Advancing automated pupillometry:a practical deep learning model utilizing infrared pupil images

AIM:To establish pupil diameter measurement algorithms based on infrared images that can be used in real-world clinical settings.METHODS:A total of 188 patients from outpatient clinic at He Eye Specialist Shenyang Hospital from Spetember to December 2022 were included,and 13470 infrared pupil images were collected for the study.All infrared images for pupil segmentation were labeled using the Labelme software.The computation of pupil diameter is divided into four steps:image pre-processing,pupil identification and localization,pupil segmentation,and diameter calculation.Two major models are used in the computation process:the modified YoloV3 and Deeplabv 3+models,which must be trained beforehand.RESULTS:The test dataset included 1348 infrared pupil images.On the test dataset,the modified YoloV3 model had a detection rate of 99.98% and an average precision(AP)of 0.80 for pupils.The DeeplabV3+model achieved a background intersection over union(IOU)of 99.23%,a pupil IOU of 93.81%,and a mean IOU of 96.52%.The pupil diameters in the test dataset ranged from 20 to 56 pixels,with a mean of 36.06±6.85 pixels.The absolute error in pupil diameters between predicted and actual values ranged from 0 to 7 pixels,with a mean absolute error(MAE)of 1.06±0.96 pixels.CONCLUSION:This study successfully demonstrates a robust infrared image-based pupil diameter measurement algorithm,proven to be highly accurate and reliable for clinical application.

复合材料J型加筋壁板制造技术研究

对J型加筋壁板共胶接技术在实际应用中所存在的主要问题进行了研究,着重介绍了R区芯材精确填充、筋条外型模成型、框架式插销垂直定位、共胶接真空袋封装等关键技术的研究情况,并对相关技术的应用进行了深入分析。研究结果表明,在精确计算的基础上,设计制造专用成型模具可实现R区填充芯材的精确制备,有利于提高加筋壁板胶接质量;选用筋条外型模成型工艺,并设计采用框架式插销垂直定位装置,可有效解决J型筋的表面成型质量、筋条尺寸精度、位置精度等问题;使用已硫化橡胶作为维形挡条、制备合适的内型面软模、优化辅助材料铺放方法,可减少表面质量问题,降低架桥、破袋风险。在应用了上述一系列制造技术后,成功制造了满足使用要求的J型加筋壁板复材零件,并初步实现该类型零件的批量化生产。

Human tracking in camera network with non-overlapping FOVs

An adaptive human tracking method across spatially separated surveillance cameras with non-overlapping fields of views （FOVs） is proposed. The method relies on the two cues of the human appearance model and spatio-temporal information between cameras. For the human appearance model, an HSV color histogram is extracted from different human body parts （head, torso, and legs）, then a weighted algorithm is used to compute the similarity distance of two people. Finally, a similarity sorting algorithm with two thresholds is exploited to find the correspondence. The spatio- temporal information is established in the learning phase and is updated incrementally according to the latest correspondence. The experimental results prove that the proposed human tracking method is effective without requiring camera calibration and it becomes more accurate over time as new observations are accumulated.

Release Model of Water-soluble Chitosan Nanoparticles for Protein Delivery

[Objective] The experiment aimed to explore release rule of water-soluble chitosan (WSC) in vitro. [Method]The bovine serum albumin(BSA) was taken as a model protein drug and some existing release models such as Kinetics model, Gompertz model, Weibull model, Higuchi model and Logistic model were used to fit the BSA release profile from WSC carriers. [Result] Except Higuchi model and Logistic model, other models could fit BSA release profile better. [Conclusion] Gompertz two-order kinetics model could fit the release of WSC nano-particles better and model parameters had practical physical meaning.

Integration of Lab model and EHOG for human appearance matching across disjoint camera views

The integration of the Lab model with the extended histogram of oriented gradients （EHOG） is proposed to improve the accuracy of human appearance matching across disjoint camera views under perturbations such as illumination changes and different viewing angles. For the Lab model that describes the global information of observations, a sorted nearest neighbor clustering method is proposed for color clustering and then a partitioned color matching method is used to calculate the color similarity between observations. The Bhattacharya distance is employed for the textural similarity calculation of the EHOG which describes the local information. The global information, which is robust to different viewing angles and scale changes, describes the observations well. Meanwhile, the use of local information, which is robust to illumination changes, can strengthen the discriminative ability of the method. The integration of global and local information improves the accuracy and robustness of the proposed matching approach. Experiments are carried out indoors, and the results show a high matching accuracy of the proposed method.

