Sensitivity Enhancement in Magnetic Resonance

NMR spectra acquired with experiments using frequency-sweeps such as the wide-band uniform-rate smooth truncation (WURST) spin-echo and Carr-Purcell-Meiboom-Gill (CPMG) sequences cannot be absorptively phased by using only conventional zeroth- and first-order phase correction. Implementation of phase correction up to the second-order is described for obtaining absorptive spectra, which have more desirable line shapes and noise properties than magnitude spectra. The relationship of the second-order phase to the parameters of frequency sweeps is derived. The second-order phasing in the frequency-domain is equivalent to a point spread in the time-domain signal. The application of second-order phase correction is demonstrated with a wideline ³⁵Cl CPMG spikelet spectrum.

With the development of machine learning in recent years, it is possible to glean much more information from an experimental data set to study matter. In this perspective, we discuss some state-of-the-art data-driven tools to analyze latent effects in data and explain their applicability in natural science, focusing on two recently introduced, physics-motivated computationally cheap tools—latent entropy and latent dimension. We exemplify their capabilities by applying them on several examples in the natural sciences and show that they reveal so far unobserved features such as, for example, a gradient in a magnetic measurement and a latent network of glymphatic channels from the mouse brain microscopy data. What sets these techniques apart is the relaxation of restrictive assumptions typical of many machine learning models and instead incorporating aspects that best fit the dynamical systems at hand.

Visual review of individual spectra in magnetic resonance spectroscopic imaging (MRSI) data benefits from the application of spectral smoothing; however, if this processing step is applied prior to spectral analysis this can impact the accuracy of the quantitation. This study aims to analyze the effect of spectral denoising and apodization smoothing on the quantitation of whole-brain MRSI data obtained at short TE.

Short-TE MRSI data obtained at 3 T were analyzed with no spectral smoothing, following (i) Gaussian apodization with values of 1, 2, 4, 6, and 8 Hz, and (ii) denoising using principal component analysis (dnPCA) with 3 different values for the number of retained principal components. The mean lobar white matter estimates for four metabolites, signal-to-noise ratio (SNR), spectral linewidth, and confidence intervals were compared to data reconstructed using no smoothing. Additionally, a voxel-wise comparison for N-acetylaspartate quantitation with different smoothing schemes was performed.

Significant pairwise differences were seen for all Gaussian smoothing methods as compared to no smoothing (p<0.001) in linewidth and metabolite estimates, whereas dnPCA methods showing no statistically significant differences in these measures. Confidence intervals decreased, and SNR increased with increasing levels of apodization smoothing or dnPCA denoising.

Mild Gaussian apodization (≤2 Hz at 3 T) can be applied with minimal (1%) errors in quantitation; however, smoothing values greater than that can significantly affect metabolite quantification. In contrast, mild to moderate dnPCA based denoising provides quantitative results that are consistent with the analysis of unsmoothed data and this method is recommended for spectral denoising.

Non-uniform sampling (NUS) provides a considerable reduction of measurement time especially for multi-dimensional experiments. This comes at the cost of additional signal processing steps to reconstruct the complete signal from the experimental data points. Despite being routinely employed in NMR for many experiments, EPR applications have not benefited from NUS due to the lack of a straightforward implementation to perform NUS in common commercial spectrometers. In this work we present a novel method to perform NUS HYSCORE experiments on commercial Bruker EPR spectrometers, along with a benchmark of modern reconstruction methods, and new processing software tools for NUS HYSCORE signals. All of this comes in the form of a free-software package: Hyscorean. Experimental NUS spectra are measured and processed with this package using different reconstruction methods and compared to their uniform sampled counterparts, thereby showcasing the method’s potential for EPR spectroscopy.

Experiments with fast repetition schemes significantly enhance the capabilities of modern NMR spectroscopy. Two schemes for heteronuclear correlation experiments that have been presented are the ASAP and the ALSOFAST method. The first method is Acceleration by Sharing Adjacent Polarization (ASAP) for samples at natural abundance isotope level. It was originally derived in the ASAP-HMQC and recently received renewed attention in the ASAP-HSQC. Sharing the polarization of active protons with the surrounding reservoir can result in seemingly instant polarization recovery and therefore enormous gains in sensitivity, but can also lead to a slight reduction of polarization and spectral intensity, depending on sample and setup. A second type of setup has been introduced with the so-called Alternate SOFAST (ALSOFAST-) HMQC and ALSOFAST-HSQC for natural abundance ${}^1$H,${}^{13}$C-correlation experiments and in the SOFAST-HMQC for ${}^1$H,${}^{15}$N-correlations. In these cases, the reservoir spins are only maintained through the pulse sequence without Hartmann-Hahn-type mixing. A model for the estimation of the available polarization in the fast repetition schemes could be a valuable tool for experimentalists and pulse sequence developers.

Starting from the well-known Ernst angle model, we derive in this article several mathematical models that describe the polarization over the course of ALSOFAST and ASAP type experiments. The models can be used to visualize the initial scans of an experiment and even more importantly, show the polarization and achievable signal intensity in the steady state of an experiment. In this way the two extreme applications of ASAP- and ALSOFAST-type acquisition schemes are covered: (i) acquisition using progressive excitation for experiments with few increments and shortest possible overall acquisition times and (ii) steady-state-type experiments with ultrahigh resolution and correspondingly large number of increments. The two resulting excitation strategies are applied to maximize SNR in different situations. To test the models, experimental data was obtained by special pulse sequences and examples are shown for different spin environments. The results show good agreement between theory and experiment.

A small flip-angle pulse direct polarization is the simplest method commonly used to quantify various compositions in many materials applications. This method sacrifices the sensitivity per scan in exchange for rapid repeating of data acquisition for signal accumulation. In addition, the resulting spectrum often encounters artifacts from background signals from probe components and/or from acoustic rings leading to a distorted baseline, especially in low-γ nuclei and wideline NMR. In this work, a multi-acquisition scheme is proposed to boost the sensitivity per scan and at the same time effectively suppress these artifacts. Here, an adiabatic inversion pulse is first applied in order to bring the magnetization from the +z to −z axis and then a small flip-angle pulse excitation is used before the data acquisition. Right after the first acquisition, the adiabatic inversion pulse is applied again to flip the magnetization back to the +z axis. The second data acquisition takes place after another small flip-angle pulse excitation. The difference between the two consecutive acquisitions cancels out any artifacts, while the wanted signals are accumulated. This acquisition process can be repeated many times before going into next scan. Therefore, by acquiring the signals multiple times in a single scan the sensitivity is improved. A mixture sample of flufenamic acid and 3,5-difluorobenzoic acid and a titanium silicate sample have been used to demonstrate the advantages of this newly proposed method.