Association of Thallium with Diabetes Risk among Patients with Hearing Loss: Result from NHANES 2013 to 2018.

To evaluate the correlation between thallium and diabetes risk among participants with hearing loss. This retrospective cohort study extracted related data such as demographic characteristics, lifestyle factors, and laboratory findings from the National Health and Nutrition Examination Survey (NHANES) database (2013-2018). Logistic regression analysis and interaction analysis were adopted to analyze the correlation between thallium and diabetes risk among patients with hearing loss. Then, the restricted cubic spline was employed to assess the nonlinear relationship between thallium and diabetes risk. The receiver operating characteristic curve and decision curve analysis were used to assess the predictive values of 3 multivariate models with or without thallium for diabetes risk. The Delong test was adopted to assess the significant change of the area under the curves (AUCs) upon thallium addition. A total of 425 participants with hearing loss were enrolled in the study: without diabetes group (n = 316) and diabetes group (n = 109). Patients with hearing loss in the diabetes group had significantly lower thallium (P < .05). The thallium was an independent predictor for diabetes risk after adjusting various covariates (P < .05). The restricted cubic spline (RCS) result showed that there was a linear correlation between thallium and diabetes risk (P nonlinear > .05). Finally, the receiver operating characteristic and decision curve analysis results revealed that adding thallium to the models slightly increased the performance in predicting diabetes risk but without significance in AUC change. Thallium was an independent predictor of diabetes risk among patients with hearing loss. The addition of thallium might help improve the predictive ability of models for risk reclassification. However, the conclusions should be verified in our cohort in the future due to the limitations inherent in the NHANES database.

Can three-dimensional nitrate structure be reconstructed from surface information with artificial intelligence? - A proof-of-concept study.

Nitrate is one of the essential variables in the ocean that is a primary control of the upper ocean pelagic ecosystem. Its three-dimensional (3D) structure is vital for understanding the dynamic and ecosystem. Although several gridded nitrate products exist, the possibility of reconstructing the 3D structure of nitrate from surface data has never been exploited. In this study, we employed two advanced artificial intelligence (AI) networks, U-net and Earthformer, to reconstruct nitrate concentration in the Indian Ocean from surface data. Simulation from an ecosystem model was utilized as the labeling data to train and test the AI networks, with wind vectors, wind stress, sea surface temperature, sea surface chlorophyll-a, solar radiation, and precipitation as the input. We compared the performance of two networks and different pre-processing methods. With the input features decomposed into climatology and anomaly components, the Earthformer achieved optimal reconstruction results with a lower normalized mean square error (NRMSE = 0.1591), spatially and temporally, outperforming U-net (NRMSE = 0.2007) and the climatology prediction (NRMSE = 0.2089). Furthermore, Earthformer was more capable of identifying interannual nitrate anomalies. With a network interpretation technique, we quantified the spatio-temporal importance of every input feature in the best case (Earthformer with decomposed inputs). The influence of different input features on nitrate concentration in the adjacent Java Sea exhibited seasonal variation, stronger than the interannual one. The feature importance highlighted the role of dynamic factors, particularly the wind, matching our understanding of the dynamic controls of the ecosystem. Our reconstruction and network interpretation technique can be extended to other ecosystem variables, providing new possibilities in studies of marine environment and ecology from an AI perspective.

Physical-biological coupling induced aggregation mechanism for the formation of high biomass red tides in low nutrient waters.

Port Shelter is a semi-enclosed bay in northeast Hong Kong where high biomass red tides are observed to occur frequently in narrow bands along the local bathymetric isobars. Previous study showed that nutrients in the Bay are not high enough to support high biomass red tides. The hypothesis is that physical aggregation and vertical migration of dinoflagellates appear to be the driving mechanism to promote the formation of red tides in this area. To test this hypothesis, we used a high-resolution estuarine circulation model to simulate the near-shore water dynamics based on in situ measured temperature/salinity profiles, winds and tidal constitutes taken from a well-validated regional tidal model. The model results demonstrated that water convergence occurs in a narrow band along the west shore of Port Shelter under a combined effect of stratified tidal current and easterly or northeasterly wind. Using particles as dinoflagellate cells and giving diel vertical migration, the model results showed that the particles aggregate along the convergent zone. By tracking particles in the model predicted current field, we estimated that the physical-biological coupled processes induced aggregation of the particles could cause 20-45 times enhanced cell density in the convergent zone. This indicated that a high cell density red tide under these processes could be initialized without very high nutrients concentrations. This may explain why Port Shelter, a nutrient-poor Bay, is the hot spot for high biomass red tides in Hong Kong in the past 25 years. Our study explains why red tide occurrences are episodic events and shows the importance of taking the physical-biological aggregation mechanism into consideration in the projection of red tides for coastal management.