Search

We develop a novel physics-adaptive machine-learning (ML) inversion scheme showing optimal generalization capabilities for field data applications. We apply the physics-driven deep-learning inversion to a massive helicopter-borne transient electromagnetic (TEM) field data set. The objective is the accurate modeling of the near surface for enhancing the exploration of low-relief structures in a sand covered desertic area. Enhanced generalization of a neural network (NN) or other ML techniques is obtained from automatic physics-based adaptive training using images of real-world data. Data reduction schemes, based on statistical sampling techniques, enable the use of small and fully informative data sets for accelerated training. Adaptive learning is implemented via an iterative physics-driven data augmentation strategy. A deterministic inversion is regularized by a penalty term built from the difference between the inverted models and the models predicted by ML. The resulting inverted models and calculated responses are used for augmenting the ML training base. The automated procedure converges rapidly, producing a trained network model, which captures the general background data statistics as well as the characteristics of field-specific geophysical data as embedded in the augmented samples. The high-resolution ML inversions with acceptable levels of data misfit are obtained for large volumes of the field geophysical data. We test several different learning models, such as artificial NNs, convolutional NN (U-Net), and the Gaussian process, which provide stable and comparable results with network-specific characteristics. The developed scheme is successfully applied to micro-TEM data providing accurate near-surface seismic corrections for the exploration of low-relief structures.

Myocardial pathologies resulting from SARS-CoV-2 infections are consistently rising with mounting case rates and reinfections; however, the precise global burden is largely unknown and will have an unprecedented impact. Understanding the mechanisms of COVID-19-mediated cardiac injury is essential toward the development of cardioprotective agents that are urgently needed. Assessing novel therapeutic strategies to tackle COVID-19 necessitates an animal model that recapitulates human disease. Here, we sought to compare SARS-CoV-2-infected animals with patients with COVID-19 to identify common mechanisms of cardiac injury. Two-month-old hamsters were infected with either the ancestral (D614) or Delta variant (B.1.617.2) of SARS-CoV-2 for 2 days, 7 days, and/or 14 days. We measured viral RNA and cytokine expression at the earlier time points to capture the initial stages of infection in the lung and heart. We assessed myocardial angiotensin-converting enzyme 2 (ACE2), the entry receptor for the SARS-CoV-2 virus, and cardioprotective enzyme, as well as markers for inflammatory cell infiltration in the hamster hearts at

Background Angiotensin-converting enzyme 2 (ACE2), the receptor for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), is highly expressed in the kidneys. Beyond serving as a crucial endogenous regulator of the renin–angiotensin system, ACE2 also possess a unique function to facilitate amino acid absorption. Our observational study sought to explore the relationship between urine ACE2 (uACE2) and renal outcomes in coronavirus disease 2019 (COVID-19). Methods In a cohort of 104 patients with COVID-19 without acute kidney injury (AKI), 43 patients with COVID-19-mediated AKI and 36 non-COVID-19 controls, we measured uACE2, urine tumour necrosis factor receptors I and II (uTNF-RI and uTNF-RII) and neutrophil gelatinase-associated lipocalin (uNGAL). We also assessed ACE2 staining in autopsy kidney samples and generated a propensity score–matched subgroup of patients to perform a targeted urine metabolomic study to describe the characteristic signature of COVID-19. Results uACE2 is increased in patients with COVID-19 and further increased in those that developed AKI. After adjusting uACE2 levels for age, sex and previous comorbidities, increased uACE2 was independently associated with a >3-fold higher risk of developing AKI [odds ratio 3.05 (95% confidence interval 1.23‒7.58), P = .017]. Increased uACE2 corresponded to a tubular loss of ACE2 in kidney sections and strongly correlated with uTNF-RI and uTNF-RII. Urine quantitative metabolome analysis revealed an increased excretion of essential amino acids in patients with COVID-19, including leucine, isoleucine, tryptophan and phenylalanine. Additionally, a strong correlation was observed between urine amino acids and uACE2. Conclusions Elevated uACE2 is related to AKI in patients with COVID-19. The loss of tubular ACE2 during SARS-CoV-2 infection demonstrates a potential link between aminoaciduria and proximal tubular injury.

