Search

Three-dimensional (3D) technology including 3D reconstruction and 3D printing technology, has been widely used in clinical treatment, especially in surgical planning, and image navigation technology, which can make surgical procedures more accurate, now is also increasingly favored by surgeons. But the combination of those 2 technologies was rarely reported. Thus, this study will preliminarily investigate the feasibility and the effect of the combination of 2 technologies in endonasal skull base surgery. Eight patients were involved in this study (from October 2016 to July 2017 at The Affiliated Hospital of Qingdao University), 5 cases of nasal skull base tumors and 3 cases of foreign body perforation. All operations were done under the assistance of 3D technology and image guidance system. Surgical discussion with patient, preoperative planning and clinical teaching were investigated between 2D images and 3D models by voting. For all cases, 3D reconstruction model and 3D printed model were deemed to be more helpful than CT/MRI images in surgical discussion with the patient; surgical simulation on 3D model in preoperative planning was largely deemed to be helpful and very helpful; and in clinical teaching, 3D models combined with image guidance system were deemed to be more helpful in understanding the disease than using 2D images. Besides, all patients recovered well after surgery, no recurrence and complications were found in the follow-up. The combination of 3D technology and electromagnetic image guidance system could improve surgical efficiency and the quality of clinical teaching.

In the traditional federated learning setting, a central server coordinates a network of clients to train one global model. However, the global model may serve many clients poorly due to data heterogeneity. Moreover, there may not exist a trusted central party that can coordinate the clients to ensure that each of them can benefit from others. To address these concerns, we present a novel decentralized framework, FedeRiCo, where each client can learn as much or as little from other clients as is optimal for its local data distribution. Based on expectation-maximization, FedeRiCo estimates the utilities of other participants' models on each client's data so that everyone can select the right collaborators for learning. As a result, our algorithm outperforms other federated, personalized, and/or decentralized approaches on several benchmark datasets, being the only approach that consistently performs better than training with local data only.

Institutions in highly regulated domains such as finance and healthcare often have restrictive rules around data sharing. Federated learning is a distributed learning framework that enables multi-institutional collaborations on decentralized data with improved protection for each collaborator’s data privacy. In this paper, we propose a communication-efficient scheme for decentralized federated learning called ProxyFL, or proxy-based federated learning. Each participant in ProxyFL maintains two models, a private model, and a publicly shared proxy model designed to protect the participant’s privacy. Proxy models allow efficient information exchange among participants using the PushSum method without the need of a centralized server. The proposed method eliminates a significant limitation of canonical federated learning by allowing model heterogeneity; each participant can have a private model with any architecture. Furthermore, our protocol for communication by proxy leads to stronger privacy guarantees using differential privacy analysis. Experiments on popular image datasets, and a pan-cancer diagnostic problem using over 30,000 high-quality gigapixel histology whole slide images, show that ProxyFL can outperform existing alternatives with much less communication overhead and stronger privacy.

Institutions in highly regulated domains such as finance and healthcare often have restrictive rules around data sharing. Federated learning is a distributed learning framework that enables multi-institutional collaborations on decentralized data with improved protection for each collaborator’s data privacy. In this paper, we propose a communication-efficient scheme for decentralized federated learning called ProxyFL, or proxy-based federated learning. Each participant in ProxyFL maintains two models, a private model, and a publicly shared proxy model designed to protect the participant’s privacy. Proxy models allow efficient information exchange among participants without the need of a centralized server. The proposed method eliminates a significant limitation of canonical federated learning by allowing model heterogeneity; each participant can have a private model with any architecture. Furthermore, our protocol for communication by proxy leads to stronger privacy guarantees using differential privacy analysis. Experiments on popular image datasets, and a cancer diagnostic problem using high-quality gigapixel histology whole slide images, show that ProxyFL can outperform existing alternatives with much less communication overhead and stronger privacy.

The current communication fault diagnosis methods mainly focus on the classifier with fault probability, which often leads to the low diagnosis efficiency. In order to overcome the above problems, mobile robot communication fault diagnosis method based on swarm intelligence algorithm is proposed in the paper. Firstly, the abnormal data is extracted after the analysis of the communication data with Kalman filter. Secondly, the supporting decision model was designed to standardize the communication exchange process and locate the fault range. Futher more, the ant colony optimization algorithm combined with particle swarm optimization was used to locate the fault area, and the multi-model hybrid method was adopted to comprehensively judge the communication fault. The different interference ratio was used in the experiment to test the performance of the proposed algorithm compared with the SVM and Bayesian model. Finally, the experiment results show the validity of mobile robort communication fault diagnosis based on swarm intelligence algorithm.

