OP-FCNN: an optronic fully convolutional neural network for imaging through scattering media.

Imaging through scattering media is a classical inverse issue in computational imaging. In recent years, deep learning(DL) methods have excelled in speckle reconstruction by extracting the correlation of speckle patterns. However, high-performance DL-based speckle reconstruction also costs huge hardware computation and energy consumption. Here, we develop an opto-electronic DL method with low computation complexity for imaging through scattering media. We design the "end-to-end" optronic structure for speckle reconstruction, namely optronic fully convolutional neural network (OP-FCNN). In OP-FCNN, we utilize lens groups and spatial light modulators to implement the convolution, down/up-sampling, and skip connection in optics, which significantly reduces the computational complexity by two orders of magnitude, compared with the digital CNN. Moreover, the reconfigurable and scalable structure supports the OP-FCNN to further improve imaging performance and accommodate object datasets of varying complexity. We utilize MNIST handwritten digits, EMNIST handwritten letters, fashion MNIST, and MIT-CBCL-face datasets to validate the OP-FCNN imaging performance through random diffusers. Our OP-FCNN reveals a good balance between computational complexity and imaging performance. The average imaging performance on four datasets achieves 0.84, 0.91, 0.79, and 16.3dB for JI, PCC, SSIM, and PSNR, respectively. The OP-FCNN paves the way for all-optical systems in imaging through scattering media.

Dynamic Analysis of an Epidemic Model Considering Personal Alert on a Complex Network.

This paper proposes a SIQRS epidemic model with birth and death on a complex network, considering individual alertness. In particular, we investigate the influence of the individual behavior in the transmission of epidemics and derive the basic reproduction number depending on birth rate, death rate, alertness rate, quarantine rate. In addition, the stabilities of the disease-free equilibrium point and endemic equilibrium point are analyzed via stability theory. It is found that the emergence of individual behavior can influence the process of transmission of epidemics. Our results show that individual alertness rate is negatively correlated with basic reproduction number, while the impact of individual alertness on infectious factor is positively correlated with basic reproduction number. When the basic reproduction number is less than one, the system is stable and the disease is eventually eradicated. Nevertheless, there is an endemic equilibrium point under the condition that the basic reproduction number is more than one. Finally, numerical simulations are carried out to illustrate theoretical results.

Effects of Community Connectivity on the Spreading Process of Epidemics.

Community structure exists widely in real social networks. To investigate the effect of community structure on the spreading of infectious diseases, this paper proposes a community network model that considers both the connection rate and the number of connected edges. Based on the presented community network, a new SIRS transmission model is constructed via the mean-field theory. Furthermore, the basic reproduction number of the model is calculated via the next-generation matrix method. The results reveal that the connection rate and the number of connected edges of the community nodes play crucial roles in the spreading process of infectious diseases. Specifically, it is demonstrated that the basic reproduction number of the model decreases as the community strength increases. However, the density of infected individuals within the community increases as the community strength increases. For community networks with weak strength, infectious diseases are likely not to be eradicated and eventually will become endemic. Therefore, controlling the frequency and range of intercommunity contact will be an effective initiative to curb outbreaks of infectious diseases throughout the network. Our results can provide a theoretical basis for preventing and controlling the spreading of infectious diseases.

Training optronic convolutional neural networks on an optical system through backpropagation algorithms.

The development of optical neural networks greatly slows the urgent demand of searching for fast computing approaches to solve big data processing. However, most optical neural networks following electronic training and optical inferencing do not really take full advantage of optical computing to reduce computational burden. Take the extensively used optronic convolutional neural networks (OPCNN) as an example, the convolutional operations still require vast computational operations in training stages on the computer. To address this issue, this study proposes the in-situ training algorithm to train the networks directly in optics. We derive the backpropagation algorithms of OPCNN hence the complicated gradient calculation in backward propagating processes can be obtained through optical computing. Both forward propagation and backward propagation are all executed on the same optical system. Furthermore, we successfully realize the introduction of optical nonlinearity in networks through utilizing photorefractive crystal SBN:60 and we also derive the corresponding backpropagation algorithm. The numerical simulation results of classification performance on several datasets validates the feasibility of the proposed algorithms. Through in-situ training, the reduction in performance resulting from the inconsistency of the plantform between training and inferencing stages can be eliminated completely. For example, we demonstrate that by using the optical training approach, OPCNN is capable of gaining a strong robustness under several misalignmed situations, which enhances the practicability of OPCNN and greatly expands its application range.

