BMC gastroenterology · 2025
Machine learning-based disease risk stratification and prediction of metabolic dysfunction-associated fatty liver disease using vibration-controlled transient elastography: Result from NHANES 2021-2023.
СтатьяНазвание 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.
Huang Liqiong, Luo Yu, Zhang Li, Wu Mengqi, Hu Lirong
Аннотация
BACKGROUND: Metabolic dysfunction-associated fatty liver disease (MAFLD) is a common chronic liver disease and represents a significant public health issue. Nevertheless, current risk stratification methods remain inadequate. The study aimed to use machine learning in the identification of significant features and the development of a predictive model to determine its usefulness in discrimination of MAFLD's risk stratification (low, moderate, and high) in adults.
METHODS: The data of the 2021-2023 NHANES database were analyzed. Vibration-controlled transient elastography measurements, including controlled attenuation parameter for the evaluation of steatosis and liver stiffness for the evaluation of fibrosis, were used for risk stratification. The participants were grouped into low-risk, moderate-risk, and high-risk groups based on specific criteria. Feature selection was conducted through Least Absolute Shrinkage and Selection Operator (LASSO) regression and random forest classification.
RESULTS: A total of 4,227 participants were included in the study. There were 16 significant predictors identified by LASSO regression, among which the top 10 predictors were demographic (age, gender, race, hypertension history), clinical (body mass index, waist circumference, hemoglobin, glycohemoglobin, lymphocyte count), and education level. The area under the receiver operating characteristic curve (AUC) of the random forest model in the validation set was 0.80, and the individual AUC was 0.83, 0.66 and 0.79 for the low-, moderate-, and high-risk groups, respectively.
CONCLUSION: Our machine learning model has excellent performance in stratification of risk for MAFLD with readily available clinical and demographic parameters. This model could be employed as a valuable screening tool to refer high-risk patients for further hepatological evaluation.
METHODS: The data of the 2021-2023 NHANES database were analyzed. Vibration-controlled transient elastography measurements, including controlled attenuation parameter for the evaluation of steatosis and liver stiffness for the evaluation of fibrosis, were used for risk stratification. The participants were grouped into low-risk, moderate-risk, and high-risk groups based on specific criteria. Feature selection was conducted through Least Absolute Shrinkage and Selection Operator (LASSO) regression and random forest classification.
RESULTS: A total of 4,227 participants were included in the study. There were 16 significant predictors identified by LASSO regression, among which the top 10 predictors were demographic (age, gender, race, hypertension history), clinical (body mass index, waist circumference, hemoglobin, glycohemoglobin, lymphocyte count), and education level. The area under the receiver operating characteristic curve (AUC) of the random forest model in the validation set was 0.80, and the individual AUC was 0.83, 0.66 and 0.79 for the low-, moderate-, and high-risk groups, respectively.
CONCLUSION: Our machine learning model has excellent performance in stratification of risk for MAFLD with readily available clinical and demographic parameters. This model could be employed as a valuable screening tool to refer high-risk patients for further hepatological evaluation.