Accurate near-surface characterization and velocity model building are important for a number of geotechnical and geophysical applications. 3D full waveform inversion can be used to generate a detailed velocity model, but must be provided with a good initial velocity model. A method for producing such an initial velocity model is explored in this paper. Assuming that the near-subsurface can be locally approximated by an effective flat-layered medium, a specialized form of acoustic full waveform inversion is proposed. The inversion is based on regularized least-squares in the Laplace-Fourier domain as a measure to mitigate the cycle-skipping problem. Forward modeling is carried out by solving a number of independent finite-difference problems in parallel, for a set of horizontal wavenumbers. The set is determined adaptively, using a pair of algorithms discussed in detail. A numerical implementation of the inverse Hankel transform, used to synthesize data in the frequency-offset domain, is also described. The forward modeling scheme is validated against analytic solutions of the Helmholtz equation, showing good agreement between the two. Full waveform inversion is tested by inverting a synthetic dataset consisting of flat layers with embedded velocity anomalies. Important features are recovered after inversion at a small number of complex frequencies, providing a detailed initial model for successive 3D full waveform inversion.

Sand-covered areas, such as desert environments, pose challenges to seismic imaging for resource exploration and monitoring. Aeolian sand dunes provide extreme variations of elastic parameters causing, among other effects, nonlinear velocity gradients proportional to compaction. Description and correction for sand velocities is typically performed through empirically derived sand curve functions. Shallow drilling with uphole traveltime measurements represents another established technique for characterizing the thickness of the sand layer. This defines a basal depositional surface across the investigated area to apply the sand correction for seismic processing operations. The accurate definition of the unconformable base of sand is crucial for the application of robust near-surface corrections. An airborne transient electromagnetic solution is proposed, where the acquisition instrumentation, acquisition parameters, and inversion strategies are tuned to the ultra-shallow depth of the target (2–20 m). The data inversion component makes extensive use of a recently introduced physics-driven DL inversion scheme. It enables boosted depth resolution in the resistivity model using all of the available data while achieving an acceptable data misfit to honor the propagation of the electromagnetic radiation. The imaged resistivity boundary provides an excellent match to the existing uphole data and indicates the presence of long-wavelength mismatch from the previously interpreted base of sand.

SARS-CoV-2 infection can lead to a variety of clinical manifestations. The occurrence of tongue swelling has recently reported in severe cases of COVID-19, and angioedema has suggested as the causative mechanism. Several factors, such as genetic predisposing factor and angiotensin-converting enzyme inhibitors (ACEI) therapies, have proposed to induce angioedema, especially as concerns patients requiring ICU treatments. Nevertheless, the question is still debated and other causes not yet recognized should be considered. Here we present a case of macroglossia occurred in a patient deceased for COVID-19 disease, who had no family history of angioedema and did not receive ACEI as antihypertensive drug. Histological and immune-histochemical analysis revealed tongue muscle atrophy with infiltrating macrophages suggesting repair mechanisms, as seen in nerve injury recovery. These new pathological findings may open new fields of study on the pathogenesis of SARS-CoV-2.

Herpes simplex virus type 2 (HSV-2) infection is the leading cause of genital lesions. Infrequently HSV-2 primary infection can spread and involve other tissue and organs, however in immunocompetent individuals extra-genital complications are rare findings. In this report we present a fatal case of fulminant myocarditis and terminal liver involvement, caused by HSV-2 infection, with a unique presentation. Diagnosis was made only post-mortem.

A 56-year-old woman presented to the emergency to be department with diarrhea, asthenia, cough, and dysgeusia. The patient had chronic obstructive pulmonary disease (COPD) and was found infected with coronavirus disease 2019 (COVID-19). On physical examination, a small basal cell carcinoma (BCC) lesion was identified on her scalp; however, following the administration of noninvasive ventilation, the appearance of both macroscopic and microscopic BCC worsened dramatically. Our findings point to positive pressure noninvasive ventilation used to treat COVID-19 associated with COPD as a possible causative agent for the progression of cutaneous BCC. (

Surface waves can be used to enhance the characterization of the shallow subsurface in desert environments. A high-resolution shear-wave velocity model is typically obtained by inverting dispersion curves, which correspond to different propagation modes of the surface waves. A common approach to estimate the dispersion curves is to manually pick the magnitude maxima from the frequency-phase velocity spectra of the seismic data. This approach is inefficient, time consuming, highly subjective, and not feasible for large surveys. Automatic picking of dispersion curves has become a topic of interest recently in the oil and gas research community, where many of the developed algorithms were inherited from the fields of image processing and machine learning. By exploring in the area of unsupervised learning, we recently derived an algorithm and workflow for fully automatic picking of surface-wave dispersion curves by employing a density-based spatial clustering technique. Our approach has been tested on the SEG Advanced Modeling Corporation Arid model synthetic data set and a field data set acquired in a desert environment. The results of the synthetic tests show that the estimated dispersion curves match the true dispersion curves with high accuracy, and they can be inverted for shear-wave velocities, successfully recovering the shallow near-surface features. The application of the method to field data provides high-resolution geology-consistent shear-wave velocity information that can be converted into a compressional-wave velocity model in agreement with uphole observations.

