6自由度步态床旁预测老年患者跌倒风险的应用

doi:10.3877/cma.j.issn.1674-134X.2026.02.008

中华关节外科杂志(电子版) ›› 2026, Vol. 20 ›› Issue (02) : 194 -205. doi: 10.3877/cma.j.issn.1674-134X.2026.02.008

临床论著

6自由度步态床旁预测老年患者跌倒风险的应用

冼明桦¹^,², 刘华章⁴, 欧阳兰飞³, 谢珍艳²^,⁵, 钟国庆²^,⁵, 林金鹏²^,⁵, 黄星³, 黄帅⁴, 谢可乐²^,⁵, 曾东禹²^,⁵, 黄文汉²^,⁵, 李丽萍¹, 王旭平³, 梁会营⁴, 张余¹^,²^,⁵^,^†(

)

¹　515041　汕头大学公共卫生学院
²　510080　广州，南方医科大学附属广东省人民医院（广东省医学科学院）骨肿瘤科
³　210046　南京，泰康仙林鼓楼医院耳鼻喉头颈外科
⁴　510080　广州，南方医科大学附属广东省人民医院（广东省医学科学院）医学大数据中心
⁵　510080　广州，广东省骨缺损功能修复与生物材料工程技术研究中心

收稿日期:2025-10-29 出版日期:2026-04-01
通信作者: 张余
基金资助:
广州市重点研发计划(2023B01J0022); 广东省重点领域研发计划(2024B0101080001); 泰康仙林鼓楼医院院内重点项目(TKKYZD20240802); 南京市卫生科技发展专项资金资助项目(ZDXX25213)

Translating gait biomechanics into bedside fall risk prediction for hospitalized older adults

Minghua Xian¹^,², Huazhang Liu⁴, Lanfei Ouyang³, Zhenyan Xie²^,⁵, Guoqing Zhong²^,⁵, Jinpeng Lin²^,⁵, Xing Huang³, Shuai Huang⁴, Kele Xie²^,⁵, Dongyu Zeng²^,⁵, Wenhan Huang²^,⁵, Liping Li¹, Xuping Wang³, Huiying Liang⁴, Yu Zhang¹^,²^,⁵^,^†()

¹ School of Public Health, Shantou University, Shantou 515041, China
² Department of Orthopedic Oncology, Guangdong Provincial People’s Hospital (Guangdong Academy of Medical Sciences), Southern Medical University, Guangzhou 510080, China
³ Department of Otorhinolaryngology and Head and Neck Surgery, Taikang Xianlin Gulou Hospital, Nanjing 210046, China
⁴ Medical Big Data Center, Guangdong Provincial People’s Hospital (Guangdong Academy of Medical Sciences), Southern Medical University, Guangzhou 510080, China
⁵ Guangdong Engineering Technology Research Center of functional repair of bone defects and biomaterials, Guangzhou 510080, China

Received:2025-10-29 Published:2026-04-01
Corresponding author: Yu Zhang

全文 ( ) 下载PDF ( ) 导出引用

引用本文：

冼明桦, 刘华章, 欧阳兰飞, 谢珍艳, 钟国庆, 林金鹏, 黄星, 黄帅, 谢可乐, 曾东禹, 黄文汉, 李丽萍, 王旭平, 梁会营, 张余. 6自由度步态床旁预测老年患者跌倒风险的应用[J/OL]. 中华关节外科杂志(电子版), 2026, 20(02): 194-205.

Minghua Xian, Huazhang Liu, Lanfei Ouyang, Zhenyan Xie, Guoqing Zhong, Jinpeng Lin, Xing Huang, Shuai Huang, Kele Xie, Dongyu Zeng, Wenhan Huang, Liping Li, Xuping Wang, Huiying Liang, Yu Zhang. Translating gait biomechanics into bedside fall risk prediction for hospitalized older adults[J/OL]. Chinese Journal of Joint Surgery(Electronic Edition), 2026, 20(02): 194-205.