Determination of Melamine by Infrared Spectroscopy Based on Nonlinear Modeling

[Objective] The aim was to develop a nonlinear model of quantitative analysis of melamine content by infrared spectroscopy and provide theoretical basis for the nondestructive detection of melamine. [Method] According to dynamics,mathematical modeling and optimization theory,linear and nonlinear models were respectively set up by taking an absorption peak of 1 550 cm-1 as characteristic absorption peak. [Result] The correlation coefficient of nonlinear model was 0.922 7 and the recovery was 96%,which showed that the nonlinear model was more accurate than linearity model with correlation coefficient of 0.904 9 and recovery of 557%. [Conclusion] It is feasible to determine melamine content by using the nonlinear model quantitatively.

Analytical model for power dissipation in cell membranes in suspensions exposed to electric fields

Due to interaction among cells, it is too complex to build an exactanalytical model for the power dissipation within the cell membrane in suspensions exposed toexternal fields. An approximate equivalence method is proposed to resolve this problem. Based on theeffective medium theory, the transmembrane voltage on cells in suspensions was investigated by theequivalence principle. Then the electric field in the cell membrane was determined. Finally,analytical solutions for the power dissipation within the cell membrane in suspensions exposed toexternal fields were derived according to the Joule principle. The equations show that theconductive power dissipation is predominant within the cell membrane in suspensions exposed todirect current or lower frequencies, and dielectric power dissipation prevails at high frequenciesexceeding the relaxation frequency of the exposed membrane.

Photoionization Mass Spectrometric and Kinetic Modeling of Low-pressure Pyrolysis of Benzene

Pyrolysis of benzene at 30 Torr was studied from 1360 K to 1820 K in this work. Synchrotron vacuum ultraviolet photoionization mass spectrometry was employed to detect the pyroly- sis products such as radicals, isomers and polycyclic aromatic hydrocarbons, and measure their mole fraction profiles versus temperature. A low-pressure pyrolysis model of benzene was developed and validated by the experimental results. Rate of production analysis was performed to reveal the major reaction networks in both fuel decomposition and aromatic growth processes. It is concluded that benzene is mainly decomposed via H-abstraction reaction to produce phenyl and partly decomposed via unimolecular decomposition reac- tions to produce propargyl or phenyl. The decomposition process stops at the formation of acetylene and polyyne species like diacetylene and 1,3,5-hexatriyne due to their high thermal stabilities. Besides, the aromatic growth process in the low-pressure pyrolysis of benzene is concluded to initiate from benzene and phenyl, and is controlled by the even carbon growth mechanism due to the inhibited formation of C5 and C7 species which play important roles in the odd carbon growth mechanism.

Photolytic Model of Six Degradable Mulch Plastic Films

[Objective] This study aimed to solve the mulch plastic film pollution problems in Xinjiang, in order to provide reliable theoretical basis for the research on the degradation mechanism of biodegradable plastic films. [Method] The effect of illumination intensity on the decomposition of decomposable mulching films was investigated through simulating the field condition in laboratory. Regression analysis was employed to fit the processes of film decompositions. [Result] The weight loss ratios of different types of plastic films were closely related with the illumination. There was no sundry produced during the degradation process, but the weight was reduced, indicating that degradation produced gas, thus, ultraviolet rays had destructive effect on mulch plastic films. Different types of plastic films showed significant differences in the degradation speeds, and under the same conditions, the degradation speed of S4# with the induction period of 30 d was the fastest. With the extension of uv illumination time, the degradation became cumulative degradation process. [Conclusion] Under the uv rays, the weight loss ratio and illumination hours is regarded as a positive correlation relationship. The significant inspection shows that the data fitted degradation model can be described by the logistic model of Y = a/（1＋b×e -ct）, and all the parameters show significant differences （P0.01）.

Computer Simulation of Compensation Method for Infrared Focal Plane Array

A major motivation for this work is to investigate a method of computer simulation for compensating fixed pattern noise of the infrared focal plane arrays. A mathematical model of the output signal of focal plane array was established; a compensating algorithm utilizing reference source was derived and simulating programs were designed. The images of compensating process verify the influence of nonuniformity of responsibility and offset on fixed pattern noise. The result show that simulating method of investigating compensation technology for focal plane arrays is feasible, the generated images and methods can be used to the study of image recognition.

