Prediction and feature selection for fatal gastrointestinal bleeding recurrence in hospital via machine learning / 中华危重病急救医学

OBJECTIVE@#To propose a method of prediction for fatal gastrointestinal bleeding recurrence in hospital and a method of feature selection via machine learning models.@*METHODS@#728 digestive tract hemorrhage samples were extracted from the first aid database of PLA General Hospital, and 343 patients among them were diagnosed as fatal gastrointestinal bleeding recurrence in hospital. A total of 64 physiological or laboratory indicators were extracted and screened. Based on the ten-fold cross-validation, Logistic regression, AdaBoost and XGBoost were used for classification prediction and comparison. XGBoost was used to search sequence features, and the key indicators for predicting fatal gastrointestinal bleeding recurrence in hospital were screened out according to the importance of the indicators during training.@*RESULTS@#Logistic regression, AdaBoost and XGBoost all get better F1.5 score under each feature input dimension, among which XGBoost had the best effect and the highest score, which was able to identify as many patients as possible who might have fatal gastrointestinal bleeding recurrence in hospital. Through XGBoost iteration results, the Top 30 indicators with high importance for predicting fatal gastrointestinal bleeding recurrence in hospital were ranked. The F1.5 scores of the first 12 key indicators peaked at iteration (0.893), including hemoglobin (Hb), calcium (CA), red blood cell count (RBC), mean platelet volume (MPV), mean erythrocyte hemoglobin concentration (MCH), systolic blood pressure (SBP), platelet count (PLT), magnesium (MG), lymphocyte (LYM), glucose (GLU, blood gas analysis), glucose (GLU, blood biochemistry) and diastolic blood pressure (DBP).@*CONCLUSIONS@#Logistic regression, AdaBoost and XGBoost could achieve the purpose of early warning for predicting fatal gastrointestinal bleeding recurrence in hospital, and XGBoost is the most suitable. The 12 most important indicators were screened out by sequential forward selection.

Prediction and feature selection for fatal gastrointestinal bleeding recurrence in hospital via machine learning / 中华危重病急救医学

Objective To propose a method of prediction for fatal gastrointestinal bleeding recurrence in hospital and a method of feature selection via machine learning models. Methods 728 digestive tract hemorrhage samples were extracted from the first aid database of PLA General Hospital, and 343 patients among them were diagnosed as fatal gastrointestinal bleeding recurrence in hospital. A total of 64 physiological or laboratory indicators were extracted and screened. Based on the ten-fold cross-validation, Logistic regression, AdaBoost and XGBoost were used for classification prediction and comparison. XGBoost was used to search sequence features, and the key indicators for predicting fatal gastrointestinal bleeding recurrence in hospital were screened out according to the importance of the indicators during training. Results Logistic regression, AdaBoost and XGBoost all get better F1.5 score under each feature input dimension, among which XGBoost had the best effect and the highest score, which was able to identify as many patients as possible who might have fatal gastrointestinal bleeding recurrence in hospital. Through XGBoost iteration results, the Top 30 indicators with high importance for predicting fatal gastrointestinal bleeding recurrence in hospital were ranked. The F1.5 scores of the first 12 key indicators peaked at iteration (0.893), including hemoglobin (Hb), calcium (CA), red blood cell count (RBC), mean platelet volume (MPV), mean erythrocyte hemoglobin concentration (MCH), systolic blood pressure (SBP), platelet count (PLT), magnesium (MG), lymphocyte (LYM), glucose (GLU, blood gas analysis), glucose (GLU, blood biochemistry) and diastolic blood pressure (DBP). Conclusions Logistic regression, AdaBoost and XGBoost could achieve the purpose of early warning for predicting fatal gastrointestinal bleeding recurrence in hospital, and XGBoost is the most suitable. The 12 most important indicators were screened out by sequential forward selection.

Ilizarov technique in combination with bone grafting and ankle traction for treatment of periarticular nonunion at the distal tibia / 中华创伤骨科杂志

Objective To evaluate the clinical efficacy of a combination of Ilizarov technique, bone grafting and ankle traction for treatment of nonunion of the distal tibia within 2 cm around the ankle joint. Methods A retrospective study was performed on the patients who had been treated at Department of Orthopaedics, Sir Run Run Shaw Hospital from September 2014 to January 2018 for nonunion of the distal tibia within 2 cm around the ankle joint. They were 8 men and 4 women, aged from 32 to 56 years ( average, 41 years). Their injury was all unilateral, involving 5 left and 7 right ankles. They had all received internal fixation at other hospitals for 6 to 18 months. After we treated them simultaneously with Ilizarov technique, bone grafting and ankle traction, they could walk with crutches as soon as 3 days after surgery in some cases. Their affected limbs were then subjected to gradual weight-bearing walking until they could walk normally without any aid 3 to 6 months after surgery. Their postoperative ankle functions were evaluated according to the ankle-hindfoot scores of the American Orthopedic Foot and Ankle Society ( AOFAS) system. Results The 12 patients were effec-tively followed up for 12 to 24 months ( average, 19 months). All their incisions healed well with no incision or pin-tract infection. All their bone nonunions healed after circular external fixation for 12 to 36 weeks. Their AOFAS scores ranged from 75 to 96 points, showing good functional recovery of the ankle joint. Conclusion A combination of Ilizarov technique, bone grafting and ankle traction is an effective treatment for nonunion of the distal tibia within 2 cm around the ankle joint, leading to preservation of the ankle function and prevention of posttraumatic ankle osteoarthritis.