目的

探讨基于下肢6自由度（6DOF）步态运动学特征的深度学习模型在老年住院患者跌倒风险床旁预测中的应用价值。

方法

采用横断面研究设计，于广东省人民医院骨科病区纳入207例老年住院患者（年龄65~93岁）。纳入标准为：年龄≥60岁、意识清楚且能独立行走≥10 m；排除严重创伤或急性并发症无法完成测试者。采集受试者下肢6DOF步态运动学数据，并依据Morse跌倒量表（MFS）评分将受试者分为低风险、中风险及高风险组。将步态时序数据转化为二维步态图像后输入注意力增强的一维卷积神经网络（1D-CNN）模型进行训练与分类，二维步态图仅用于模型可视化展示。模型性能通过准确率（accuracy）、精确率（precision）、召回率（recall）及F1得分进行评估，并通过受试者工作特征曲线（ROC）及混淆矩阵进行分析。

结果

本研究共纳入207例患者，其中高风险59例、中风险72例、低风险76例。模型在独立测试集（n=41）上的分类准确率为0.878，精确率为0.874，召回率为0.897，F1值为0.882。高风险组召回率为1.000，未出现高危跌倒个体漏检。ROC曲线分析显示该模型在不同风险等级上均具有良好的区分能力。

结论

基于6DOF步态运动学特征构建的注意力增强1D-CNN模型可有效实现老年住院患者跌倒风险分层，并具有较高的预测准确性与临床可解释性，可为病房环境下的床旁跌倒风险筛查提供客观评估工具。

关键词: 步态, 跌倒, 深度学习, 老年人

Objective

To investigate the clinical application of a deep-learning model based on six-degree-of-freedom (6DOF) lower-limb gait kinematics for bedside prediction of fall risk in hospitalized older adults.

Methods

This cross-sectional study enrolled 207 hospitalized older adults (65 to 93 years) from the orthopedic wards of Guangdong Provincial People’s Hospital. Inclusion criteria: age≥60 years, clear consciousness, and the ability to walk independently for at least 10 m. Patients unable to complete gait testing due to severe trauma or acute complications were excluded. Lower-limb 6DOF gait kinematic data were collected using a motion-capture system. According to the Morse fall scale (MFS), participants were categorized into low-, medium-, and high-risk groups. Gait time-series data were transformed into two-dimensional gait maps and analyzed using an attention-enhanced one-dimensional convolutional neural network (1D-CNN). Model performance was evaluated using accuracy, precision, recall, and F1 score, and further assessed with receiver operating characteristic (ROC) curves and confusion matrix analysis.

Results

A total of 207 participants were included, comprising 59 high-risk, 72 medium-risk, and 76 low-risk individuals. On the independent test set (n=41), the model achieved an accuracy of 0.878, precision of 0.874, recall of 0.897, and an F1 score of 0.882. The recall for the high-risk group reached 1.000, indicating no missed high-risk fallers. ROC analysis demonstrated good discriminative ability across different risk levels.

Conclusions

The attention-enhanced 1D-CNN model based on 6DOF gait kinematics can effectively stratify fall risk in hospitalized older adults with high predictive performance and clinical interpretability. This approach provides an objective bedside tool for fall-risk screening in hospital settings.

Key words: Gait, Accidental falls, Deep learning, Aged

表1 入口统计学与人体测量特征组间比较

Table 1 Comparisons of demographic and anthropometric characteristics among groups

组别 Groups	例数 Number of cases	年龄［岁，（ $x ¯$ ±s）］ Age (year)	身高［cm，（ $x ¯$ ±s）］Height	体重［kg，（ $x ¯$ ±s）］Weight	BMI ［kg/m²，（ $x ¯$ ±s）］	性别［n（%）］ Gender
组别 Groups	例数 Number of cases	年龄［岁，（ $x ¯$ ±s）］ Age (year)	身高［cm，（ $x ¯$ ±s）］Height	体重［kg，（ $x ¯$ ±s）］Weight	BMI ［kg/m²，（ $x ¯$ ±s）］	男性 Male	女性 Female
低风险组 Low-risk	76	68±5	161±8	63±10	24±3	35（44.3）	41（32.0）
中风险组 Medium-risk	72	69±5	160±7	63±9	25±3	26（32.9）	46（35.9）
高风险组 High-risk	59	69±7	159±9	60±9	24±4	18（22.8）	41（32.0）
统计量 Statistical values		F=1.75	F=1.25	F=2.04	F=0.87	χ²=3.59
P值		>0.05	>0.05	>0.05	>0.05	>0.05

表1 入口统计学与人体测量特征组间比较

Table 1 Comparisons of demographic and anthropometric characteristics among groups

