images enhances bone contrast by suppressing soft tissue signals. A human cadaver skull and a healthy adult human subject were scanned at 3 T field strength (Siemens Prisma, Erlangen, Germany) within 6 minutes. Semi-automatic segmentation of bone voxels was performed using ITK-SNAP, leading to 3D renderings of the skulls. For comparison, CT scans were performed with 1 mm isotropic resolution. The biometric accuracy was assessed by measuring eight anatomic distances in CT- and MRI-based 3D renderings, using Mimics software (Materialise®, Ghent, Belgium). Lin’s Concordance Correlation test was applied to assess agreement between measurements from MR-based and CT based 3D skull renderings. Results: Compared to CT, the 3D rendered MR images depict most craniofacial features (e.g., zygomatic arch), although some voxels were erroneously included or excluded in the renderings. MR-based measurements differed by mean of -6mm from CT-based measurements. Lin’s Concordance Correlation Coefficients for MR vs CT were 0.996 (cadaver, 95% CI 0.991–1.001), and 0.991 (adult subject, 95% CI 0.980–1.001). Conclusion: The proposed dual-RF dual-echo 3D UTE imaging technique produces high-resolution bone-specific images within a clinically feasible imaging time, leading to clear visualization of craniofacial skeletal structures. Concordance coefficients of 0.991–0.996 suggest good reliability of the method compared to CT. The method is being optimized for more accurate 3D renderings and evaluated in pediatric patients.

Optimized analysis of sharpening characteristics of a compact RF pulse source based on a gyro-magnetic nonlinear transmission line for ultrawideband electromagnetic pulse application

We constructed a compact high-power RF pulse generator based on a gyro-magnetic nonlinear transmission line (GNLTL) to produce a high-voltage pulse with a sub-nanosecond rise time and a relatively high repetition rate, which shows great potential for application in the high-power ultrawideband electromagnetic effect, etc. The influence of incident pulse parameters (rise time and voltage amplitude) and line length on the sharpening characteristics of the GNLTL were investigated experimentally to optimize the rising rate of the modulated pulse front. Based on the GNLTL equivalent circuit model consisting of an LC ladder network, the rise time, the voltage conversion coefficient and the rising rate properties of a modulated pulse were also numerically analyzed in a wider range. The results show that a > 90 kV RF pulse with a rise time of 350 ps and a repetition rate of 1 kHz in burst mode is produced by the GNLTL at an axial biasing magnetic field of 22 kA m−1 and a line length of 30 cm under the condition of a 70 kV incident pulse. Applying a faster and higher incident pulse is conducive to improving the sharpening effect of the GNLTL. Furthermore, within a certain range, increasing the line length of the GNLTL not only reduces the rise time, but increases the voltage conversion coefficient and the rising rate of a modulated pulse. Furthermore, considering the energy loss of ferrite rings, there is an optimal line length to obtain the fastest rising rate of a modulated pulse front edge.

Scattering matrix imaging pulse design for real-time respiration and cardiac motion monitoring.

PurposeThe scattering matrix (S‐matrix) of a parallel transmit (pTx) coil is sensitive to physiological motion but requires additional monitoring RF pulses to be measured. In this work, we present and evaluate pTx RF pulse designs that simultaneously excite for imaging and measure the S‐matrix to generate real‐time motion signals without prolonging the image sequence.Theory and MethodsThree pTx waveforms for measuring the S‐matrix were identified and superimposed onto the imaging excitation RF pulses: (1) time division multiplexing, (2) frequency division multiplexing, and (3) code division multiplexing. These 3 methods were evaluated in healthy volunteers for scattering sensitivity and image artefacts. The S‐matrix and real‐time motion signals were calculated on the image calculation environment of the MR scanner. Prospective cardiac triggers were identified in early systole as a high rate of change of the cardiac motion signal. Monitoring accuracy was compared against electrocardiogram or the imaged diaphragm position.ResultsAll 3 monitoring approaches measure the S‐matrix during image excitation with quality correlated to input power. No image artefacts were observed for frequency multiplexing, and low energy artefacts were observed in the other methods. The accuracy of the achieved prospective cardiac gating was 15 ± 16 ms for breath hold and 24 ± 17 ms during free breathing. The diaphragm position prediction accuracy was 1.3 ± 0.9 mm. In all volunteers, good quality cine images were acquired for breath hold scans and dual gated CINEs were demonstrated.ConclusionThe S‐matrix can be measured during image excitation to generate real‐time cardiac and respiratory motion signals for prospective gating. No artefacts are introduced when frequency division multiplexing is used.

Magnetic resonance Spectrum simulator (MARSS), a novel software package for fast and computationally efficient basis set simulation.

