ABSTRACT
Recently, segmentation-based scene text detection methods have drawn extensive attention in the scene text detection field, because of their superiority in detecting the text instances of arbitrary shapes and extreme aspect ratios, profiting from the pixel-level descriptions. However, the vast majority of the existing segmentation-based approaches are limited to their complex post-processing algorithms and the scale robustness of their segmentation models, where the post-processing algorithms are not only isolated to the model optimization but also time-consuming and the scale robustness is usually strengthened by fusing multi-scale feature maps directly. In this paper, we propose a Differentiable Binarization (DB) module that integrates the binarization process, one of the most important steps in the post-processing procedure, into a segmentation network. Optimized along with the proposed DB module, the segmentation network can produce more accurate results, which enhances the accuracy of text detection with a simple pipeline. Furthermore, an efficient Adaptive Scale Fusion (ASF) module is proposed to improve the scale robustness by fusing features of different scales adaptively. By incorporating the proposed DB and ASF with the segmentation network, our proposed scene text detector consistently achieves state-of-the-art results, in terms of both detection accuracy and speed, on five standard benchmarks.
ABSTRACT
BACKGROUND: Coronavirus disease 2019 (COVID-19) is caused by severe acute respiratory syndrome coronavirus 2, and outbreaks have occurred worldwide. Laboratory test results are an important basis for clinicians to determine patient condition and formulate treatment plans. METHODS: Fifty-two thousand six hundred forty-four laboratory test results with continuous values of adult inpatients who were diagnosed with COVID-19 and hospitalized in the Fifth Hospital in Wuhan between 16 January 2020 and 18 March 2020 were compiled. The first and last test results were compared between survivors and non-survivors with variance test or Welch test. Laboratory test variables with significant differences were then included in the temporal change analysis. RESULTS: Among 94 laboratory test variables in 82 survivors and 25 non-survivors with COVID-19, white blood cell count, neutrophil count/percentage, mean platelet volume, platelet distribution width, platelet-large cell percentage, hypersensitive C-reactive protein, procalcitonin, D-dimer, fibrin (ogen) degradation product, middle fluorescent reticulocyte percentage, immature reticulocyte fraction, lactate dehydrogenase were significantly increased (P < 0.05), and lymphocyte count/percentage, monocyte percentage, eosinophil percentage, prothrombin activity, low fluorescent reticulocyte percentage, plasma carbon dioxide, total calcium, prealbumin, total protein, albumin, albumin-globulin ratio, cholinesterase, total cholesterol, nonhigh-density/low-density/small-dense-low-density lipoprotein cholesterol were significantly decreased in non-survivors compared with survivors (P < 0.05), in both first and last tests. Prothrombin time, prothrombin international normalized ratio, nucleated red blood cell count/percentage, high fluorescent reticulocyte percentage, plasma uric acid, plasma urea nitrogen, cystatin C, sodium, phosphorus, magnesium, myoglobin, creatine kinase (isoenzymes), aspartate aminotransferase, alkaline phosphatase, glucose, triglyceride were significantly increased (P < 0.05), and eosinophil count, basophil percentage, platelet count, thrombocytocrit, antithrombin III, red blood cell count, haemoglobin, haematocrit, total carbon dioxide, acidity-basicity, actual bicarbonate radical, base excess in the extracellular fluid compartment, estimated glomerular filtration rate, high-density lipoprotein cholesterol, apolipoprotein A1/ B were significantly decreased in non-survivors compared with survivors (P < 0.05), only in the last tests. Temporal changes in 26 variables, such as lymphocyte count/percentage, neutrophil count/percentage, and platelet count, were obviously different between survivors and non-survivors. CONCLUSIONS: By the comprehensive usage of the laboratory markers with different temporal changes, patients with a high risk of COVID-19-associated death or progression from mild to severe disease might be identified, allowing for timely targeted treatment.