组别 Groups	例数 Number of cases	年龄［岁，（ $x ¯$ ±s）］ Age (year)	身高［cm，（ $x ¯$ ±s）］Height	体重［kg，（ $x ¯$ ±s）］Weight	BMI ［kg/m²，（ $x ¯$ ±s）］	性别［n（%）］ Gender
组别 Groups	例数 Number of cases	年龄［岁，（ $x ¯$ ±s）］ Age (year)	身高［cm，（ $x ¯$ ±s）］Height	体重［kg，（ $x ¯$ ±s）］Weight	BMI ［kg/m²，（ $x ¯$ ±s）］	男性 Male	女性 Female
低风险组 Low-risk	76	68±5	161±8	63±10	24±3	35（44.3）	41（32.0）
中风险组 Medium-risk	72	69±5	160±7	63±9	25±3	26（32.9）	46（35.9）
高风险组 High-risk	59	69±7	159±9	60±9	24±4	18（22.8）	41（32.0）
统计量 Statistical values		F=1.75	F=1.25	F=2.04	F=0.87	χ²=3.59
P值		>0.05	>0.05	>0.05	>0.05	>0.05

图1 受试者疾病类型分布

Figure 1 Distribution of disease types among participants

图2 预测模型构建过程

Figure 2 Workflow of model development

图3 模型内部结构

Figure 3 Structure of model

表2 模型的分类及评价

Table 2 Model classification andassessment

分类 Categories	样本量 Sample Size	准确性 Accuracy	精确度 Precision	召回得分 Recall	F1得分 F1 Score
低风险 Low-risk	17		0.875	0.824	0.848
中风险 Medium-risk	15		0.929	0.867	0.897
高风险 High-risk	9		0.818	1.000	0.900
整体模型 Overall model	41	0.878	0.874	0.897	0.882

表2 模型的分类及评价

Table 2 Model classification andassessment

分类 Categories	样本量 Sample Size	准确性 Accuracy	精确度 Precision	召回得分 Recall	F1得分 F1 Score
低风险 Low-risk	17		0.875	0.824	0.848
中风险 Medium-risk	15		0.929	0.867	0.897
高风险 High-risk	9		0.818	1.000	0.900
整体模型 Overall model	41	0.878	0.874	0.897	0.882

表3 模型方法的比较

Table 3 Comparisons on models

模型 Models	模型内容 Contents	准确度 Accuracy
A	Basic CNN	0.659
B	LSTM	0.61
C（所用模型）	Residual attention CNN (900)	0.878
D	Multi-model fusion	0.829

表3 模型方法的比较

Table 3 Comparisons on models

模型 Models	模型内容 Contents	准确度 Accuracy
A	Basic CNN	0.659
B	LSTM	0.61
C（所用模型）	Residual attention CNN (900)	0.878
D	Multi-model fusion	0.829

图4 模型训练与验证期间的损失与准确率变化趋势注：图中曲线包括训练集损失值（loss）、验证集损失值（val_loss）、训练集准确率（acc）和验证集准确率（val_acc）；X轴为训练轮次，Y轴为对应的损失值或准确率

Figure 4 The trend of loss and accuracy changes during model training and validation Note: the curves in the figure are training set loss value (loss), validation set loss value (val_loss), training set accuracy (acc), and validation set accuracy (val_cc); X-axis represents the training epochs, while Y-axis represents the corresponding loss value or accuracy

图5 模型在测试集上的混淆矩阵注：矩阵的行代表真实类别，列代表模型预测类别；矩阵中的数值表示样本数量

Figure 5 The confusion matrix of the model on the test set Note: the rows of the matrix represent the true class, and the columns represent the predicted class of the model; the values in the matrix represent the number of samples

图6 Model DC（本模型）与其他基线模型在测试集上的ROC（受试者工作特征曲线）注：X轴为假阳性率（1-特异度），Y轴为真阳性率（灵敏度）

Figure 6 ROC (receiver operating characteristic curves) of model DC and other baseline models on the test set Note: X-axis represents the false positive rate (1-specificity), while the Y-axis represents the trueoositive rate (sensitivity)

图7 模型的特征可视化结果。图A为DOF1（内收/外展自由度）的特征热图；图B为DOF3（屈曲/伸展自由度）的特征热图注：每个子图上半部分为原始步态信号，下半部分由注意力热图（红色条带）和提取特征（蓝色曲线）构成，红色条带的深浅反映了该时间步特征对模型分类的重要性

Figure 7 Visualization of model features. A is the characteristic heat map of DOF1 (inward/outward degrees of freedom); B is the characteristic heat map of DOF3 (degree of freedom in flexion/extension) Note: the upper half of each subgraph is the original gait signal, while the lower half consists of an attention heatmap (red stripe) and extracted features (blue curve), the depth of the red stripe reflects the importance of the time step feature for model classification

