MAFE-Net: retinal vessel segmentation based on a multiple attention-guided fusion mechanism and ensemble learning network.

The precise and automatic recognition of retinal vessels is of utmost importance in the prevention, diagnosis and assessment of certain eye diseases, yet it brings a nontrivial uncertainty for this challenging detection mission due to the presence of intricate factors, such as uneven and indistinct curvilinear shapes, unpredictable pathological deformations, and non-uniform contrast. Therefore, we propose a unique and practical approach based on a multiple attention-guided fusion mechanism and ensemble learning network (MAFE-Net) for retinal vessel segmentation. In conventional UNet-based models, long-distance dependencies are explicitly modeled, which may cause partial scene information loss. To compensate for the deficiency, various blood vessel features can be extracted from retinal images by using an attention-guided fusion module. In the skip connection part, a unique spatial attention module is applied to remove redundant and irrelevant information; this structure helps to better integrate low-level and high-level features. The final step involves a DropOut layer that removes some neurons randomly to prevent overfitting and improve generalization. Moreover, an ensemble learning framework is designed to detect retinal vessels by combining different deep learning models. To demonstrate the effectiveness of the proposed model, experimental results were verified in public datasets STARE, DRIVE, and CHASEDB1, which achieved F1 scores of 0.842, 0.825, and 0.814, and Accuracy values of 0.975, 0.969, and 0.975, respectively. Compared with eight state-of-the-art models, the designed model produces satisfactory results both visually and quantitatively.

Distribution, migration, and removal of N-containing products during polyurethane pyrolysis: A review.

Due to the wide applications of polyurethane (PU), production is constantly increasing, accounting for 8% of produced plastics. PU has been regarded as the 6th most used polymer in the world. Improper disposal of waste PU will result in serious environmental consequences. The pyrolysis of polymers is one of the most commonly used disposal methods, but PU pyrolysis easily produces toxic and harmful nitrogen-containing substances due to its high nitrogen content. This paper reviews the decomposition pathways, kinetic characteristics, and migration of N-element by product distribution during PU pyrolysis. PU ester bonds break to produce isocyanates and alcohols or decarboxylate to produce primary amines, which are then further decomposed to MDI, MAI, and MDA. The nitrogenous products, including NH3, HCN, and benzene derivatives, are released by the breakage of C-C and C-N bonds. The N-element migration mechanism is concluded. Meanwhile, this paper reviews the removal of gaseous pollution from PU pyrolysis and discusses the removal mechanism in depth. Among the catalysts for pollutant removal, CaO has the most superior catalytic performance and can convert fuel-N to N2 by adsorption and dehydrogenation reactions. At the end of the review, new challenges for the utilization and high-quality recycling of PU are presented.

Effect of diesel blended with di-n-butyl ether/1-octanol on combustion and emission in a heavy-duty diesel engine.

Two kinds of C8 isomers, di-n-butyl ether (DNBE) and 1-octanol, as potential oxygen-containing alternative fuels, show important value in the trade-off between efficiency and emission. In the present work, the effects of DNBE/1-octanol with different proportions (0, 10%, and 20%) blended into diesel on the combustion characteristics, fuel economy, and emission characteristics in a six-cylinder heavy-duty diesel engine were studied at low, medium, and high loads. 1-Octanol with a 20% blending ratio showed different combustion characteristics in the cylinder compared with the other fuels. The economic analysis showed that the brake specific fuel consumption of DNBE-diesel blend fuels was higher than that of 1-octanol-diesel blend fuels, while brake thermal efficiency was the opposite tendency. The emissions of nitrogen oxides (NOx), hydrocarbons (HC), and carbon monoxide (CO) were affected by the types of blend fuels, blending ratios, and loads. In comparison with 1-octanol-diesel blend fuels, the addition of DNBE in diesel promoted the emission of nitrogen oxides, but inhibited the emissions of soot, HC, and CO. DNBE- and 1-octanol-diesel blend fuels increased the weighted brake specific fuel consumption but decreased the weighted brake thermal efficiency compared with diesel in the World Harmonized Stationary Cycle test cycle of Euro VI regulation. The weighted NOx, HC, soot, and CO emissions of blend fuels depended on the types of blend fuels and blend ratios. The weighted NOx, HC, and soot emissions were reduced by blending 1-octanol into diesel, while the weighted CO emission was increased. The weighted CO and soot emissions of diesel blended with DNBE were reduced than that of diesel.