Rural Carbon Emissions Based on Community Management

Rural community is an effective way to simultaneously achieve manage- ment of ecological resources and protection of natural resources, depending on its management mode based on cooperation, motivation, local experiences and extensi- ble boundary. In the research, carbon emission in community is a kind of ecological resources and negative externality is the cause of increase of rural carbon emission. Compared with governmental ruling and marketing approach, community manage- ment mode proves more effective to solve the problem of negative externality of carbon emission in community. Furthermore, rural carbon, emission was analyzed in detail on basis of community management mode and extending mode of carbon emis- sion in rural areas. In addition, some policies and suggestions were proposed to im- prove community management of carbon emission in rural areas, providing an ef- fective way for low-carbon economy in rural areas.

Nucleation of Kinetic Ising Model Under Oscillating Field

We have studied the nucleation process of a two-dimensional kinetic Ising model subject to a bias oscillating external field, focusing on how the nucleation time depends on the oscillation frequency. It is found that the nucleation time shows a clear-cut minimum with the variation of oscillation frequency, wherein the average size of the critical nuclei is the smallest, indicating that an oscillating external field with an optimal frequency can be much more favorable to the nucleation process than a constant field. We have also investigated the effect of the initial phase of the external field, which helps to illustrate the occurrence of such an interesting finding.

Simple Model for Gas Holdup and Liquid Velocity of Annular Photocatalytic External-Loop Airlift Reactor Under both Bubble and Developing Slug Flow

Based on the momentum conservation approach, a theoretical model was developed to predict the superficial liquid velocity, and a correlation equation was established to calculate the gas holdup of an annular external-loop airlift reactor(AELAR)in the bubble flow and developing slug flow pattern. Experiments were performed by using tap-water and silicone oil with the viscosity of 2.0 mm^2/s(2cs-SiO)and 5.0 mm^2/s(5cs-SiO)as liquid phases. The effects of liquid viscosity and flow pattern on the AELAR performance were investigated. The predictions of the proposed model were in good agreement with the experimental results of the AELAR. In addition, the comparison of the experimental results shows that the proposed model has good accuracy and could be used to predict the gas holdup and liquid velocity of an AELAR operating in bubble and developing flow pattern.

Computational Exploration of Conformations of Glycine-Arginine and a Deduced Model on Global Minimum Configurations of Dipeptides in Gas Phase

An extensive computational study on the conformations of gaseous dipeptide glycine- arginine, GlyArg, has been performed. A large number of trail structures were generated by systematically sampling the potential energy surface （PES） of OlyArg. The trial structures were successively optimized with the methods of PM3, HF/3-21G＊, BHandHLYP/6-31G＊ and BHandHLYP/6-311＋＋G＊＊ in order to reliably find the low energy conformations. The eonformational energies were finally determined with the methods of BHandHLYP, cam- B3LYP, B97D, and MP2 using the basis set of 6-311＋＋G（3df,3pd）. The results establish firmly that gaseous GlyArg exists primarily in its canonical form, in sharp contrast with ArgGly that adopts the zwitterionic form. Important data such as the rotational constants, dipole moments, vertical ionization energies, temperature distributions and IR spectra of the low energy conformers are represented for the understanding of the future experiments. Moreover, considering the global minima of all amino acids and many dipeptides, combined with the hydrophobicities of amino acids, a model predicting whether the global minimum configuration of a dipeptide is canonical or zwitterionic is developed.

Foreign Direct Investment with Chinese Characteristics:A Middle Path Between Ownership-Location-Internalization Model and Linkage-Leverage-Learning Model

The majority of multinational enterprises （MNEs） traditionally originate from developed countries. In the last ten years, however, there has been dramatic growth in foreign direct investment （FDI） from China. It is a comparatively new phenomenon that challenges the classic FDI theories. In this paper, we review the pros and cons of two important theories, known as the Owner- ship-Location-Internalization （0LI） model and Linkage-Leverage-Learning （LLL） model, and use the statistical data and company case studies from China to test the plausibility of these two models. We believe that neither of them suits totally： the OLI model is quite use- fill for understanding FDI from China to developing economies, while the LLL model is more powerful for explaining the FDI to de- veloped economies. We argue that the companies from China attain a very advantageous position as intermediates in the global economy They may catch up with the first movers if they integrate OLI-led and LLL-led FDI within one firm. This combination can bring to- gether the most advanced knowledge acquired in developed economies with the knowledge about adaptation needs and the needs for cost reduction in production as expressed in developing economies. It may also accelerate the knowledge transfer globally. We thus fill a gap in research into the geographical pattern of Chinese FDI and offer a deeper understanding of the internationalization of Chinese MNEs and revolving knowledge transfer.