Here, we successfully developed nanostructured PtNi particles supported on nitrogen-doped carbon (NC), which were obtained by carbonization of metal-organic frameworks under different temperatures, forming the nano-PtNi/NC-600, nano-PtNi/NC-800, nano-PtNi/NC-900 and nano-PtNi/NC-1000 catalysts. For hydrosilylation of 1-octene, we found that the nano-PtNi/NC-1000 catalyst exhibits higher activity for anti-Markovnikov hydrosilylation of 1-octene than those of nano-PtNi/NC-600, nano-PtNi/NC-800, nano-PtNi/NC-900 catalysts. Experiments have verified that benefiting from obvious charge transfer from nano-PtNi particles to NC support carbonized at 1,000 °C, the nano-PtNi/NC-1000 catalyst achieved almost complete conversion and produce exclusive adduct for anti-Markovnikov hydrosilylation of 1-octene. Importantly, the nano-PtNi/NC-1000 catalyst exhibited good reusability for the hydrosilylation reaction. This work provides a new path to optimize electronic structure of catalysts by support modification to enhance electron transfer between metal active species and supports for highly catalytic performance.

To investigate the effect of testosterone propionate injection on the condition and prognosis of patients with sepsis.The clinical data of 61 sepsis patients admitted to the department of intensive care medicine, Weinan Central Hospital from June 2009 to October 2019 were retrospectively analyzed. All patients were treated with anti-infection, control of infection sources, organ function support, nutrition enhancement and supportive treatment. On the basis of routine treatment, observation group was given 100 mg of testosterone propionate injection for deep intramuscular injection twice a week (twice in total), and control group was not given testosterone propionate injection. The general information and laboratory indexes before treatment were observed, and the testosterone, albumin, acute physiology and chronic health evaluation II (APACHE II), sequential organ failure assessment (SOFA) score after treatment, intensive care unit (ICU) hospitalization time, total hospitalization cost, mechanical ventilation time, 28-day all-cause mortality and other indicators of the patients in two groups were compared.There were no significant differences between the two groups in gender, age and other baseline data and laboratory indexes before treatment. After treatment, in observation group the testosterone (μg/L: 3.69±2.38 vs. 2.85±0.90) and albumin (g/L: 39.87±1.98 vs. 26.25±4.13) were significantly higher than those in control group. Total hospitalization expenses (ten thousand Yuan: 10.14±3.22 vs. 12.10±3.91), APACHE II (13.71±2.13 vs. 23.23±2.52), SOFA (4.45±1.57 vs. 9.97±2.65), ICU hospitalization time (days: 12.36±4.37 vs. 14.03±3.86) and mechanical ventilation time (days: 3.00±1.85 vs. 7.00±2.50) were significantly lower than those in control group (all P0.05), and the difference in 28-day all-cause mortality of two groups was not significant [3.2% (1/31) vs. 13.3% (4/30), P0.05].Testosterone propionate injection can increase albumin level, shorten the time of mechanical ventilation, and improve the condition and prognosis of patients with sepsis.

Fenton-like catalysis has found a broad application for the removal of drug compounds from wastewater. Herein, we presented a synthesis method for the novel coordinatively unsaturated cobalt-iron bimetallic based metal-organic frameworks as heterogeneous catalysts for Fenton-like catalysis in amoxicillin (AMX) degradation process under the presence of H2O2. In order to optimize the process, the influences of catalyst dosages, H2O2 concentrations, initial pH, and the coexisting anions on AMX degradation were studied. Results showed that the complete AMX degradation was found at pH: 7, H2O2: 3 mM, catalyst dose: 0.1 g/L. After five cycle times, coordinatively unsaturated cobalt-iron bimetallic based metal-organic frameworks can still perform substantial Fenton-catalytic activity (100.0 % within 30 min). This project highlights the synthesized catalyst with outstanding performances for Fenton catalytic degradation of drug compounds. The cobalt-iron based organic frameworks with coordinated sites indicated significant antibacterial performance in the disc diffusion analysis.

The hydrosilylation reaction is one of the largest-scale application of homogeneous catalysis and is widely used to enable the commercial manufacture of silicon products. However, considerable issues including disposable platinum consumption, undesired side reactions and unacceptable catalyst residues still remain. Here, we synthesize a heterogeneous partially charged single-atom platinum supported on anatase TiO2 (Pt1δ+/TiO2) catalyst via an electrostatic-induction ion exchange and two-dimensional confinement strategy, which can catalyze hydrosilylation reaction with almost complete conversion and produce exclusive adduct. Density functional theory calculations reveal that unexpected property of Pt1δ+/TiO2 originates from atomic dispersion of active species and unique partially positive charge Ptδ+ electronic structure that conventional nanocatalysts do not possess. The fabrication of single-atom Pt1δ+/TiO2 catalyst accomplishes a reasonable use of Pt through recycling and maximum atom-utilized efficienc...

In this work, we prepared a reduced graphene oxide (RGO)/poly(3,4-ethylenedioxythiophene) (PEDOT) hybrid composite with well defined nanostructure. The graphene oxide (GO) was first deposited on substrate through the Langmuir–Blodgett (LB) deposition, which provided a tunable and ordered GO arrangement on substrate. Then the GO LB films were reduced to RGO by following thermal treatment, and a ultrathin conducting polymer (CP) PEDOT was directly coated on RGO through a vapor phase polymerization process. The RGO/PEDOT nanocomposite exhibits excellent electrical conductivity about 377.2 S/cm. Electrochemical activity investigation revealed that this nanocomposite exhibits 213 F/g high specific capacitance at a 0.5 A/g current density and shows better capacitance retention rate than pure PEDOT. The detailed study also confirmed that the arrangement of RGO shows distinct influence on the electrical and electrochemical properties of obtained nanocomposite. Large area RGO/PEDOT nanocomposite with high conductivity and electrochemical activity can be deposited on different substrates. Such high conductivity and electrochemical activity RGO/CP nanocomposite shows promising application future in organic and flexible electrode materials for sustainable energy storage.

