Fully asynchronous stochastic coordinate descent: a tight lower bound on the parallelism achieving linear speedup

Mathematical Programming - We seek tight bounds on the viable parallelism in asynchronous implementations of coordinate descent that achieves linear speedup. We focus on asynchronous coordinate...     

Transfer learning in deep neural network based under-sampled MR image reconstruction

In Magnetic Resonance Imaging (MRI), the success of deep learning-based under-sampled MR image reconstruction depends on: (i) size of the training dataset, (ii) generalization capabilities of the trained neural network. Whenever there is a mismatch between the training and testing data, there is a need to retrain the neural network from scratch with thousands of MR images obtained using the same protocol. This may not be possible in MRI as it is costly and time consuming to acquire data. In this research, a transfer learning approach i.e. end-to-end fine tuning is proposed for U-Net to address the data scarcity and generalization problems of deep learning-based MR image reconstruction. First the generalization capabilities of a pre-trained U-Net (initially trained on the human brain images of 1.5 T scanner) are assessed for: (a) MR images acquired from MRI scanners of different magnetic field strengths, (b) MR images of different anatomies and (c) MR images under-sampled by different acceleration factors. Later, end-to-end fine tuning of the pre-trained U-Net is proposed for the reconstruction of the above-mentioned MR images (i.e. (a), (b) and (c)). The results show successful reconstructions obtained from the proposed method as reflected by the Structural SIMilarity index, Root Mean Square Error, Peak Signal-to-Noise Ratio and central line profile of the reconstructed images.     

“Quad-band flexible magnesium zinc ferrite (MgZnFe2O4)-based double negative metamaterial for microwave applications”

This article presents vertically coupled, rectangular complementary split-ring resonator-shaped quad-band double-negative (DNG) metamaterial unit cells, that is, having both negative permittivity and permeability, which redirect negative refractive and also are not found in nature. The metamaterial is fabricated on magnesium zinc ferrite-based flexible microwave substrates, and the flexible substrates are chosen with two different concentrations of magnesium (Mg) denoted by Mg₃₀ and Mg₅₀ for 30% and 50% of Mg, which possess dielectric constants of 4.32 and 3.15 and loss tangents of 0.003 and 0.005, respectively. The proposed metamaterials are demonstrated by utilizing the CST microwave simulator, and their effective parameters are extracted according to the Nicolson-Ross-Wire method. With Mg_30, the prepared, flexible metamaterial shows measured resonances at 3.70 GHz, 7 GHz, 8.60 GHz, and 9.78 GHz, whereas with Mg₅₀ it shows the measured resonances at 4.10 GHz, 7.70 GHz, 9.33 GHz, and 10.62 GHz. Very good effective medium ratios (EMR) along with DNG properties are obtained, namely 6.5 and 5.85 for Mg₃₀ and Mg_50, respectively, with a physical dimension of 12.5 × 9.5 mm² for both of the unit cells_. Also, the electric field, magnetic field, and surface current distribution at different resonances and the polarization insensitivity at different polarization angles were observed. Thus, the designed new flexible substrate microwave materials based on DNG metamaterials are potential candidates for S-, C- and X-band applications, as well as for flexible microwave technologies.     

A finite difference method for the static limit analysis of masonry domes under seismic loads

Meccanica - The static limit analysis of axially symmetric masonry domes subject to pseudo-static seismic forces is addressed. The stress state in the dome is represented by the shell stress...