Supplementary Material for: AI-Driven Fall Prediction Across Generations: Integrating Deep Learning and Machine Learning for Young, Middle-Aged, and Older Adults

Introduction Falls occur in all age groups and represent a significant public health concern. Previous studies have implemented artificial intelligence (AI), including machine learning (ML) and deep learning (DL) algorithms for fall risk prediction, but the comparative performance between models and the applicability for younger populations remains unclear. This study aims to develop and compare different ML/DL models and identify key predictive features across age groups. Methods We enrolled 1441 community-dwelling adults aged over 20 years in southern Taiwan and collected demographic, clinical, and physical performance data. Participants were categorized based on fall history. Five ML models (KNN, RF, GBDT, XGBoost, and CatBoost) and two DL (GRU, AGRU) models were trained and evaluated using accuracy, precision, recall, F1-score, and the area under the receiver operating characteristic curve (AUROC). Feature importance was interpreted using SHapley Additive exPlanations (SHAP) values in the best-performing model. Age-stratified subgroup analyses were conducted for groups aged 20-45, 46-65, and >65 years. Results The AGRU model achieved the highest accuracy (91.39%) and AUROC (0.934) in the overall group and outperformed other models across all subgroups. Feature importance analysis revealed pulse rate, living alone, systolic blood pressure, 5-times Sit-to-Stand test, and sex as major predictors of falls in the overall group. The top five predicting factors varied across age groups. Conclusion We developed a robust and interpretable DL model for identifying fall risk across different age groups. Age-specific risk factors highlight the need for tailored preventive strategies. External validation using an independent dataset demonstrated moderate generalizability. Larger and more diverse datasets for validation and integration of sequential or sensor-based data are essential for practical applications.

引言 跌倒事件在所有年龄群体中均有发生，是一项重大的公共卫生关切问题。既往研究已将人工智能（AI）、机器学习（ML）及深度学习（DL）算法应用于跌倒风险预测，但不同模型间的性能对比以及其在年轻群体中的适用性仍不明确。本研究旨在开发并对比不同的机器学习/深度学习模型，并明确不同年龄群体中的关键预测特征。研究方法 本研究纳入了中国台湾南部1441名年龄20岁以上的社区常住成年人，收集其人口统计学、临床及身体机能相关数据。研究对象根据跌倒史进行分组。本研究训练并评估了5种机器学习模型（K近邻（KNN）、随机森林（RF）、梯度提升决策树（GBDT）、极端梯度提升（XGBoost）及分类梯度提升树（CatBoost））与2种深度学习模型（门控循环单元（GRU）、注意力门控循环单元（AGRU）），评估指标包括准确率、精确率、召回率、F1值以及受试者工作特征曲线下面积（AUROC）。针对表现最优的模型，采用沙普利可加解释（SHAP）值对特征重要性进行解读。本研究针对20~45岁、46~65岁及65岁以上三个年龄分层亚组开展亚组分析。结果 注意力门控循环单元（AGRU）模型在全人群中取得了最高的准确率（91.39%）与受试者工作特征曲线下面积（0.934），且在所有亚组中均优于其他模型。特征重要性分析显示，在全人群中，脉搏频率、独居状态、收缩压、5次坐站试验结果以及性别为跌倒的主要预测因素。不同年龄群体的前五大预测因素存在差异。结论 本研究开发了一款稳健且可解释的深度学习模型，可用于识别不同年龄群体的跌倒风险。针对不同年龄的特异性风险因素表明，需制定个性化的预防策略。采用独立数据集开展的外部验证显示，该模型具备中等程度的泛化能力。若要实现实际应用，需使用规模更大、多样性更强的数据集开展验证，并整合时序数据或基于传感器的相关数据。