No abstract

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

No abstract

BackgroundHyperlipidaemia may be a potential risk factor for the occurrence of intestinal polyps. This study aimed to evaluate correlation between lipidaemia and the formation of colorectal polyps.MethodsOne hundred and fourteen patients with colorectal polyps and forty-eight healthy controls were included in this study. Colonoscopies were performed for all patients and controls within 1 week before blood samples were taken. The concentrations of serum lipids and lipoproteins were measured simultaneously using an automatic biochemical analyser. The colorectal lesions were classified based on pathological characteristics, and four types were identified in the study: hyperplastic polyp (HP), tubular adenoma (TA), tubulovillous adenoma (TVA) and adenoma with high-grade dysplasia (A-HGD). Advanced adenoma was classified according to the number, size and histological type of polyps.ResultsThe value of low-density lipoprotein cholesterol (LDL-C) was significantly higher in the group with advanced adenoma than in the controls (p < 0.05). Moreover, the LDL-C values in the HP and TA groups were higher when compared to that of controls (p < 0.05). Obesity, age, and increased TG and LDL-C were independent risk factors for the formation of colorectal polyps. The cut-off values of triglyceride (TG) and LDL-C to distinguish polyp patients from healthy controls were 0.96 mmol/L (AUC = 0.604, p = 0.036) and 3.05 mmol/L (AUC = 0.654, p = 0.002). The combined use of increased LDL-C and TG levels to distinguish polyp patients was effective, with a sensitivity of 50.0% and a specificity of 89.6% (AUC = 0.733, p < 0.01).ConclusionsColorectal polyps are more often found in obese and older patients. Increased LDL-C and TG were correlated with the occurrence of polyps. Combination of the two serum indicators was useful to assess risk of colorectal lesions, maybe more effective in screening hyperplastic polyp, tubular adenoma and advanced adenoma.

59Purpose: COVID-19 has become global threaten. CT acts as an important method of 60 diagnosis. However, human-based interpretation of CT imaging is time consuming. 61More than that, substantial inter-observer-variation cannot be ignored. We aim at 62 developing a diagnostic tool for artificial intelligence (AI)-based classification of CT 63 images for recognizing COVID-19 and other common infectious diseases of the lung. 64Experimental Design: In this study, images were retrospectively collected and 65 prospectively analyzed using machine learning. CT scan images of the lung that show 66 or do not show COVID-19 were used to train and validate a classification framework 67 based on convolutional neural network. Five conditions including COVID-19 68 pneumonia, non-COVID-19 viral pneumonia, bacterial pneumonia, pulmonary 69 tuberculosis, and normal lung were evaluated. Training and validation set of images 70 were collected from Wuhan Jin Yin-Tan Hospital whereas test set of images were 71 collected from Zhongshan Hospital Xiamen University and the fifth Hospital of 72 Wuhan. 73 Results: Accuracy, sensitivity, and specificity of the AI framework were reported. For 74 test dataset, accuracies for recognizing normal lung, COVID-19 pneumonia, 75 non-COVID-19 viral pneumonia, bacterial pneumonia, and pulmonary tuberculosis 76 were 99.4%, 98.8%, 98.5%, 98.3%, and 98.6%, respectively. For the test dataset, 77 accuracy, sensitivity, specificity, PPV, and NPV of recognizing COVID-19 were 78 98.8%, 98.2%, 98.9%, 94.5%, and 99.7%, respectively. 79 Conclusions:The performance of the proposed AI framework has excellent 80 performance of recognizing COVID-19 and other common infectious diseases of the 81 lung, which also has balanced sensitivity and specificity. 82 83