Machine learning, and specifically deep-learning (DL) techniques applied to geophysical inverse problems, is an attractive subject, which has promising potential and, at the same time, presents some challenges in practical implementation. Some obstacles relate to scarce knowledge of the searched geologic structures, a problem that can limit the interpretability and generalizability of the trained DL networks when applied to independent scenarios in real applications. Commonly used (physics-driven) least-squares optimization methods are very efficient local optimization techniques but require good starting models close to the correct solution to avoid local minima. We have developed a hybrid workflow that combines both approaches in a coupled physics-driven/DL inversion scheme. We exploit the benefits and characteristics of both inversion techniques to converge to solutions that typically outperform individual inversion results and bring the solution closer to the global minimum of a nonconvex inverse problem. The completely data-driven and self-feeding procedure relies on a coupling mechanism between the two inversion schemes taking the form of penalty functions applied to the model term. Predictions from the DL network are used to constrain the least-squares inversion, whereas the feedback loop from inversion to the DL scheme consists of the network retraining with partial results obtained from inversion. The self-feeding process tends to converge to a common agreeable solution, which is the result of two independent schemes with different mathematical formalisms and different objective functions on the data and model misfit. We determine that the hybrid procedure is converging to robust and high-resolution resistivity models when applied to the inversion of the synthetic and field transient electromagnetic data. Finally, we speculate that the procedure may be adopted to recast the way we solve inverse problems in several different disciplines.

It is not yet entirely clear what is the relevance of skin symptoms and what clinical implications are related to their appearance in COVID‐19 patients. We describe two cases of COVID‐19‐associated pneumonia, which presented skin manifestations in advanced stage of illness, when nasopharyngeal swabs became negative for SARS‐CoV‐2. The first case presented erythematous, maculopapular lesions; the second developed petechial, vesicular and blood‐encrusted lesions on the limbs. Histopathology documented perivascular lymphocytic infiltrates, with prevalent CD4+ T‐cells in both patients. The research of SARS‐CoV‐2 in tissues with real time RT‐PCR was negative. Basal keratinocytes displayed C4d deposits in one case, who developed laboratory signs indicative of a procoagulative condition at the same time as the skin rash. Skin manifestations during SARS‐CoV‐2 infection seem to be clinically relevant and further studies are necessary to assess if they are linked to systemic complications, lack of viral clearance or cascades of immune responses induced by the virus, even in patients affected by mild pneumonia.

Land seismic velocity modeling is a difficult task largely related to the description of the near-surface complexities. Full-waveform inversion (FWI) is the method of choice for achieving high-resolution velocity mapping, but its application to land seismic data faces difficulties related to the complex physics, unknown and spatially varying source signatures, and low signal-to-noise ratio (S/N) in the data. Large parameter variations occur in the near surface at various scales causing severe kinematic and dynamic distortions of the recorded wavefield. Some of the parameters can be incorporated in the inversion model, whereas others, due to subresolution dimensions or unmodeled physics, need to be corrected through data preconditioning; a topic not well described for land data FWI applications. We have developed novel algorithms and workflows for surface-consistent data preconditioning using the transmitted portion of the wavefield, S/N enhancement by generation of common midpoint (CMP)-based virtual supershot gathers (VSGs), and robust 1.5D Laplace-Fourier full-waveform inversion (LF-FWI). Our surface-consistent scheme solves residual kinematic corrections and amplitude anomalies via scalar compensation or deconvolution of the near-surface response. S/N enhancement is obtained through the statistical evaluation of volumetric prestack responses at the CMP position, or VSGs. These are inverted via a novel 1.5D acoustic LF-FWI scheme using the Helmholtz wave equation and Hankel domain forward modeling. Inversion is performed with nonlinear conjugate gradients. The method is applied to a complex structure-controlled wadi area exhibiting faults, dissolution, collapse, and subsidence in which the high-resolution FWI velocity modeling helps clarify the geologic interpretation. The developed algorithms and automated workflows provide an effective solution for massive FWI of land seismic data that can be embedded in typical near-surface velocity analysis procedures.

