Recognition of eye diseases based on deep neural networks for transfer learning and improved D-S evidence theory.

BACKGROUND: Human vision has inspired significant advancements in computer vision, yet the human eye is prone to various silent eye diseases. With the advent of deep learning, computer vision for detecting human eye diseases has gained prominence, but most studies have focused only on a limited number of eye diseases. RESULTS: Our model demonstrated a reduction in inherent bias and enhanced robustness. The fused network achieved an Accuracy of 0.9237, Kappa of 0.878, F1 Score of 0.914 (95% CI [0.875-0.954]), Precision of 0.945 (95% CI [0.928-0.963]), Recall of 0.89 (95% CI [0.821-0.958]), and an AUC value of ROC at 0.987. These metrics are notably higher than those of comparable studies. CONCLUSIONS: Our deep neural network-based model exhibited improvements in eye disease recognition metrics over models from peer research, highlighting its potential application in this field. METHODS: In deep learning-based eye recognition, to improve the learning efficiency of the model, we train and fine-tune the network by transfer learning. In order to eliminate the decision bias of the models and improve the credibility of the decisions, we propose a model decision fusion method based on the D-S theory. However, D-S theory is an incomplete and conflicting theory, we improve and eliminate the existed paradoxes, propose the improved D-S evidence theory(ID-SET), and apply it to the decision fusion of eye disease recognition models.

Carbon, nitrogen, and phosphorus metabolic relationships and reaction mechanisms in SBBR processes in the plateau habitat.

In this study, two laboratory-scale SBBR reactors were established in a plateau habitat. Using high flux sequencing, the SBBR process was compared by natural sediment and autotrophic sludge to characterize the functional modules and functional genes of carbon, nitrogen, and phosphorus metabolism under different working conditions and to analyze the reaction mechanism. The results showed that all the functional modules of carbon metabolism and nitrogen metabolism were found in the SBBR process, except for methane metabolism, which occurred at 25 °C in tank 2, the functional modules related to methane metabolism are enhanced at all working conditions. Except for methane metabolism, all functional genes in tank 2 are inhibited by different working conditions, whereas tank 1 shows a slight enhancement. The different working conditions in nitrogen metabolism demonstrate inhibition of functional modules and functional genes in both tanks. Oxidative phosphorylation was missing five functional modules, except for M00153, where only two genes, K00424 and K22501, are missing, all of the required genes are missing in the other four functional modules. Overall the different conditions demonstrated some inhibition in both reaction tanks of the SBBR process. It is preferable to use self-cultivated sludge for membrane acclimation when operating the SBBR process in a plateau habitat. The findings of this study can be used to further research microbial carbon, nitrogen, and phosphorus metabolism mechanisms in SBBR processes in plateau habitats.