Nonlinear Contact Between Inner Walls of Deep Sea Pipelines in Buckling Process

In order to study buckling propagation mechanism in deep sea pipelines, the contact between pipeline's inner walls in buckling process was studied. A two-dimensional ring model was used to represent the pipeline and a nonlinear spring model was adopted to simulate the contact between inner walls. Based on the elastoplastic constitutive relationship and the principle of virtual work theory, the coupling effect of pipeline's nonlinear large deformation and wall contact was included in the theoretical analysis with the aid of MATLAB, and the application scope of the theoretical model was also discussed. The calculated results show that during the loading process, the change in external pressure is closely related to the distribution of section stress, and once the walls are contacting each other, the external pressure increases and then remains stable after it reaches a specific value. Without fracture, the pipeline section will stop showing deformation. The results of theoretical calculations agree well with those of numerical simulations. Finally, in order to ensure reliability and accuracy of the theoretical results, the collapse pressure and propagation pressure were both verified by numerical simulations and experiments. Therefore, the theoretical model can be used to analyze pipeline's buckling deformation and contact between pipeline's inner walls, which forms the basis for further research on three-dimensional buckling propagation.

Influence of Velocity Shell Correlations on Anomalous Scaling Exponents of Passive Scalars in Shell Models

We propose a new approach to the old-standing problem of the anomaly of the scaling exponents of passive scalars of turbulence. Different to the original problem, the distribution function of the prescribed random velocity field is multi-dimensional normal and delta-correlated in time. Here, our random velocity field is spatially correlative. For comparison, we also give the result obtained by the Gaussian random velocity field without spatial correlation. The anomalous scaling exponents H（p） of passive scalar advected by two kinds of random velocity above are determined for structure function up to p= 15 by numerical simulations of the random shell model with Runge-Kutta methods to solve the stochastic differential equations. We observed that the H（p） ＇s obtained by the multi-dimeasional normal distribution random velocity are more anomalous than those obtained by the independent Gaussian random velocity.

The torsion model of soft tissue in virtual surgery

In order to solve the problem of real-time soft tissue torsion simulation in virtual surgeries,a torsion model based on coil spring is proposed to actualize real-time interactions and applications in virtual surgeries. The proposed model is composed of several connected coil springs in series. The sum of torsion deformation on every coil is equivalent to the soft tissue surface deformation. The calculation of the model is simple because the method for calculating the torsion deformation for each coil spring is the same. The virtual surgery simulation system is established on PHANTOM OMNI haptic device based on the Open GL 3 D graphic interface and VC + + software,and it is used to simulate the torsion deformation of virtual legs and arms. Experimental results show that the proposed model can effectively simulate the torsion deformation of soft tissue while being of real-time performance and simplicity,which can well meet requirements of virtual operation simulations.

Empirical study on the antecedents of multinational enterprises’ oversea subsidiaries staffing

Based on probing into the literature on multinational enterprise (MNE) staffing, we set up a concept model for MNEs’ subsidiary staffing by two groups of influencing factors: the national differences bwteen the parent country and the host country, and the strategies employed by MNEs. We also tested the model and proposed propositions by a sample evaluation method, specifically with 1 000 copies of questionnaires given out to managers or directors of MNEs’ subsidiaries in China Mainland and resulting in 151 sets of valid answers. The empirical study supports that national differences between the parent country and the host country and the strategies employed by MNEs do have impact on the subsidiary staffing, and MNE headquarters should make different staffing plans according to the difference of nations and strategies. We welcome testing our model by peer researchers in other country.

In Vitro Model of Human Choroidal Neovascular

Choroidal capillary endothelia cell （CEC） plays a critical role in the development of choroidal neovascularization which is one of the major causes of blindness. An effective method for CEC cultivation was proposed. The isolation of human choroidal CECs using micro dissection followed by the use of superparamagnetic beads （Dynabeads） coated with the CD 31, which selectively binds to the endothelial cell surface. Cells bound to beads were isolated using a magnetic particle concentrator. The CECs were planted into type Ⅳ collagen coated 24 well plates. The results show that the primary cultured CEC is induced to tube formation in collagen Ⅳ coated environment, which can be presented as an in vitro model of choroidal neovascularization.