The repeated maternal separation (RMS) is a useful experimental model useful in rodents to study the long-term influence of early-life stress on brain neurophysiology. We here investigated the influence of RMS exposure on hippocampal inhibitory and excitatory synaptic transmission, long-term synaptic plasticity and the related potential alterations in learning and memory performance in adult male and female C57Bl/6J mice. Mice were separated daily from their dam for 360 min, from postnatal day 2 (PND2) to PND17, and experiments were performed at PND 60. Patch-clamp recordings in hippocampal CA1 pyramidal neurons revealed a significant enhancement of GABAergic miniature IPSC (mIPSC) frequency, and a decrease in the amplitude of glutamatergic mEPSCs in male mice exposed to RMS. Only a slight but significant reduction in the amplitude of GABAergic mIPSCs was observed in females exposed to RMS compared to the relative controls. A marked increase in long-term depression (LTD) at CA3-CA1 glutamatergic synapses and in the response to the CB1r agonist win55,212 were detected in RMS male, but not female mice. An impaired spatial memory and a reduced preference for novelty was observed in males exposed to RMS but not in females. A single injection of β-ethynyl estradiol at PND2, prevented the changes observed in RMS male mice, suggesting that estrogens may play a protective role early in life against the exposure to stressful conditions. Our findings strengthen the idea of a sex-dependent influence of RMS on long-lasting modifications in synaptic transmission, effects that may be relevant for cognitive performance.

Executive abilities are frequently impaired in patients with chronic obstructive pulmonary disease (COPD). We aimed at investigating the association between trail making test (TMT) and survival. Our series consisted of 68 stable COPD outpatients followed-up every 6 months for 52.6 ± 27.6 months. Enrolled patients underwent a baseline comprehensive neuropsychological assessment, including mini-mental state exam, attentional matrices, digit span, Rey auditory verbal learning, Rey-Osterrieth complex figure, copy drawing, tokens test, verbal fluency, category fluency, frontal assessment battery, Raven's progressive matrices, TMT-A, -B and -B-A. The association between neuropsychological deficits and overall mortality was investigated by Cox regression. During follow-up period, 41 patients (60.3%) died. After adjusting for potential confounders, TMT-B was significantly associated with mortality (HR = 2.42, 95% CI = 1.10-5.31), along with age (HR = 1.06, 95% CI = 1.0-1.13), overall comorbidity (HR = 1.29, 95% CI = 1.02-1.62) and use of noninvasive ventilation (HR = 2.16, 95% CI = 1.05-4.45). Defective TMT-B may be associated with long-term mortality in patients with stable COPD.

The sudden worldwide outbreak of Coronavirus Disease 2019 (COVID-19) has certainly provided new challenges in the management of acute ischaemic stroke, and the risk-benefit ratio of intravenous thrombolysis in COVID-19 positive patients is not well known. We describe four COVID-19 patients treated with intravenous thrombolysis for acute ischaemic stroke. Although rt-PA administration is the main therapeutic strategy, our patients experienced unpredictable complications and showed atypical features: the overall mortality was very high. In conclusion, in this article, we provide information about these cases and discuss the possible explanation behind this trend.

Purpose The aim of our study was to assess respiratory function at the time of clinical recovery and 6 weeks after discharge in patients surviving to COVID-19 pneumonia. Methods Our case series consisted of 13 patients with COVID-19 pneumonia. Results At the time of clinical recovery, FEV1 (2.07 ± 0.72 L) and FVC (2.25 ± 0.86 L) were lower compared to lower limit of normality (LLN) values (2.56 ± 0.53 L, p = 0.004, and 3.31 ± 0.65 L, p < 0.001, respectively), while FEV1/FVC (0.94 ± 0.07) was higher compared to upper limit of normality (ULN) values (0.89 ± 0.01, p = 0.029). After 6 weeks pulmonary function improved but FVC was still lower than ULN (2.87 ± 0.81, p = 0.014). Conclusion These findings suggest that COVID-19 pneumonia may result in clinically relevant alterations in pulmonary function tests, with a mainly restrictive pattern. Electronic supplementary material The online version of this article (10.1007/s15010-020-01474-9) contains supplementary material, which is available to authorized users.

We developed a novel surface-consistent processing framework that applies to transmitted wavefields (e.g., diving waves/refracted waves). Automatic velocity estimation from traveltime and full-waveform inversion, residual time shift corrections, and near-surface amplitude deconvolution are performed on the raw seismic gathers with minimal data preconditioning. In this implementation of the transmission-based surface-consistent analysis, we introduce a complete prestack and preprocessing workflow to fully analyze the near surface and enable highly specialized processes for what concerns land data imaging and reservoir characterization. Billions of traveltimes and traces can be analyzed automatically, providing enormous benefits in the execution time and in the accuracy and robustness of obtained results. The new paradigm for near-surface analysis opens the door to a better utilization of land seismic data for reservoir imaging and characterization.

