Search

Additional file 3: Table S1. The performance of models with and without NSPs for identifying patients with unfavorable outcome at 3 months. Table S2. Bivariate correlation between neutrophil counts and neutrophil elastase (NE), cathepsin G (CTSG), as well as proteinase 3 (PR3) plasma concentrations in the subgroups. Table S3. The Youden index of neutrophils serine proteinases for identifying patients with unfavorable outcome at 3 months in different groups. Table S4. Comparisons of dichotomous neutrophil elastase (NE), and proteinase 3 (PR3) levels according to the modified Rankin Scale (mRS) at 3 months in the whole study population (n = 342). Table S5. Logistic regression analysis and additional predictive value of the model including neutrophil elastase (NE) and proteinase 3 (PR3) concentration for patients with unfavorable outcome (mRS > 2) at 3 months in the subgroup of patients with available infarct volume (n = 212). Table S6. Comparisons of neutrophil serine proteinases plasma concentrations between patients eligible for IV recombinant tissue plasminogen activator (rtPA) or not. Table S7. Comparisons of neutrophil serine proteinases plasma concentrations between patients received bridging therapy or not. Table S8. Demographic and clinical data for the subgroup of patients received IV recombinant tissue plasminogen activator (rtPA) treatment (n = 119). Table S9. Comparison of dichotomous neutrophil elastase (NE) and proteinase 3 (PR3) levels in the subgroup of patients received rtPA treatment according to the modified Rankin Scale (mRS) at 3 months. Table S10. Correlation between neutrophil serine proteinases plasma concentrations and baseline characteristics in the whole study population (n = 342).

Self-centering joints always show good seismic performance and self-centering capacity. However, there is rare study on fire response of self-centering joint. This paper presents a comprehensive experimental study on the fire response of self-centering beam-column joints as well as their post-fire seismic performance. In total, two joints were designed and fabricated. One was used as reference and the other subjected to constant vertical loads on the beam ends was tested under the ISO-834 fire. The fire test terminated after 171min without reaching failure limit states. The test results indicate that the fire resistance of self-centering joint is satisfactory. Then both joints were tested under quasi-static loading. The fire exposure significantly affected the post-fire seismic performance of self-centering joint. Nevertheless, self-centering joint could sustain severe seismic excitation after fire exposure.

Automatic image colorization is a challenging problem. Due to the high illness and multi-modal uncertainty, directly training a deep neural network usually leads to incorrect semantic colors and low color richness. Recent transformer-based methods can deliver better results, but they often rely on manually designed priors, which are hard to implement and suffer from poor generalization ability. Moreover, they tend to introduce serious color bleeding effects since color attention is performed on single-scale features, thus fail to exploit sufficient semantic information. To address these issues, we propose DDColor, a new end-to-end method with dual decoders for image colorization. Our approach includes a multi-scale image decoder and a transformer-based color decoder. The former restores the spatial resolution of the image, while the latter establishes the correlation between color and semantic representations via cross-attention. Rather than using additional priors, our two decoders work together to leverage multi-scale image features to guide optimization of adaptive color queries, significantly alleviating color bleeding effects. In addition, a simple yet effective colorfulness loss is introduced to further enhance the color richness of generated results. Our extensive experiments demonstrate that DDColor achieves significantly superior performance to existing state-of-the-art works both quantitatively and qualitatively. Codes will be made publicly available at https://github.com/piddnad/DDColor.