Analysis of Vulnerability on Weighted Power Networks under Line Breakdowns.

Vulnerability is a major concern for power networks. Malicious attacks have the potential to trigger cascading failures and large blackouts. The robustness of power networks against line failure has been of interest in the past several years. However, this scenario cannot cover weighted situations in the real world. This paper investigates the vulnerability of weighted power networks. Firstly, we propose a more practical capacity model to investigate the cascading failure of weighted power networks under different attack strategies. Results show that the smaller threshold of the capacity parameter can enhance the vulnerability of weighted power networks. Furthermore, a weighted electrical cyber-physical interdependent network is developed to study the vulnerability and failure dynamics of the entire power network. We perform simulations in the IEEE 118 Bus case to evaluate the vulnerability under various coupling schemes and different attack strategies. Simulation results show that heavier loads increase the likelihood of blackouts and that different coupling strategies play a crucial role in the cascading failure performance.

Optronic convolutional neural networks of multi-layers with different functions executed in optics for image classification.

Although deeper convolutional neural networks (CNNs) generally obtain better performance on classification tasks, they incur higher computation costs. To address this problem, this study proposes the optronic convolutional neural network (OPCNN) in which all computation operations are executed in optics, and data transmission and control are executed in electronics. In OPCNN, we implement convolutional layers with multi input images by the lenslet 4f system, downsampling layers by optical-strided convolution and obtaining nonlinear activation by adjusting the camera's curve and fully connected layers by optical dot product. The OPCNN demonstrates good performance on the classification tasks in simulations and experiments and achieves better performance than other current optical convolutional neural networks by comparison due to the more complex architecture. The scalability of OPCNN contributes to building deeper networks when facing complicated datasets.

3-Arylamino-quinoxaline-2-carboxamides inhibit the PI3K/Akt/mTOR signaling pathways to activate P53 and induce apoptosis.

Thirty-eight new 3-arylaminoquinoxaline-2-carboxamide derivatives were in silico designed, synthesized and their cytotoxicity against five human cancer cell lines and one normal cells WI-38 were evaluated. Molecular mechanism studies indicated that N-(3-Aminopropyl)-3-(4-chlorophenyl) amino-quinoxaline-2-carboxamide (6be), the compound with the most potent anti-proliferation can inhibit the PI3K-Akt-mTOR pathway via down regulating the levels of PI3K, Akt, p-Akt, p-mTOR and simultaneously inhibit the phosphorylation of Thr308 and Ser473 residues in Akt kinase to servers as a dual inhibitor. Further investigation revealed that 6be activate the P53 signal pathway, modulated the downstream target gene of Akt kinase such p21, p27, Bax and Bcl-2, caused the fluctuation of intracellular ROS, Ca2+ and mitochondrial membrane potential to induce cell cycle arrest and apoptosis in MGC-803 cells. 6be also display moderate anti-tumor activity in vivo while displaying no obvious adverse signs during the drug administration. The results suggest that 3-arylaminoquinoxaline-2-carboxamide derivatives might server as new scaffold for development of PI3K-Akt-mTOR inhibitor.

Temperature retrieval from Rayleigh-Brillouin scattering profiles measured in air.

In order to investigate the performance of two different algorithms for retrieving temperature from Rayleigh-Brillouin (RB) line shapes, RB scattering measurements have been performed in air at a wavelength of 403 nm, for a temperature range from 257 K to 330 K, and atmospherically relevant pressures from 871 hPa to 1013 hPa. One algorithm, based on the Tenti S6 line shape model, shows very good accordance with the reference temperature. In particular, the absolute difference is always less than 2 K. A linear correlation yields a slope of 1.01 ± 0.02 and thus clearly demonstrates the reliability of the retrieval procedure. The second algorithm, based on an analytical line shape model, shows larger discrepancies of up to 9.9 K and is thus not useful at its present stage. The possible reasons for these discrepancies and improvements of the analytical model are discussed. The obtained outcomes are additionally verified with previously performed RB measurements in air, at 366 nm, temperatures from 255 K to 338 K and pressures from 643 hPa to 826 hPa [Appl. Opt. 52, 4640 (2013)]. The presented results are of relevance for future lidar studies that might utilize RB scattering for retrieving atmospheric temperature profiles with high accuracy.

