PurposeTo explore the changes of bacterial flora in anophthalmic patients wearing ocular prosthesis (OP) and the microbiome diversity in conditions of different OP materials.MethodsA cross-sectional clinical study was conducted, involving 19 OP patients and 23 healthy subjects. Samples were collected from the upper, lower palpebral, caruncle, and fornix conjunctiva. 16S rRNA sequencing was applied to identify the bacterial flora in the samples. The eye comfort of each OP patient was determined by a questionnaire. In addition, demographics information of each participant was also collected.ResultsThe diversity and richness of ocular flora in OP patients were significantly higher than that in healthy subjects. The results of flora species analysis also indicated that in OP patients, pathogenic microorganisms such as Escherichia Shigella and Fusobacterium increased significantly, while the resident flora of Lactobacillus and Lactococcus decreased significantly. Within the self-comparison of OP patients, compared with Polymethyl Methacrylate (PMMA), prosthetic material of glass will lead to the increased colonization of opportunistic pathogens such as Alcaligenes, Dermabacter and Spirochaetes, while gender and age have no significant impact on ocular flora.ConclusionsThe ocular flora of OP patients was significantly different from that of healthy people. Abundant colonization of pathogenic microorganisms may have an important potential relationship with eye discomfort and eye diseases of OP patients. PMMA, as an artificial eye material, demonstrated potential advantages in reducing the colonization of opportunistic pathogens.

Sepsis is a systemic inflammatory reaction caused by various infectious or noninfectious factors, which can lead to shock, multiple organ dysfunction syndrome, and death. It is one of the common complications and a main cause of death in critically ill patients. At present, the treatments of sepsis are mainly focused on the controlling of inflammatory response and reduction of various organ function damage, including anti-infection, hormones, mechanical ventilation, nutritional support, and traditional Chinese medicine (TCM). Among them, Xuebijing injection (XBJI) is an important derivative of TCM, which is widely used in clinical research. However, the molecular mechanism of XBJI on sepsis is still not clear. The mechanism of treatment of “bacteria, poison and inflammation” and the effects of multi-ingredient, multi-target, and multi-pathway have still not been clarified. For solving this issue, we designed a new systems pharmacology strategy which combines target genes of XBJI and the pathogenetic genes of sepsis to construct functional response space (FRS). The key response proteins in the FRS were determined by using a novel node importance calculation method and were condensed by a dynamic programming strategy to conduct the critical functional ingredients group (CFIG). The results showed that enriched pathways of key response proteins selected from FRS could cover 95.83% of the enriched pathways of reference targets, which were defined as the intersections of ingredient targets and pathogenetic genes. The targets of the optimized CFIG with 60 ingredients could be enriched into 182 pathways which covered 81.58% of 152 pathways of 1,606 pathogenetic genes. The prediction of CFIG targets showed that the CFIG of XBJI could affect sepsis synergistically through genes such as TAK1, TNF-α, IL-1β, and MEK1 in the pathways of MAPK, NF-κB, PI3K-AKT, Toll-like receptor, and tumor necrosis factor signaling. Finally, the effects of apigenin, baicalein, and luteolin were evaluated by in vitro experiments and were proved to be effective in reducing the production of intracellular reactive oxygen species in lipopolysaccharide-stimulated RAW264.7 cells, significantly. These results indicate that the novel integrative model can promote reliability and accuracy on depicting the CFIGs in XBJI and figure out a methodological coordinate for simplicity, mechanism analysis, and secondary development of formulas in TCM.

The automatic segmentation of retinal images obtained by optical coherence tomography is increasingly important for ophthalmologists to diagnose and monitor many kinds of ophthalmic diseases. U-Net is the most widely used deep learning network in retinal segmentation, but the limited number of data-flow paths made it hard to capture complex features. We proposed here an optimized Mobile Ladder U-Net (MLU-Net), which consists of a Ladder Connection for increasing the network’s data-flow paths and a depthwise separable convolution for reducing the model’s parameters. Experiments on 100 B-scans from 10 human eyes demonstrated that the 9 retinal layer boundaries can be segmented accurately with the MLU-Net. In addition, compared with the original U-Net and LadderNet, the segmentation result of our method is closest to the expert label.