Clinical epidemiology · 2026
Machine Learning Model Predicts New-Onset Lower Extremity Deep Vein Thrombosis After Pelvic Fracture Surgery and Targeted Diagnosis.
СтатьяНазвание и аннотация: Перевод пока не подготовлен
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Fu Haoyuan, Dong Qi, Li Guoqiang, Zhao Kuo, Hou Zhiyong
Аннотация
BACKGROUND: Postoperative new-onset deep vein thrombosis (PNO-DVT) is a common and serious complication after pelvic fracture surgery, significantly affecting recovery and quality of life. Traditional risk assessment tools lack precision, whereas machine learning offers improved predictive capability. This study evaluated machine learning models for predicting PNO-DVT following pelvic fracture surgery.
METHODS: Clinical data from 745 patients treated between 2016 and 2019 were analyzed. Demographic characteristics, laboratory indicators, surgical variables, and scoring systems were collected. Univariate logistic regression, LASSO regression, and multivariate logistic regression identified 12 independent risk factors. The dataset was split 7:3 into training and test sets. Six machine learning models (logistic regression, SVM, random forest, XGBoost, LightGBM, and AdaBoost) were constructed.
RESULTS: Univariate analysis identified GGT, HDL-C, ApoB, TCO2, GLU, GAP, EOS, and MPV as potential predictors. Multivariate analysis confirmed age, nephropathy, BMI, fracture reduction, intraoperative blood loss, red blood cell transfusion, HDL-C, ApoB, GAP, and MPV as independent risk factors. Among all models, XGBoost achieved the highest AUC (0.8633), demonstrating superior predictive performance. Model accuracy ranged from 0.6502 to 0.9283.
CONCLUSION: Machine learning (particularly XGBoost) provides effective prediction of PNO-DVT after pelvic fracture surgery. Key predictors such as age, intraoperative blood loss, and BMI enable accurate risk stratification and support early preventive intervention.
METHODS: Clinical data from 745 patients treated between 2016 and 2019 were analyzed. Demographic characteristics, laboratory indicators, surgical variables, and scoring systems were collected. Univariate logistic regression, LASSO regression, and multivariate logistic regression identified 12 independent risk factors. The dataset was split 7:3 into training and test sets. Six machine learning models (logistic regression, SVM, random forest, XGBoost, LightGBM, and AdaBoost) were constructed.
RESULTS: Univariate analysis identified GGT, HDL-C, ApoB, TCO2, GLU, GAP, EOS, and MPV as potential predictors. Multivariate analysis confirmed age, nephropathy, BMI, fracture reduction, intraoperative blood loss, red blood cell transfusion, HDL-C, ApoB, GAP, and MPV as independent risk factors. Among all models, XGBoost achieved the highest AUC (0.8633), demonstrating superior predictive performance. Model accuracy ranged from 0.6502 to 0.9283.
CONCLUSION: Machine learning (particularly XGBoost) provides effective prediction of PNO-DVT after pelvic fracture surgery. Key predictors such as age, intraoperative blood loss, and BMI enable accurate risk stratification and support early preventive intervention.