Estimation of Orientation and Camera Parameters from Cryo-Electron Microscopy Images with Variational Autoencoders and Generative Adversarial Networks. (arXiv:1911.08121v1 [q-bio.QM])

Cryo-electron microscopy (cryo-EM) is capable of producing reconstructed 3D images of biomolecules at near-atomic resolution. As such, it represents one of the most promising imaging techniques in structural biology. However, raw cryo-EM images are only highly corrupted – noisy and band-pass filtered – 2D projections of the target 3D biomolecules. Reconstructing the 3D molecular shape…

Enhancing the Extraction of Interpretable Information for Ischemic Stroke Imaging from Deep Neural Networks. (arXiv:1911.08136v1 [eess.IV])

When artificial intelligence is used in the medical sector, interpretability is a crucial factor to consider. Diagnosis based on machine decision produced by a black-box neural network, sometimes lacking clear rationale, is unlikely to be clinically adopted for fear of potentially dire consequences arising from unexplained misdiagnosis. In this work, we implement Layer-wise Relevance Propagation…

High-resolution crystal structure of gelsolin domain 2 in complex with the physiological calcium ion. (arXiv:1911.08188v1 [q-bio.BM])

The second domain of gelsolin (G2) hosts mutations responsible for a hereditary form of amyloidosis. The active form of gelsolin is Ca2+-bound; it is also a dynamic protein, hence structural biologists often rely on the study of the isolated G2. However, the wild type G2 structure that have been used so far in comparative studies…

The structure of N184K amyloidogenic variant of gelsolin highlights the role of the H-bond network for protein stability and aggregation properties. (arXiv:1911.08194v1 [q-bio.BM])

Mutations in the gelsolin protein are responsible for a rare conformational disease known as AGel amyloidosis. Four of these mutations are hosted by the second domain of the protein (G2): D187N/Y, G167R and N184K. The impact of the latter has been so far evaluated only by studies on the isolated G2. Here we report the…

Three-dimensional cell culture model for hepatocytes opens a new avenue of real world research on liver. (arXiv:1911.08231v1 [q-bio.TO])

3-demensional (3D) culture model is a valuable in vitro tool to study liver biology, metabolism, organogenesis, tissue morphology, drug discovery and cell-based assays. Compelling evidence suggests that cells cultured in 3D model exhibit superior liver-specific functions over the conventional 2-dimentional (2D) culture in evaluating hepatobiliary drug disposition and drug-induced hepatotoxicity due to the in vivo-like…

Parametric Sparse Bayesian Dictionary Learning for Multiple Sources Localization with Propagation Parameters Uncertainty and Nonuniform Noise. (arXiv:1911.08021v1 [eess.SP])

Received signal strength (RSS) based source localization method is popular due to its simplicity and low cost. However, this method is highly dependent on the propagation model which is not easy to be captured in practice. Moreover, most existing works only consider the single source and the identical measurement noise scenario, while in practice multiple…

Driver Identification Based on Vehicle Telematics Data using LSTM-Recurrent Neural Network. (arXiv:1911.08030v1 [cs.LG])

Despite advancements in vehicle security systems, over the last decade, auto-theft rates have increased, and cyber-security attacks on internet-connected and autonomous vehicles are becoming a new threat. In this paper, a deep learning model is proposed, which can identify drivers from their driving behaviors based on vehicle telematics data. The proposed Long-Short-Term-Memory (LSTM) model predicts…

Universal digital filtering for denoising volumetric retinal OCT and OCT angiography in 3D shearlet domain. (arXiv:1911.08060v1 [eess.IV])

Retinal optical coherence tomography (OCT) and OCT angiography (OCTA) suffer from the degeneration of image quality due to speckle noise and bulk-motion noise, respectively. Because the cross-sectional retina has distinctive features in OCT and OCTA B-scans, existing digital filters that can denoise OCT efficiently are unable to handle the bulk-motion noise in OCTA. In this…

Optimal V2G and Route Scheduling of Mobile Energy Storage Devices Using a Linear Transit Model to Reduce Electricity and Transportation Energy Losses. (arXiv:1911.08073v1 [eess.SY])

Mobile energy storage devices (MESDs) operate as medium- or large-sized batteries that can be loaded onto electric trucks and connected to charging stations to provide various ancillary services for distribution grids. This paper proposes a new strategy for MESD operation, in which their power outputs and paths are co-optimally scheduled to minimize the total energy…

Partial AUC optimization based deep speaker embeddings with class-center learning for text-independent speaker verification. (arXiv:1911.08077v1 [cs.LG])

Deep embedding based text-independent speaker verification has demonstrated superior performance to traditional methods in many challenging scenarios. Its loss functions can be generally categorized into two classes, i.e., verification and identification. The verification loss functions match the pipeline of speaker verification, but their implementations are difficult. Thus, most state-of-the-art deep embedding methods use the identification…