Analysis of Rayleigh-Brillouin spectral profiles and Brillouin shifts in nitrogen gas and air.

On the basis of experimental Rayleigh-Brillouin scattering data in gaseous nitrogen and air, simulations are performed to describe the observed frequency profiles in analytical form. The experimental data pertain to a λ = 366 nm scattering wavelength, a 90° scattering angle, pressures of 1 and 3 bar, and temperatures in the range 250 - 340 K. Two different models are used to represent the RB-profiles, to distinguish the RB-peaks, and to obtain the Brillouin shift associated with the acoustic waves generated in a gaseous medium. Calculations in the framework of V3 and G3 models, exhibiting composite profiles of three distinct peaks of Voigt or Gaussian functions, are compared to observation. Fitting results show that the V3 model yields an improvement over the G3 model. This mathematical model provides an even better representation of the observed profiles than the Tenti S6 model, which is considered to be the optimum representation in terms of physical parameters. For the derivation of Brillouin shifts, both models perform well at high gas pressure, while at lower pressures, the V3 model yields a higher accuracy than the G3 model.

A systematic study of Rayleigh-Brillouin scattering in air, N2, and O2 gases.

Spontaneous Rayleigh-Brillouin scattering experiments in air, N2, and O2 have been performed for a wide range of temperatures and pressures at a wavelength of 403 nm and at a 90° scattering angle. Measurements of the Rayleigh-Brillouin spectral scattering profile were conducted at high signal-to-noise ratio for all three species, yielding high-quality spectra unambiguously showing the small differences between scattering in air, and its constituents N2 and O2. Comparison of the experimental spectra with calculations using the Tenti S6 model, developed in the 1970s based on linearized kinetic equations for molecular gases, demonstrates that this model is valid to high accuracy for N2 and O2, as well as for air. After previous measurements performed at 366 nm, the Tenti S6 model is here verified for a second wavelength of 403 nm, and for the pressure-temperature parameter space covered in the present study (250-340 K and 0.6-3 bars). In the application of the Tenti S6 model, based on the transport coefficients of the gases, such as thermal conductivity κ, internal specific heat capacity c(int) and shear viscosity η, as well as their temperature dependencies taken as inputs, values for the more elusive bulk viscosity η(b) for the gases are derived by optimizing the model to the measurements. It is verified that the bulk viscosity parameters obtained from previous experiments at 366 nm are valid for wavelengths of 403 nm. Also for air, which is treated as a single-component gas with effective gas transport coefficients, the Tenti S6 treatment is validated for 403 nm as for the previously used wavelength of 366 nm, yielding an accurate model description of the scattering profiles for a range of temperatures and pressures, including those of relevance for atmospheric studies. It is concluded that the Tenti S6 model, further verified in the present study, is applicable to LIDAR applications for exploring the wind velocity and the temperature profile distributions of the Earth's atmosphere. Based on the present findings at 90° scattering and the determination of η(b) values, predictions can be made on the spectral profiles for a typical LIDAR backscatter geometry. These Tenti S6 predictions for Rayleigh-Brillouin scattering deviate by some 7% from purely Gaussian profiles at realistic sub-atmospheric pressures occurring at 3-5 km altitude in the Earth's atmosphere.

Temperature-dependent bulk viscosity of nitrogen gas determined from spontaneous Rayleigh-Brillouin scattering.

Values for the bulk viscosity η(b) of molecular nitrogen gas (N2) were derived from spontaneous Rayleigh-Brillouin scattering at ultraviolet wavelengths (λ=366.8 nm) and at a 90° scattering angle. Analysis of the scattering profiles yields values showing a linear increasing trend, ranging from η(b)=0.7×10(-5) to 2.0×10(-5) kg·m(-1)·s(-1) in the temperature interval from 255 to 340 K. The present values, pertaining to hypersound acoustics at frequencies in the gigahertz domain, are found to be in agreement with results from acoustic attenuation experiments in N2 performed at megahertz frequencies.

Rayleigh-Brillouin scattering profiles of air at different temperatures and pressures.

