Prediction of fatal adverse prognosis in patients with fever-related diseases based on machine learning: A retrospective study

Zhao, Chun-Hong; Wu, Hui-Tao; Che, He-Bin; Song, Ya-Nan; Zhao, Yu-Zhuo; Li, Kai-Yuan; Xiao, Hong-Ju; Zhai, Yong-Zhi; Liu, Xin; Lu, Hong-Xi; Li, Tan-Shi

After selecting the optimal number by the RFE method with decision tree model, we got the highest accuracy when there were 15 factors in model, and then 15 was taken as the coefficient of RFE to select the specific factors. The obtained 15 factors through these processes were: heart rate (HR), respiratory rate (RR), systolic blood pressure (SBP), diastolic blood pressure (DBP), pulse oxygen saturation (SPO 2 ), temperature (T), creatine kinase myocardial isoenzyme (CK-MB), total bilirubin (TBIL), lactate dehydrogenase (LDH), serum amylase (AMY), serum lipase (LIP), cardiac troponin T (CTnT), aerum kalium (K), total protein (TP), and albumin (ALB). After that, the Pearson correlation test was done, with the details shown in Figure 1 The results were expressed as the median (interquartile range) or mean ± standard deviation. Table 2 and Figure 2 ]. Tenfold cross-validation was performed on the logistic and bagging models with better comprehensive performance, and their ROC-AUC were 0.80 and 0.87, respectively. In validation part, the decision tree model got the highest accuracy and F1-score, while the bagging model got the highest sensitivity and ROC-AUC.