Clinical interventions in aging · 2025
Machine Learning-Based Prediction of Post-PKP Frailty: A Retrospective Cohort Study.
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.
Xu Dingjun, Fan Ziwei, Li Zhiyuan, Jia Mengxian, Fang Xiang, Shen Yizhe, Zhou Quan, Xie Changnan, Teng Honglin
Аннотация
BACKGROUND: Frailty and osteoporotic vertebral compression fractures (OVCFs) exhibit bidirectional causality, yet the impact of percutaneous kyphoplasty (PKP) on frailty progression remains unclear. This study developed machine learning (ML) models to predict post-PKP frailty and identify key predictors.
METHODS: A retrospective cohort of 4599 PKP patients was categorized into frailty/non-frailty groups based on two-year follow-up. Variables included preoperative baseline data, imaging parameters (fracture number/segments, Genant classification, T2 hyperintensity), clinical characteristics (osteoporosis severity, Visual Analogue Scale scores, residual low back pain [LBP]), and surgical details. After data splitting (4:1 ratio), features were selected to train and optimize ML models, with performance evaluated via area under the curve (AUC). The ML model with the best performance was selected as our final model while using it for external validation. SHAP analysis determined predictor contributions.
RESULTS: Key features (residual LBP, Genant classification, etc) informed model development. Hyperparameter optimization enhanced performance, with Extreme Gradient Boost achieving superior prediction (AUC 0.950, 95% CI 0.934-0.965). The model still maintains a good performance in the external test set, with an AUC of 0.845 (95% CI 0.805-0.884). SHAP identified residual LBP, Genant classification, and postoperative recumbency duration as top predictors.
CONCLUSION: ML models effectively predict post-PKP frailty, highlighting modifiable risk factors. Standardized anti-osteoporosis therapy, residual LBP prevention, and reduced postoperative recumbency may mitigate frailty risk.
METHODS: A retrospective cohort of 4599 PKP patients was categorized into frailty/non-frailty groups based on two-year follow-up. Variables included preoperative baseline data, imaging parameters (fracture number/segments, Genant classification, T2 hyperintensity), clinical characteristics (osteoporosis severity, Visual Analogue Scale scores, residual low back pain [LBP]), and surgical details. After data splitting (4:1 ratio), features were selected to train and optimize ML models, with performance evaluated via area under the curve (AUC). The ML model with the best performance was selected as our final model while using it for external validation. SHAP analysis determined predictor contributions.
RESULTS: Key features (residual LBP, Genant classification, etc) informed model development. Hyperparameter optimization enhanced performance, with Extreme Gradient Boost achieving superior prediction (AUC 0.950, 95% CI 0.934-0.965). The model still maintains a good performance in the external test set, with an AUC of 0.845 (95% CI 0.805-0.884). SHAP identified residual LBP, Genant classification, and postoperative recumbency duration as top predictors.
CONCLUSION: ML models effectively predict post-PKP frailty, highlighting modifiable risk factors. Standardized anti-osteoporosis therapy, residual LBP prevention, and reduced postoperative recumbency may mitigate frailty risk.