［1］	Scott AJ. The longevity society［J］. Lancet Healthy Longev, 2021, 2（12）: e820-e827.
［2］	Li W, Keegan THM, Sternfeld B, et al. Outdoor falls among middle-aged and older adults: a neglected public health problem［J］. Am J Public Health, 2006, 96（7）: 1192-1200.
［3］	Igual R, Medrano C, Plaza I. Challenges, issues and trends in fall detection systems［J/OL］. Biomed Eng Online, 2013, 12: 66. DOI: 10.1186/1475-925X-12-66.
［4］	Akyol AD. Falls in the elderly: what can be done?［J］. Int Nurs Rev, 2007, 54（2）: 191-196.
［5］	Tanaka B, Sakuma M, Ohtani M, et al. Incidence and risk factors of hospital falls on long-term care wards in Japan［J］. J Eval Clin Pract, 2012, 18（3）: 572-577.
［6］	de Almeida Abreu HC, Almeida Oliveira Reiners A, de Souza Azevedo RC, et al. Incidence and predicting factors of falls of older inpatients［J/OL］. Rev Saúde Pública, 2015, 49: 37. DOI: 10.1590/s0034-8910.2015049005549.
［7］	Qian XX, Chau PH, Fong DYT, et al. Post-hospital falls among the older population: the temporal pattern in risk and healthcare burden［J］. J Am Med Dir Assoc, 2023, 24（10）: 1478-1483.e2.
［8］	Harrington L, Luquire R, Vish N, et al. Meta-analysis of fall-risk tools in hospitalized adults［J］. J Nurs Adm, 2010, 40（11）: 483-488.
［9］	Chow SKY, Lai CKY, Wong TKS, et al. Evaluation of the Morse Fall Scale: applicability in Chinese hospital populations［J］. Int J Nurs Stud, 2007, 44（4）: 556-565.
［10］	Shimada H, Hirata R, Katsuki NE, et al. Validation of a fall predictive model for inpatients in Japanese long term care hospitals［J］. Int J Med Sci, 2025, 22（12）: 2877-2883.
［11］	Mao A, Su J, Ren M, et al. Risk prediction models for falls in hospitalized older patients: a systematic review and meta-analysis［J/OL］. BMC Geriatr, 2025, 25（1）: 29. DOI: 10.1186/s12877-025-05688-0.
［12］	Bae J-H, Seo J-W, Kim DY. Deep-learning model for predicting physical fitness in possible sarcopenia: analysis of the Korean physical fitness award from 2010 to 2023［J/OL］. Front Public Health, 2023, 11: 1241388. DOI: 10.3389/fpubh.2023.1241388.
［13］	Lindberg DS, Prosperi M, Bjarnadottir RI, et al. Identification of important factors in an inpatient fall risk prediction model to improve the quality of care using EHR and electronic administrative data: a machine-learning approach［J/OL］. Int J Med Inform, 2020, 143: 104272. DOI: 10.1016/j.ijmedinf.2020.104272.
［14］	Tunca C, Salur G, Ersoy C. Deep learning for fall risk assessment with inertial sensors: utilizing domain knowledge in spatio-temporal gait parameters［J］. IEEE J Biomed Health Inform, 2020, 24（7）: 1994-2005.
［15］	Kelly D, Condell J, Gillespie J, et al. Improved screening of fall risk using free-living based accelerometer data［J/OL］. J Biomed Inform, 2022, 131: 104116. DOI: 10.1016/j.jbi.2022.104116.
［16］	Lockhart TE, Soangra R, Yoon H, et al. Prediction of fall risk among community-dwelling older adults using a wearable system［J/OL］. Sci Rep, 2021, 11: 20976. DOI: 10.1038/s41598-021-00458-5.
［17］	Kokkotis C, Apostolidis K, Menychtas D, et al. Explainable Siamese neural networks for detection of high fall risk older adults in the community based on gait analysis［J/OL］. J Funct Morphol Kinesiol, 2025, 10（1）: 73. DOI: 10.3390/jfmk10010073.
［18］	Adeli V, Korhani N, Sabo A, et al. Ambient monitoring of gait and machine learning models for dynamic and short-term falls risk assessment in people with dementia［J］. IEEE J Biomed Health Inform, 2023, 27（7）: 3599-3609.