Rayleigh-Brillouin (RB) scattering profiles for air have been recorded for the temperature range from 255 to 340 K and the pressure range from 640 to 3300 mbar, covering the conditions relevant for the Earth's atmosphere and for planned atmospheric light detection and ranging (LIDAR) missions. The measurements performed at a wavelength of λ=366.8 nm detect spontaneous RB scattering at a 90° scattering angle from a sensitive intracavity setup, delivering scattering profiles at a 1% rms noise level or better. The experimental results have been compared to a kinetic line-shape model, the acclaimed Tenti S6 model, considered to be most appropriate for such conditions, under the assumption that air can be treated as an effective single-component gas with temperature-scaled values for the relevant macroscopic transport coefficients. The elusive transport coefficient, the bulk viscosity η(b), is effectively derived by a comparing the measurements to the model, yielding an increased trend from 1.0 to 2.5×10(-5) kg·m(-1)·s(-1) for the temperature interval. The calculated (Tenti S6) line shapes are consistent with experimental data at the level of 2%, meeting the requirements for the future RB-scattering LIDAR missions in the Earth's atmosphere. However, the systematic 2% deviation may imply that the model has a limit to describe the finest details of RB scattering in air. Finally, it is demonstrated that the RB scattering data in combination with the Tenti S6 model can be used to retrieve the actual gas temperatures.

Rotational diffusion and alignment of short gold nanorods in an external electric field.

We present measurements of the polarized extinction of gold nanorod suspensions exposed to an external electric field. By employing an amplitude modulated field in combination with lock-in detection we resolve changes in the optical density as low as 10(-6) in an integration time of 10 s. This sensitivity allows us to probe the partial alignment of small gold nanorods with an aspect ratio of 2.5 and a width ranging from 13 nm to 28 nm. The degree of orientation scales as the square of the electric field strength, as expected for an induced dipole moment in an external field. By varying the modulation frequency we measure the rotation diffusion constant of different samples, which are in excellent agreement with the calculated values for a short cylinder.

Regulation of homocysteine metabolism and methylation in human and mouse tissues.

Hyperhomocysteinemia is an independent risk factor for cardiovascular disease. Homocysteine (Hcy) metabolism involves multiple enzymes; however, tissue Hcy metabolism and its relevance to methylation remain unknown. Here, we established gene expression profiles of 8 Hcy metabolic and 12 methylation enzymes in 20 human and 19 mouse tissues through bioinformatic analysis using expression sequence tag clone counts in tissue cDNA libraries. We analyzed correlations between gene expression, Hcy, S-adenosylhomocysteine (SAH), and S-adenosylmethionine (SAM) levels, and SAM/SAH ratios in mouse tissues. Hcy metabolic and methylation enzymes were classified into two types. The expression of Type 1 enzymes positively correlated with tissue Hcy and SAH levels. These include cystathionine beta-synthase, cystathionine-gamma-lyase, paraxonase 1, 5,10-methylenetetrahydrofolate reductase, betaine:homocysteine methyltransferase, methionine adenosyltransferase, phosphatidylethanolamine N-methyltransferases and glycine N-methyltransferase. Type 2 enzyme expressions correlate with neither tissue Hcy nor SAH levels. These include SAH hydrolase, methionyl-tRNA synthase, 5-methyltetrahydrofolate:Hcy methyltransferase, S-adenosylmethionine decarboxylase, DNA methyltransferase 1/3a, isoprenylcysteine carboxyl methyltransferases, and histone-lysine N-methyltransferase. SAH is the only Hcy metabolite significantly correlated with Hcy levels and methylation enzyme expression. We established equations expressing combined effects of methylation enzymes on tissue SAH, SAM, and SAM/SAH ratios. Our study is the first to provide panoramic tissue gene expression profiles and mathematical models of tissue methylation regulation.

A Rayleigh-Brillouin scattering spectrometer for ultraviolet wavelengths.

A spectrometer for the measurement of spontaneous Rayleigh-Brillouin (RB) scattering line profiles at ultraviolet wavelengths from gas phase molecules has been developed, employing a high-power frequency-stabilized UV-laser with narrow bandwidth (2 MHz). The UV-light from a frequency-doubled titanium:sapphire laser is further amplified in an enhancement cavity, delivering a 5 W UV-beam propagating through the interaction region inside a scattering cell. The design of the RB-scattering cell allows for measurements at gas pressures in the range 0-4 bars and at stably controlled temperatures from -30 °C to 70 °C. A scannable Fabry-Perot analyzer with instrument resolution of 232 MHz probes the RB profiles. Measurements on N(2) and SF(6) gases demonstrate that the high signal-to-noise ratio is achievable with the instrument at the 1% level at the peak amplitude of the scattering profile.