Accurate velocity models for the near surface and overburden are needed for seismic processing and reliable depth imaging. Seismic with multiphysics data, well logs, and geology information need to be quantitatively integrated to obtain high-resolution velocity models. We detail our development and application of the joint wavefield inversion software platform, which enables flexible algorithmic schemes for the integration of multiparameter data and constraints. Inversion is performed in cascade or simultaneously using a variety of input data to constrain the velocity field reconstruction at multiple scales. Coupling mechanisms based on structure similarity together with rock-physics relations are optimally combined to boost resolution and enhance accuracy of the inverted velocity models. Ill-posed inversion problems are then solved using extensive geologic and rock-physics regularization instead of relying on smoothness constraints alone. We detail workflows and algorithms to guide the application of multiparameter joint inversion for velocity model building whether the input data are seismic traveltimes, electromagnetics (time/frequency domains), gravity, and/or surface waves. Extensive applications of multiparameter joint inversion are presented for a variety of complex geologic scenarios in which various multiparameter coupling strategies are illustrated. Robust velocity modeling and enhanced seismic imaging in time and depth domains are obtained as a result, proving the importance of multiphysics integration for reliable earth model parameter estimation.

We have developed a new framework for performing surface-consistent amplitude balancing and deconvolution of the near-surface attenuation response. Both approaches rely on the early arrival waveform of a seismic recording, which corresponds to the refracted or, more generally speaking, to the transmitted energy from a seismic source. The method adapts standard surface-consistent amplitude compensation and deconvolution to the domain of refracted/transmitted waves. A sorting domain specific for refracted energy is extended to the analysis of amplitude ratios of each trace versus a reference average trace to identify amplitude residuals that are inverted for surface consistency. The residual values are either calculated as a single scalar value for each trace or as a function of frequency to build a surface-consistent deconvolution operator. The derived operators are then applied to the data to obtain scalar amplitude balancing or amplitude balancing with spectral shaping. The derivation of the operators around the transmitted early arrival waveforms allows for deterministically decoupling the near-surface attenuation response from the remaining seismic data. The developed method is fully automatic and does not require preprocessing of the data. As such, it qualifies as a standard preprocessing tool to be applied at the early stages of seismic processing. Applications of the developed method are provided for a case in a complex, structure-controlled wadi, for a seismic time-lapse [Formula: see text] land monitoring case, and for an exploration area with high dunes and sabkhas producing large frequency-dependent anomalous amplitude responses. The new development provides an effective tool to enable better reservoir characterization and monitoring with land seismic data.

Reservoir characterization and monitoring represent some of the most ambitious goals for geophysical methods. Several challenges are involved, including sensitivity to the parameter changes and resolution of the obtained results. Electromagnetic (EM) methods are attractive for reservoir applications due to the high sensitivity of the resistivity parameter to oil/water saturations. Crosswell EM and surface-to-borehole EM provide opportunities for reservoir monitoring. The EM inverse problem, however, is highly nonconvex and ill-posed so as to necessitate significant preconditioning in the form of a priori information and regularization that impact resolution. We explore the use of machine-learning (ML) techniques in the form of deep-learning neural networks for implementing EM-based reservoir monitoring coupled with a dynamic fluid flow simulator. A crosswell acquisition setup is modeled in the framework of a realistic water-alternating-gas reservoir simulation scenario for enhanced oil recovery. Several reservoir saturation instances are generated and converted into resistivity, and corresponding crosswell EM data are generated using an electric source and a multicomponent (electric-magnetic) receiver assemblage. The U-Net deep-learning network is modified for the purpose of training and validation in which saturation models and the corresponding EM data are used. We also test the sensitivity of the deep-learning inversion to multiple EM components, noise in the data, generalization problems, and 3D reconstruction ability in which we use 3D convolutional neural network layers. In all cases, ML inversion proves to be robust with good resilience to increased noise levels. Prediction results indicate excellent reconstruction capabilities with resolution comparable to the reservoir models used by the simulator. Our results suggest that ML inversion through deep learning can become an efficient approach to data-driven and physics-constrained reservoir monitoring in which the sensitivity of EM-based techniques to fluid saturations can be fully exploited without compromising the resolution and accuracy of the results.