International journal of medical informatics · 2025
Comparison of machine learning and logistic regression models for predicting emergence delirium in elderly patients: A prospective study.
СтатьяСравнительное исследованиеПоддержка исследования: вне правительства СШАНазвание и аннотация: Перевод в очереди
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.
Lu Yufan, Li Ying, Chi Shengqiang, Feng Yan, Li Gaowei, Lin Xuezheng, Jin Jie, Wang Ying
Аннотация
OBJECTIVE: To compare the performance of machine learning and logistic regression algorithms in predicting emergence delirium (ED) in elderly patients.
METHODS: A prospective study was carried out in a Chinese teaching tertiary hospital and collected the details of 1045 patients who underwent noncardiac surgery with general anesthesia. Characteristic variables related to ED were selected by least absolute shrinkage and selection operator (LASSO). Finally, seven machine learning models (gradient boosting machine, extreme gradient boosting, light gradient boosting machine, support vector machine, decision tree, neural network, and random forest) and logistic regression were used in the training set, and the predictive performance of the models was validated in the test set.
RESULTS: ED was identified in 316 (30.2%) patients. The logistic regression model performed better than the machine learning models (area under the curve [AUC] of 0.790, 95% confidence interval [CI] 0.736-0.843). Besides, the calibration curve indicated good consistency between predicted and actual ED probabilities, and decision curve analysis demonstrated that the logistic regression model could bring clinical benefits.
CONCLUSION: The optimal application of logistic regression can provide rapid and efficient risk prediction of ED for medical workers so that reasonable prevention and treatment measures can be taken.
METHODS: A prospective study was carried out in a Chinese teaching tertiary hospital and collected the details of 1045 patients who underwent noncardiac surgery with general anesthesia. Characteristic variables related to ED were selected by least absolute shrinkage and selection operator (LASSO). Finally, seven machine learning models (gradient boosting machine, extreme gradient boosting, light gradient boosting machine, support vector machine, decision tree, neural network, and random forest) and logistic regression were used in the training set, and the predictive performance of the models was validated in the test set.
RESULTS: ED was identified in 316 (30.2%) patients. The logistic regression model performed better than the machine learning models (area under the curve [AUC] of 0.790, 95% confidence interval [CI] 0.736-0.843). Besides, the calibration curve indicated good consistency between predicted and actual ED probabilities, and decision curve analysis demonstrated that the logistic regression model could bring clinical benefits.
CONCLUSION: The optimal application of logistic regression can provide rapid and efficient risk prediction of ED for medical workers so that reasonable prevention and treatment measures can be taken.