Efficient optimization with higher-order ising machines.

A prominent approach to solving combinatorial optimization problems on parallel hardware is Ising machines, i.e., hardware implementations of networks of interacting binary spin variables. Most Ising machines leverage second-order interactions although important classes of optimization problems, such as satisfiability problems, map more seamlessly to Ising networks with higher-order interactions. Here, we demonstrate that higher-order Ising machines can solve satisfiability problems more resource-efficiently in terms of the number of spin variables and their connections when compared to traditional second-order Ising machines. Further, our results show on a benchmark dataset of Boolean k-satisfiability problems that higher-order Ising machines implemented with coupled oscillators rapidly find solutions that are better than second-order Ising machines, thus, improving the current state-of-the-art for Ising machines.

Learning and Inference in Sparse Coding Models With Langevin Dynamics.

We describe a stochastic, dynamical system capable of inference and learning in a probabilistic latent variable model. The most challenging problem in such models-sampling the posterior distribution over latent variables-is proposed to be solved by harnessing natural sources of stochasticity inherent in electronic and neural systems. We demonstrate this idea for a sparse coding model by deriving a continuous-time equation for inferring its latent variables via Langevin dynamics. The model parameters are learned by simultaneously evolving according to another continuous-time equation, thus bypassing the need for digital accumulators or a global clock. Moreover, we show that Langevin dynamics lead to an efficient procedure for sampling from the posterior distribution in the L0 sparse regime, where latent variables are encouraged to be set to zero as opposed to having a small L1 norm. This allows the model to properly incorporate the notion of sparsity rather than having to resort to a relaxed version of sparsity to make optimization tractable. Simulations of the proposed dynamical system on both synthetic and natural image data sets demonstrate that the model is capable of probabilistically correct inference, enabling learning of the dictionary as well as parameters of the prior.

Bioactivity and Antibacterial Behaviors of Nanostructured Lithium-Doped Hydroxyapatite for Bone Scaffold Application.

The material for bone scaffold replacement should be biocompatible and antibacterial to prevent scaffold-associated infection. We biofunctionalized the hydroxyapatite (HA) properties by doping it with lithium (Li). The HA and 4 Li-doped HA (0.5, 1.0, 2.0, 4.0 wt.%) samples were investigated to find the most suitable Li content for both aspects. The synthesized nanoparticles, by the mechanical alloying method, were cold-pressed uniaxially and then sintered for 2 h at 1250 °C. Characterization using field-emission scanning electron microscopy (FE-SEM) revealed particle sizes in the range of 60 to 120 nm. The XRD analysis proved the formation of HA and Li-doped HA nanoparticles with crystal sizes ranging from 59 to 89 nm. The bioactivity of samples was investigated in simulated body fluid (SBF), and the growth of apatite formed on surfaces was evaluated using SEM and EDS. Cellular behavior was estimated by MG63 osteoblast-like cells. The results of apatite growth and cell analysis showed that 1.0 wt.% Li doping was optimal to maximize the bioactivity of HA. Antibacterial characteristics against Escherichia coli (E. coli) and Staphylococcus aureus (S. aureus) were performed by colony-forming unit (CFU) tests. The results showed that Li in the structure of HA increases its antibacterial properties. HA biofunctionalized by Li doping can be considered a suitable option for the fabrication of bone scaffolds due to its antibacterial and unique bioactivity properties.

A comparative study on solubility improvement of tetracycline and dexamethasone by poly(propylene imine) and polyamidoamine dendrimers: An insight into cytotoxicity and cell proliferation.

Many of new chemical discovered in pharmaceutical industry are hydrophobic compounds. Various techniques have been used to overcome solubility problems of hydrophobic drugs in aqueous media. In the meantime, dendrimers have been considered for sustainability, nanoscale size, high carry capacity, tunable terminal functional groups in terms of drug delivery and solubility. In this work, we have synthesized poly(propylene imine) (PPI) dendrimer up to fifth generation using reduction of nitrile groups after Michael addition and also, polyamidoamine (PAMAM) dendrimer up to fourth generation using Michael addition and amidation reactions. fourth and fifth generations of PPI dendrimer and fourth and third generations of PAMAM dendrimer in different concentrations were used to evaluate the solubility of two hydrophobic drugs (tetracycline and dexamethasone). Furthermore, cytotoxicity of dendrimers and dendrimers/drugs hybrids was studied. The results showed that with increasing concentrations and also the generation of dendrimers, the solubility of these two hydrophobic drugs was increased. Cytotoxicity study through MTT assay against Osteoblast-like cell line (MG-63 cells) showed that dendrimers were relatively cytotoxic where adding dexamethasone caused higher cytotoxicity. However, tetracycline showed no significant effect on cytotoxicity whereas prevented cell proliferation.

Design of optical neural networks with component imprecisions.

For the benefit of designing scalable, fault resistant optical neural networks (ONNs), we investigate the effects architectural designs have on the ONNs' robustness to imprecise components. We train two ONNs - one with a more tunable design (GridNet) and one with better fault tolerance (FFTNet) - to classify handwritten digits. When simulated without any imperfections, GridNet yields a better accuracy (∼98%) than FFTNet (∼95%). However, under a small amount of error in their photonic components, the more fault tolerant FFTNet overtakes GridNet. We further provide thorough quantitative and qualitative analyses of ONNs' sensitivity to varying levels and types of imprecisions. Our results offer guidelines for the principled design of fault-tolerant ONNs as well as a foundation for further research.

