Satellite cloud image segmentation based on lightweight convolutional neural network.

More than 50% of the images captured by optical satellites are covered by clouds, which reduces the available information in the images and seriously affects the subsequent applications of satellite images. Therefore, the identification and segmentation of cloud regions come to be one of the most important problems in current satellite image processing. Due to the complexity and variability of satellite images, especially when the ground is covered with snow, the boundary information of cloud regions is difficult to be accurately identified. The fast and accurate segmentation of cloud regions is a difficult point in the current research. We propose a lightweight convolutional neural network. Firstly, channel attention is used to optimize the effective information in the feature maps as a way to improve the network's ability to extract semantic information at each scale. Then, we fuse high and low-dimensional feature maps to enhance the network's ability to obtain small-scale semantic information. In addition, the feature aggregation module automatically adjusts the input multi-level feature weights to highlight the details of different features. Finally, we design the fully connected conditional random field to solve the problem that some noise in the input image and local minima during training is passed to the output layer resulting in the loss of edge features. Experimental results show that the proposed method achieves 0.9695 and 0.8218 for overall accuracy and recall, respectively, which has higher segmentation accuracy with the shortest time consumption compared with other state-of-the-art methods.

Effects of the duration of postresuscitation hyperoxic ventilation on neurological outcome and survival in an asphyxial cardiac arrest rat model.

Cardiac arrest leads to sudden cessation of oxygen supply and cerebral hypoxia occurs when there is not sufficient oxygen supplied to the brain. Current Guidelines for adult cardiopulmonary resuscitation (CPR) and emergency cardiovascular care recommend the use of 100% oxygen during resuscitative efforts to maximize the probability of achieving the return of spontaneous circulation (ROSC). However, the optimal strategy for oxygen management after ROSC is still debatable. The aim of the present study was to evaluate the effects of the duration of post-resuscitation hyperoxic ventilation on neurological outcomes in asphyxial cardiac arrest rats treated with targeted temperature management (TTM). Asphyxia was induced by blocking the endotracheal tube in 80 adult male Sprague-Dawley rats. CPR begun after 7 min of untreated cardiac arrest. Animals were randomized to either the normoxic control under normothermia (NNC) group or to one of the 4 experimental groups (n = 16 each) immediately after ROSC: ventilated with 100% oxygen for 0 (O2_0h), 1 (O2_1h), 3 (O2_3h), or 5 (O2_5h) h and ventilated with room air thereafter under TTM. Physiological variables were recorded at baseline and during the 6 h postresuscitation monitoring period. Animals were closely observed for 96 h to assess neurologic recovery and survival. There were no significant differences in baseline measurements between groups, and all animals were successfully resuscitated. There were significant interactions between the duration of 100% oxygen administration and hemodynamics as well as, myocardial and cerebral injuries. Among all the durations of hyperoxic ventilation investigated, significantly lower neurological deficit scores and higher survival rates were observed in the O2_3h group than in the NNC group. In conclusion, postresuscitation hyperoxic ventilation leads to improved PaO2, PaCO2, hemodynamic, myocardial and cerebral recovery in asphyxial cardiac arrest rats treated with TTM. However, the beneficial effects of high concentration-oxygen are duration dependent and ventilation with 100% oxygen during induced hypothermia contributes to improved neurological recovery and survival after 96 h.