摘要
一位新闻记者-机器人与机器学习的工作人员新闻编辑每日新闻-肺部疾病和疾病的新研究-肺栓塞是一篇报道的主题。摘要:根据《新闻周刊》编辑在湖北发表的新闻报道,研究表明:“肺栓塞(PE)是自身免疫性炎性风湿性疾病(aiird)患者中死亡率高的严重急性心血管综合征,准确预测和及时干预对提高生存率至关重要。”我们的新闻记者从华中科技大学的研究中获得了一句话：“然而，在AIRD患者中，实用的PE早期预测和风险评估系统明显缺乏。”在培训队列中，60例AIRD患者患有PE，180例年龄、性别、对同济医院2014～2022年7254例AIIRD患者进行了疾病匹配非PE病例的筛选。采用单变量Logistic回归(LR)和最小绝对回归和选择算子(LASSO),采用机器学习(ML)方法,包括随机前ST(RF),支持向量机(SVM),神经网络(NN),logistic回归(LR),梯度提升决策树(GBDT),对临床特征进行进一步训练。在训练队列中,分别采用单变量LR和LASO策略筛选出24个和13个临床特征,5个ML模型(RF、SVM、NN、LR和GBDT)表现出良好的性能。训练队列(ROC)曲线下面积(AUC)为0.962～1.000，验证队列为0.969～0.999，训练队列CART模型和C5.0模型的AUCs分别为0.850和0.932，以d-二聚体作为预筛选指标，改良的C5.0模型在训练队列中AUC超过0.948，在验证队列中AUC超过0.925，这些结果明显优于单独使用D-二聚体水平。
Abstract
By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-New research on Lung Diseases and Cond itions - Pulmonary Embolism is the subject of a report. According to news report ing out of Hubei, People's Republic of China, by NewsRx editors, research stated , "Pulmonary embolism (PE) is a severe and acute cardiovascular syndrome with hi gh mortality among patients with autoimmune inflammatory rheumatic diseases (AII RDs). Accurate prediction and timely intervention play a pivotal role in enhanci ng survival rates." Our news journalists obtained a quote from the research from the Huazhong Univer sity of Science and Technology, "However, there is a notable scarcity of practic al early prediction and risk assessment systems of PE in patients with AIIRD. In the training cohort, 60 AIIRD with PE cases and 180 age-, gender-, and disease- matched AIIRD non-PE cases were identified from 7254 AIIRD cases in Tongji Hospi tal from 2014 to 2022. Univariable logistic regression (LR) and least absolute s hrinkage and selection operator (LASSO) were used to select the clinical feature s for further training with machine learning (ML) methods,including random fore st (RF), support vector machines (SVM), neural network (NN), logistic regression (LR), gradient boosted decision tree (GBDT), classification and regression tree s (CART), and C5.0 models. The performances of these models were subsequently va lidated using a multicenter validation cohort. In the training cohort, 24 and 13 clinical features were selected by univariable LR and LASSO strategies, respect ively. The five ML models (RF, SVM, NN, LR, and GBDT) showed promising performan ces, with an area under the receiver operating characteristic (ROC) curve (AUC) of 0.962-1.000 in the training cohort and 0.969-0.999 in the validation cohort. CART and C5.0 models achieved AUCs of 0.850 and 0.932, respectively, in the trai ning cohort. Using D-dimer as a pre-screening index, the refined C5.0 model achi eved an AUC exceeding 0.948 in the training cohort and an AUC above 0.925 in the validation cohort. These results markedly outperformed the use of D-dimer level s alone."
NETL
NSTL