［19］	Hommen JM, Batista JP, Bollheimer LC, et al. Movement patterns during gait initiation in older adults with various stages of frailty: a biomechanical analysis［J/OL］. Eur Rev Aging Phys Act, 2024, 21（1）: 1. DOI: 10.1186/s11556-024-00335-w.
［20］	Ullrich M, Roth N, Küderle A, et al. Fall risk prediction in Parkinson’s disease using real-world inertial sensor gait data［J］. IEEE J Biomed Health Inform, 2023, 27（1）: 319-328.
［21］	赵钰婷, 张景焜, 杨轲欣, 等. 认知与动作双任务对肌少症老年人步态特征及动态稳定性的影响［J/OL］. 医用生物力学, 2025: 1-15.
［22］	Mehdizadeh S, Van Ooteghem K, Gulka H, et al. A systematic review of center of pressure measures to quantify gait changes in older adults［J/OL］. Exp Gerontol, 2021, 143: 111170. DOI: 10.1016/j.exger.2020.111170.
［23］	Roshdibenam V, Jogerst GJ, Butler NR, et al. Machine learning prediction of fall risk in older adults using timed up and go test kinematics［J/OL］. Sensors, 2021, 21（10）: 3481. DOI: 10.3390/s21103481.
［24］	Wang S, Pai Y-C, Bhatt T. Neuromuscular mechanisms of motor adaptation to repeated gait-slip perturbations in older adults［J/OL］. Sci Rep, 2022, 12（1）: 19851. DOI: 10.1038/s41598-022-23051-w.
［25］	Yin L, Chen P, Xu J, et al. Validity and reliability of inertial measurement units for measuring gait kinematics in older adults across varying fall risk levels and walking speeds［J/OL］. BMC Geriatr, 2025, 25（1）: 336. DOI: 10.1186/s12877-025-05993-8.
［26］	Nouredanesh M, Kukreja SL, Tung J. Detection of compensatory balance responses using wearable electromyography sensors for fall-risk assessment［C］//2016 38th Annual International Conference of the IEEE Engineeringin Medicineand Biology Society（EMBC）. August 16-20, 2016, Orlando, FL, USA. IEEE, 2016: 1680-1683.
［27］	Nouredanesh M, Tung J. IMU, sEMG, or their cross-correlation and temporal similarities: Which signal features detect lateral compensatory balance reactions more accurately?［J/OL］. Comput Meth Programs Biomed, 2019, 182: 105003. DOI: 10.1016/j.cmpb.2019.105003.
［28］	Savadkoohi M, Oladunni T, Thompson LA. Deep neural networks for human’s fall-risk prediction using force-plate time series signal［J/OL］. Expert Syst Appl, 2021, 182. DOI: 10.1016/j.eswa.2021.115220.
［29］	Caicedo PE, Rengifo CF, Rodriguez LE, et al. Dataset for gait analysis and assessment of fall risk for older adults［J/OL］. Data Brief, 2020, 33: 106550. DOI: 10.1016/j.dib.2020.106550.
［30］	张余, 黄文汉, 曾小龙, 等. 从“步态图” 学会读懂你的膝［J/OL］. 中华关节外科杂志（电子版）, 2020, 14（1）: 1-3.
［31］	Lim ZK, Connie T, Goh MKO, et al. Fall risk prediction using temporal gait features and machine learning approaches［J/OL］. Front Artif Intell, 2024, 7: 1425713. DOI: 10.3389/frai.2024.1425713.
［32］	Wang S, Nguyen TK, Bhatt T. Trip-related fall risk prediction based on gait pattern in healthy older adults: amachine-learning approach［J/OL］. Sensors, 2023, 23（12）: 5536. DOI: 10.3390/s23125536.
［33］	Windolf M, Götzen N, Morlock M. Systematic accuracy and precision analysis of video motion capturing systems-exemplified on the Vicon-460 system［J］. J Biomech, 2008, 41（12）: 2776-2780.
［34］	Benoit DL, Ramsey DK, Lamontagne M, et al. Effect of skin movement artifact on knee kinematics during gait and cutting motions measured in vivo［J］. Gait Posture, 2006, 24（2）: 152-164.
［35］	Fung PW, Mok KM, Leow RS, et al. Knee kinematics of ACL-deficient patients: a development of a portable motion analysis system［J］. Jhse, 2018, 13（4）: 832-842. DOI: 10.14198/jhse.2018.134.11.
［36］	Mundt M, Thomsen W, David S, et al. Assessment of the measurement accuracy of inertial sensors during different tasks of daily living［J］. J Biomech, 2019, 84: 81-86.