Effect of grafting ratio of poly(propylene imine) dendrimer onto gold nanoparticles on the properties of colloidal hybrids, their DOX loading and release behavior and cytotoxicity.

A facile method via "grafting to" approach was used to synthesize hybrid gold-dendrimer nanoparticles. To this end, gold nanoparticles (GNPs) were synthesized via Turkevich method and 5th-generation cystamine-cored poly(propylene imine) (PPI) dendrimer was synthesized by iterative Michael addition and hydrogenation reactions. To prepare hybrid nanoparticles, aqueous solution of dendrimer was poured into colloidal solution of GNPs to form gold-S interactions which resulted in hybrid gold-dendrimer nanoparticles. UV-VIS-NIR and Raman spectroscopies, dynamic light scattering (DLS), thermogravimetric analysis (TGA) and field-emission scanning electron microscopy (FE-SEM) were utilized to confirm the surface modification of GNPs by PPI dendrimer. Cytotoxicity study through MTT assay against human fibroblast (FBS) cells showed appropriate proliferation of cells in presence of hybrid nanoparticles whereas higher grafting ratio of dendrimers induced more toxicity due to existence of peripheral amine groups. Also, hybrid gold-dendrimer nanoparticles were used as DOX nanocarriers. Results showed that carriers did not release the drug at pH = 7.4 significantly while up to 92.8% of drug release was measured at pH = 5.3. Also, higher grafting ratio limited the drug release due to shielding effect of grafted dendrimers.

Composite electrospun nanofibers of reduced graphene oxide grafted with poly(3-dodecylthiophene) and poly(3-thiophene ethanol) and blended with polycaprolactone.

In this paper, an effective method was employed for preparation of nanofibers using conducting polymer-functionalized reduced graphene oxide (rGO). First, graphene oxide (GO) was obtained from graphite by Hommer method. GO was reduced to rGO by NaBH4 and covalently functionalized with a 3-thiophene acetic acid (TAA) by an esterification reaction to reach 3-thiophene acetic acid-functionalized reduced graphene oxide macromonomer (rGO-f-TAAM). Afterward, rGO-f-TAAM was copolymerized with 3-dodecylthiophene (3DDT) and 3-thiophene ethanol (3TEt) to yield rGO-f-TAA-co-PDDT (rGO-g-PDDT) and rGO-f-TAA-co-P3TEt (rGO-g-PTEt), which were confirmed by Fourier transform infrared spectra. The grafted materials depicted better electrochemical properties and superior solubilities in organic solvents compared to GO and rGO. The soluble rGO-g-PDDT and rGO-g-PTEt composites blended with polycaprolactone were fabricated by electrospinning, and then cytotoxicity, hydrophilicity, biodegradability and mechanical properties were investigated. The grafted rGO composites exhibited a good electroactivity behavior, mainly because of the enhanced electrochemical performance. The electrospun nanofibers underwent degradation about 7 wt% after 40 days, and the fabricated scaffolds were not able to induce cytotoxicity in mouse osteoblast MC3T3-E1 cells. The soluble conducting composites developed in this study are utilizable in the fabrication of nanofibers with tissue engineering application.

ViSAPy: a Python tool for biophysics-based generation of virtual spiking activity for evaluation of spike-sorting algorithms.

BACKGROUND: New, silicon-based multielectrodes comprising hundreds or more electrode contacts offer the possibility to record spike trains from thousands of neurons simultaneously. This potential cannot be realized unless accurate, reliable automated methods for spike sorting are developed, in turn requiring benchmarking data sets with known ground-truth spike times. NEW METHOD: We here present a general simulation tool for computing benchmarking data for evaluation of spike-sorting algorithms entitled ViSAPy (Virtual Spiking Activity in Python). The tool is based on a well-established biophysical forward-modeling scheme and is implemented as a Python package built on top of the neuronal simulator NEURON and the Python tool LFPy. RESULTS: ViSAPy allows for arbitrary combinations of multicompartmental neuron models and geometries of recording multielectrodes. Three example benchmarking data sets are generated, i.e., tetrode and polytrode data mimicking in vivo cortical recordings and microelectrode array (MEA) recordings of in vitro activity in salamander retinas. The synthesized example benchmarking data mimics salient features of typical experimental recordings, for example, spike waveforms depending on interspike interval. COMPARISON WITH EXISTING METHODS: ViSAPy goes beyond existing methods as it includes biologically realistic model noise, synaptic activation by recurrent spiking networks, finite-sized electrode contacts, and allows for inhomogeneous electrical conductivities. ViSAPy is optimized to allow for generation of long time series of benchmarking data, spanning minutes of biological time, by parallel execution on multi-core computers. CONCLUSION: ViSAPy is an open-ended tool as it can be generalized to produce benchmarking data or arbitrary recording-electrode geometries and with various levels of complexity.