We identify a few unique issues that are important for performing a nonlinear refraction traveltime tomography effectively. These include accuracy of traveltime and raypath calculation for a turning ray and physical information in a refraction traveltime curve. Consequently, we develop a shortest path raytracing method with an optimized node distribution that can accurately calculate refraction traveltimes and raypaths in any velocity model. We find that minimizing misfit of refraction traveltimes with the least-squares criterion does not account for the whole physical meaning of a refraction traveltime curve. We therefore pose a different nonlinear inverse problem that explicitly minimizes misfits of both traveltimes (integrated slownesses) and traveltime gradients (apparent slownesses). As a result, we enhance the resolution of the tomographic inversion as well as the convergence speed. We regularize our inverse problem with the Tikhonov method as opposed to applying ad hoc smoothing to keep the inversion stable. The use of the Tikhonov regularization avoids solving an ill-posed problem and allows us to invert an infinite number of unknowns. We apply this tomographic technique to image the shallow velocity structure at a coastal site near Boston, Massachusetts. The results are consistent with a local boring survey.

We have developed rapid 3-D do resistivity forward modeling and inversion algorithms that use conjugate gradient relaxation techniques. In the forward network modeling calculation, an incomplete Cholesky decomposition for preconditioning and sparse matrix routines combine to produce a fast and efficient algorithm (approximately 2 minutes CPU time on a Sun SPARCstation 2 for 50 x 50 x 20 blocks). The side and bottom boundary conditions are scaled impedance conditions that take into account the local current flow at the boundaries as a result of any configuration of current sources.For the inversion, conjugate gradient relaxation is used to solve the maximum likelihood inverse equations. Since conjugate gradient techniques only require the results of the sensitivity matrix A or its transpose A T multiplying a vector, we are able to bypass the actual computation of the sensitivity matrix and the inversion of ATA, thus greatly decreasing the time needed to do 3-D inversions.We demonstrate 3-D resistivity tomographic imaging using pole-pole resistivity data collected during an experiment for a leakage monitoring system near evaporation ponds at the Mojave Generating Station in Laughlin, Nevada.

Deep learning has achieved tremendous success in numerous industrial applications. As training a good model often needs massive high-quality data and computation resources, the learned models often have significant business values. However, these valuable deep models are exposed to a huge risk of infringements. For example, if the attacker has the full information of one target model including the network structure and weights, the model can be easily finetuned on new datasets. Even if the attacker can only access the output of the target model, he/she can still train another similar surrogate model by generating a large scale of input-output training pairs. How to protect the intellectual property of deep models is a very important but seriously under-researched problem. There are a few recent attempts at classification network protection only.In this paper, we propose the first model watermarking framework for protecting image processing models. To achieve this goal, we leverage the spatial invisible watermarking mechanism. Specifically, given a black-box target model, a unified and invisible watermark is hidden into its outputs, which can be regarded as a special task-agnostic barrier. In this way, when the attacker trains one surrogate model by using the input-output pairs of the target model, the hidden watermark will be learned and extracted afterward. To enable watermarks from binary bits to high-resolution images, both traditional and deep spatial invisible watermarking mechanism are considered. Experiments demonstrate the robustness of the proposed watermarking mechanism, which can resist surrogate models learned with different network structures and objective functions. Besides deep models, the proposed method is also easy to be extended to protect data and traditional image processing algorithms.

We have developed rapid 3-D do resistivity forward modeling and inversion algorithms that use conjugate gradient relaxation techniques. In the forward network modeling calculation, an incomplete Cholesky decomposition for preconditioning and sparse matrix routines combine to produce a fast and efficient algorithm (approximately 2 minutes CPU time on a Sun SPARCstation 2 for 50 x 50 x 20 blocks). The side and bottom boundary conditions are scaled impedance conditions that take into account the local current flow at the boundaries as a result of any configuration of current sources.For the inversion, conjugate gradient relaxation is used to solve the maximum likelihood inverse equations. Since conjugate gradient techniques only require the results of the sensitivity matrix A or its transpose A T multiplying a vector, we are able to bypass the actual computation of the sensitivity matrix and the inversion of ATA, thus greatly decreasing the time needed to do 3-D inversions.We demonstrate 3-D resistivity tomographic imaging using pole-pole resistivity data collected during an experiment for a leakage monitoring system near evaporation ponds at the Mojave Generating Station in Laughlin, Nevada.

Ferroptosis is an iron- and lipotoxicity-dependent form of regulated cell death (RCD). It is morphologically and biochemically distinct from characteristics of other cell death. This modality has been intensively investigated in recent years due to its involvement in a wide array of pathologies, including cancer, neurodegenerative diseases, and acute kidney injury. Dysregulation of ferroptosis has also been linked to various liver diseases and its modification may provide a hopeful and attractive therapeutic concept. Indeed, targeting ferroptosis may prevent the pathophysiological progression of several liver diseases, such as hemochromatosis, nonalcoholic steatohepatitis, and ethanol-induced liver injury. On the contrary, enhancing ferroptosis may promote sorafenib-induced ferroptosis and pave the way for combination therapy in hepatocellular carcinoma. Glutathione peroxidase 4 (GPx4) and system xc− have been identified as key players to mediate ferroptosis pathway. More recently diverse signaling pathways have also been observed. The connection between ferroptosis and other forms of RCD is intricate and compelling, where discoveries in this field advance our understanding of cell survival and fate. In this review, we summarize the central molecular machinery of ferroptosis, describe the role of ferroptosis in non-cancer hepatic disease conditions and discuss the potential to manipulate ferroptosis as a therapeutic strategy.

Abstract. We develop a rapid nonlinear travel time tomography method that simultaneously inverts refraction and reflection travel times on a regular velocity grid. For travel time and ray path calculations, we apply a wave front method employing graph theory. The first-arrival refraction travel times are calculated on the basis of cell velocities, and the later refraction and reflection travel times are computed using both cell velocities and given interfaces. We solve a regularized nonlinear inverse problem. A Laplacian operator is applied to regularize the model parameters (cell slownesses and reflector geometry) so that the inverse problem is valid for a continuum. The travel times are also regularized such that we invert travel time curves rather than travel time points. A conjugate gradient method is applied to minimize the nonlinear objective function. After obtaining a solution, we perform nonlinear Monte Carlo inversions for uncertainty analysis and compute the posterior model covariance. In numerical experiments, we demonstrate that combining the first arrival refraction travel times with later reflection travel times can better reconstruct the velocity field as well as the reflector geometry. This combination is particularly important for modeling crustal structures where large velocity variations occur in the upper crust. We apply this approach to model the crustal structure of the California Borderland using ocean bottom seismometer and land data collected during the Los Angeles Region Seismic Experiment along two marine survey lines. Details of our image include a high-velocity zone under the Catalina Ridge, but a smooth gradient zone between Catalina Ridge and San Clemente Ridge. The Moho depth is about 22 km with lateral variations.

No abstract

No abstract