BackgroundThis study aimed to evaluate the predictive power of five available delayed graft function (DGF)-prediction models for kidney transplants in the Chinese population.ResultsAmong the five models, the Irish 2010 model scored the best in performance for the Chinese population. Irish 2010 model had an area under the receiver operating characteristic (ROC) curve of 0.737. Hosmer-Lemeshow goodness-of-fit test showed that the Irish 2010 model had a strong correlation between the calculated DGF risk and the observed DGF incidence (p = 0.887). When Irish 2010 model was used in the clinic, the optimal upper cut-off was set to 0.5 with the best positive likelihood ratio, while the lower cut-off was set to 0.1 with the best negative likelihood ratio. In the subgroup of donor aged ≤ 5, the observed DGF incidence was significantly higher than the calculated DGF risk by Irish 2010 model (27% vs. 9%).Materials and MethodsA total of 711 renal transplant cases using deceased donors from China Donation after Citizen's Death Program at our center between February 2007 and August 2016 were included in the analysis using the five predictive models (Irish 2010, Irish 2003, Chaphal 2014, Zaza 2015, Jeldres 2009).ConclusionsIrish 2010 model has the best predictive power for DGF risk in Chinese population among the five models. However, it may not be suitable for allograft recipients whose donor aged ≤ 5-year-old.

Swarm robots search for multiple targets in collaboration in unknown environments has been addressed in this paper. An improved grouping strategy based on constriction factors Particle Swarm Optimization is proposed. Robots are grouped under this strategy after several iterations of stochastic movements, which considers the influence range of targets and environmental information they have sensed. The group structure may change dynamically and each group focuses on searching one target. All targets are supposed to be found finally. Obstacle avoidance is considered during the search process. Simulation compared with previous method demonstrates the adaptability, accuracy and efficiency of the proposed strategy in multiple targets searching.

No abstract

No abstract

Deep neural network approaches have made remarkable progress in many machine learning tasks. However, the latest research indicates that they are vulnerable to adversarial perturbations. An adversary can easily mislead the network models by adding well-designed perturbations to the input. The cause of the adversarial examples is unclear. Therefore, it is challenging to build a defense mechanism. In this paper, we propose an image-to-image translation model to defend against adversarial examples. The proposed model is based on a conditional generative adversarial network, which consists of a generator and a discriminator. The generator is used to eliminate adversarial perturbations in the input. The discriminator is used to distinguish generated data from original clean data to improve the training process. In other words, our approach can map the adversarial images to the clean images, which are then fed to the target deep learning model. The defense mechanism is independent of the target model, and the structure of the framework is universal. A series of experiments conducted on MNIST and CIFAR10 show that the proposed method can defend against multiple types of attacks while maintaining good performance.