The aim of this study was to develop a novel software platform for the simulation of magnetic resonance spin systems, capable of simulating a large number of spatial points (1283 ) for large in vivo spin systems (up to seven coupled spins) in a time frame of the order of a few minutes. The quantum mechanical density-matrix formalism is applied, a coherence pathway filter is utilized for handling unwanted coherence pathways, and the 1D projection method, which provides a substantial reduction in computation time for a large number of spatial points, is extended to include sequences of an arbitrary number of RF pulses. The novel software package, written in MATLAB, computes a basis set of 23 different metabolites (including the two anomers of glucose, seven coupled spins) with 1283 spatial points in 26min for a three-pulse experiment on a personal desktop computer. The simulated spectra are experimentally verified with data from both phantom and in vivo MEGA-sLASER experiments. Recommendations are provided regarding the various assumptions made when computing a basis set for in vivo MRS with respect to the number of spatial points simulated and the consideration of relaxation.

PENGARUH VARIASI TIME REPETITION (TR) TERHADAP KUALITAS CITRADAN INFORMASI CITRA PADA PEMERIKSAAN MRI LUMBALSEKUENS T2 FSE POTONGAN SAGITAL

Background : TR (Time Repetition) is an MRI parameter. TR controls the amount of longitudinal magnetization recovery before the next RF pulse. TR parameter is one of parameters that can affect image quality and image information. Long TR arrangements can increase SNR (Signal To Noise Ratio), CNR (Contrast To Noise Ratio) and provide better contrast, but increase the examination time. The purpose of this research is to know the effect of TR to image quality, difference of image information and TR optimal on MRI Lumbal sequence T2 FSE sagittal. Methods : The type of this research is quantitative research with experimental approach. The sample of research used is 6 probandus. TR variations used are TR 2000, 2500, 3000, 3500, 4000 with other parameters controlled. Aspects of image quality and image information values assessed in this study include Corpus Vertebrae, Discus Vertebrae, Cerebro Spinal Fluid (CSF), Medulla Spinalis, and Soft Tissue. Processing and analysis of image quality data is done by data normality test using shapiro-wilk, then test influence correlation pearson, linear regression and repeated anova. and for image information is done normality data test using shapiro-wilk, and friedman test. Result : The results of the test analysis showed that there was a TR effect on SNR image quality (p = 0.001) and CNR (p = 0.007) with (p <0.05) on all assessed organs. There is a difference in image information of MRI Lumbal T2 FSE sequence of sagittal pieces on Time Repetition variation with significance value (p <0.05). The optimal TR value on image quality is found on TR 4000 variation with mean of highest mean rank that is 42.82. While the optimal TR value on the image information is on the variation of TR 3000 with the highest mean mean rank is 3.25. TR 3000 is considered to produce better image information and is considered optimal on MRI Lumbal sequences of T2 FSE sagittal fragments compared to other TR variations with a duration of examination time of 2 minutes 30 seconds. Conclusion: Based on the results that good variation TR used on MRI Lumbal examination sequence T2 FSE sagital view is TR 3000 with best image qualit, informative image information and short scanning time.

Accelerated imaging with segmented 2D pulses using parallel imaging and virtual coils.

Large magnetic field inhomogeneity can be a significant cause of spatial flip-angle variation when using ordinary, limited-bandwidth RF pulses. Multidimensional RF pulses are particularly sensitive to inhomogeneity due to their extended pulse length, which decreases their bandwidth. Previously, it was shown that, by breaking a 2D pulse into multiple undersampled k-space segments, the excitation bandwidth can be increased at the expense of increased imaging time. The present study shows how this increased imaging time can be offset by undersampling acquisition k-space in a phase-encoded dimension that is in the direction of excitation segmentation. Data from each segment are viewed as originating from "virtual receive coils" rather than multiple physical coils. The undersampled data are reconstructed using parallel imaging techniques (e.g. as in GRAPPA). The method was tested in vivo with brain imaging at both 3 T and 4 T, and used in conjunction with a 32-channel head coil and conventional GRAPPA on the 3 T data. Relationships with existing techniques and future applications are discussed.

High-Power Piezoelectric Tuner Driver for Lorentz Force Compensation

Superconducting Radio Frequency (SRF) cavities are used in modern accelerators to efficiently accelerate particles. When cavity is supplied with pulsed RF field it undergoes a mechanical strain due to the Lorentz force. The resulting deformation causes dynamic detuning whose amplitude depends on mechanical properties of the cavity, RF pulse rate and their profile. This effect causes considerable loss of acceleration performance. Therefore, it is usually actively compensated, most commonly with fast piezoelectric actuators. MicroTCA.4 standard was developed to accommodate control and data acquisition electronic systems of large-scale physics applications. The paper presents a design of high-power amplifier implemented using the MicroTCA.4 technology. The design of the driver was optimized for driving large-capacitance piezo actuators. Several possible architectures of the driver are presented and compared, taking into consideration the power and cooling limitations of MicroTCA.4. The design of a two-channel piezo driver and its initial laboratory test results are also discussed.