Digestive and liver disease : official journal of the Italian Society of Gastroenterology and the Italian Association for the Study of the Liver · 2026
Metabolic factor-based machine learning model for mortality prediction in acute hepatitis E: Development and validation from a dual-center cohort.
СтатьяВалидационное исследованиеМногоцентровое исследованиеНазвание 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.
Fu Haoshuang, Song Shuying, Xiao Yuelin, Du Bingying, Zhao Gangde, Zhou Tianhui, Du Yanan
Аннотация
BACKGROUND: Hepatitis E virus (HEV) infection remains a major cause of liver failure with high short-term mortality, yet predictive models incorporating systemic metabolic factors are limited.
AIMS: We aimed to develop a machine learning model incorporating systemic metabolic parameters for mortality prediction in HEV patients.
METHODS: A total of 510 HEV patients from two medical centers were retrospectively enrolled and grouped into training, internal validation and external validation cohorts. Metabolic parameters (total cholesterol, HDL, LDL and diabetes) were integrated using SVM to generate a Metabolism Score. This score and other clinical variables (infection status, WBC, creatinine, age, platelet, albumin, TBIL, EGFR, INR, and GGT) were selected using LASSO regression. Five machine-learning models were developed to predict 28-day and 90-day mortality.
RESULTS: Among all models, logistic regression (LR) showed the best performance, with AUROCs of 0.87, 0.98, and 0.84 for 28-day mortality and 0.86, 0.88, and 0.89 for 90-day mortality across the training, validation, and external cohorts, respectively. The LR model outperformed MELD, demonstrated good calibration and net clinical benefits, and was visualized as a nomogram.
CONCLUSION: The LR model incorporating systemic metabolic factors accurately predicted short-term mortality in HEV patients and may facilitate early risk stratification and personalized management.
AIMS: We aimed to develop a machine learning model incorporating systemic metabolic parameters for mortality prediction in HEV patients.
METHODS: A total of 510 HEV patients from two medical centers were retrospectively enrolled and grouped into training, internal validation and external validation cohorts. Metabolic parameters (total cholesterol, HDL, LDL and diabetes) were integrated using SVM to generate a Metabolism Score. This score and other clinical variables (infection status, WBC, creatinine, age, platelet, albumin, TBIL, EGFR, INR, and GGT) were selected using LASSO regression. Five machine-learning models were developed to predict 28-day and 90-day mortality.
RESULTS: Among all models, logistic regression (LR) showed the best performance, with AUROCs of 0.87, 0.98, and 0.84 for 28-day mortality and 0.86, 0.88, and 0.89 for 90-day mortality across the training, validation, and external cohorts, respectively. The LR model outperformed MELD, demonstrated good calibration and net clinical benefits, and was visualized as a nomogram.
CONCLUSION: The LR model incorporating systemic metabolic factors accurately predicted short-term mortality in HEV patients and may facilitate early risk stratification and personalized management.