In this work, conducting polymers poly(3,4-ethylenedioxythiophene) (PEDOT), PEDOT/carbon nanotubes (CNTs), and PEDOT/reduced graphene oxide (RGO) were prepared via an in situ chemical vapor phase polymerization (VPP) process. Experiment results showed that PEDOT and PEDOT nanocomposites were uniformly constructed in oxidant and oxidant nanocomposite films through a modifying template effect. The VPP PEDOT and its nanocomposites were built on aluminium film as supercapaitor electrode materials and electrochemical capacitive properties were investigated by using cycle voltammetry and charge/discharge techniques. The VPP PEDOT exhibited a specific capacitance of 92 F/g at a current density of 0.2 A/g. The VPP PEDOT composites consisting of CNTs and RGO displayed specific capacitances of 137 and 156 F/g, respectively, at the same current density. For VPP nanocomposites, more than 80 % of initial capacitance was retained after 1,000 charge/discharge cycles, suggesting a good cycling stability for electrochemical electrode materials. The good capacitive performance of the conducting polymer nanocomposites are contributed to the synergic effect of the two components.

Layer-ordered and ultrathin films of conducting polymer poly(3,4-ethylene dioxythiophene) (PEDOT) was prepared through a chemical vapor phase polymerization method. The chemical polymerization of 3, 4-ethylenedioxythiophene monomer was initiated in as-prepared oxidant LB films,and PEDOT nanofilms with layer-ordered structure was constructed. UV-Vis absorption spectrum and Fourier transform infrared spectroscopy was used to confirm an interface polymerization of PEDOT in as-prepared LB films. The results of X-ray diffraction and secondary ion mass spectrometry revealed that conductive PEDOT ultrathin layers were well located at different planes of LB films. The film deposition surface pressure and chemical polymerization time of PEDOT monomer in as-prepared LB films showed distinct influence on surface morphology and conductive performance of the polymerized PEDOT LB films. This layer-ordered conducting polymer ultrathin films was deposited on ITO surface as hole injection layer for organic light-emitting diodes, and the luminescence performance of devices was improved as well.

To observe the clinical and pathological features of adenoid cystic carcinoma(ACC) of external auditory canal (EAC, and analyze the possible factors related to prognosis.One out of 8 patients with ACC of EAC underwent tumor open biopsy. 5 patients underwent the extensive tumor resection, and 2 patients underwent the modified lateral temporal bone resection (1 of the 2 subjoined a total parotidectomy). Five patients received the postoperative radiotherapy.The time of follow up was 5-97 months. Two patients were loss to follow-up, 5 patients with disease free survival, and 1 patient survival with tumor.In this study, ACC of EAC is more often observed in female. The most common clinical manifestations are otalgia and neoplasm which grows slowly. The disease has high misdiagnosis rate. Surgical treatment is effective for it.

The study was performed to assess the potential role of GSTM1, GSTT1 and GSTP1 polymorphisms in the risk of nasopharyngeal cancer in Chinese population.We collected 182 cases undergoing pathologic examination and 366 controls from the affiliated hospital of Medical College of Qingdao University from April 2006 to July 2010. Genotyping was based upon duplex polymerase-chain-reactions with the PCR- CTPP method.More smokers were found in NPC patients than controls, and a higher IgA/VCA+. Individuals carrying null GSTM1 and GSTT1 had 1.76 and 2.01 fold risk of NPC when compared with non-null genotypes, respectively. A non-significant increase risk of NPC was found in individuals with 1b/1b genotype when compared with 1a/1a genotype (OR=1.32, 95%CI=0.60-2.94). When compared with non-null GSTM1 and GSTT1 genotypes, the combination of null/null GSTM1 and GSTT1 genotypes showed moderate increased risk of NPC (OR=3.03, 95% CI=1.74-5.08).Our study provides evidence that genetic deletion of GSTM1 and GSTT1 may contribute to increased susceptibility to NPC in Chinese population, while GSTP1 may not. Our findings provide information relevant to the prevention of NPC.

We report the use of zinc phthalocyanine (ZnPc) thin films as an optical gas sensor. The ZnPc thin films were fabricated by Langmuir-Blodgett technique onto quartz substrates. The responses of the thin films to various volatile organic compounds (VOCs) were recorded by using optical absorption spectroscopy under static conditions. It was found that the thin films presented a reversible changes in the absorption spectra after exposure to the test gas. The strongest gas-sensing response was observed when the films were exposed to pyridine. The experimental results indicate that the phthalocyanine can easily react with the VOCs which include electron donating group or electron withdrawing group. And the responses are in good agreement with the intensity of electron donating or electron withdrawing ability.