［37］	Hollingworth AL, Hutchinson JB, Bechtel JR, et al. Knee deep in Error: Unraveling the accuracy limitations of consumer-grade wearable motion sensors with limited degrees of freedom computing knee joint angles［J/OL］. ComputBiolMed, 2025, 193: 110469. DOI: 10.1016/j.compbiomed.2025.110469.
［38］	Zeng X, Zhu Y, Xie Z, et al. 3D knee kinematic parameters effectively diagnose knee osteoarthritis and assess its therapeutic strategy［J/OL］. Adv Intell Syst, 2022, 4（6）: 2100161. DOI: 10.1002/aisy.202100161.
［39］	Zhong G, Huang S, Zhang Z, et al. Diagnosis of generalized joint hypermobility with gait patterns using a deep neural network［J/OL］. Comput Biol Med, 2023, 164: 107360. DOI: 10.1016/j.compbiomed.2023.107360.
［40］	Wang G, Zeng X, Lai G, et al. Efficient subject-independent detection of anterior cruciate ligament deficiency based on marine predator algorithm and support vector machine［J］. IEEE J Biomed Health Inform, 2022, 26（10）: 4936-4947.
［41］	Zhang Y, Yao Z, Wang S, et al. Motion analysis of Chinese normal knees during gait based on a novel portable system［J］. Gait Posture, 2015, 41（3）: 763-768.
［42］	Zhang Y, Huang W, Yao Z, et al. Anterior cruciate ligament injuries alter the kinematics of knees with or without meniscal deficiency［J］. Am J Sports Med, 2016, 44（12）: 3132-3139.
［43］	王文锦, 田斐, 李柠薇, 等. 新型膝关节运动分析系统的研制及临床应用［J/OL］. 中华关节外科杂志（电子版）, 2020, 14（1）: 78-84.
［44］	Elfring R, de la Fuente M, Radermacher K. Assessment of optical localizer accuracy for computer aided surgery systems［J］. Comput Aided Surg, 2010, 15（1－3）: 1-12.
［45］	Wang S, Zeng X, Huangfu L, et al. Validation of a portable marker-based motion analysis system［J/OL］. J Orthop Surg Res, 2021, 16（1）: 425. DOI: 10.1186/s13018-021-02576-2.
［46］	Cruz-Jimenez M. Normal changes in gait and mobility problems in the elderly［J］. Phys Med Rehabil Clin N Am, 2017, 28（4）: 713-725.
［47］	Morse JM, Black C, Oberle K, et al. A prospective study to identify the fall-prone patient［J］. Soc Sci Med, 1989, 28（1）: 81-86.
［48］	Brodie MA, Coppens MJ, Ejupi A, et al. Comparison between clinical gait and daily-life gait assessments of fall risk in older people［J］. Geriatr Gerontol Int, 2017, 17（11）: 2274-2282.
［49］	McFarlane-Kolb H. Falls risk assessment, multitargeted interventions and the impact on hospital falls［J］. Int J Nurs Pract, 2004, 10（5）: 199-206.
［50］	Liu L, Hou Y, He J, et al. An energy-efficient fall detection method based on FD-DNN for elderly people［J/OL］. Sensors, 2020, 20（15）: 4192. DOI: 10.3390/s20154192.
［51］	Han T, Kang W, Choi G. IR-UWB sensor based fall detection method using CNN algorithm［J/OL］. Sensors, 2020, 20（20）: 5948. DOI: 10.3390/s20205948.
［52］	Zhang J, Li Z, Liu Y, et al. An effective deep learning framework for fall detection: model development and study design［J/OL］. J Med Internet Res, 2024, 26: e56750. DOI: 10.2196/56750.
［53］	Wang S, Miranda F, Wang Y, et al. Near-fall detection in unexpected slips during over-ground locomotion with body-worn sensors among older adults［J/OL］. Sensors, 2022, 22（9）: 3334. DOI: 10.3390/s22093334.
［54］	Zhou Z, Liu Z, Liu Y, et al. TCS-Fall: Cross-individual fall detection system based on channel state information and time-continuous stack method［J/OL］. Digit Health, 2024, 10: 20552076241259047. DOI: 10.1177/20552076241259047.
［55］	Maiora J, Rezola-Pardo C, García G, et al. Older adult fall risk prediction with deep learning and timed up and go（TUG）test data［J/OL］. Bioengineering, 2024, 11（10）: 1000. DOI: 10.3390/bioengineering11101000.
［56］	Lin J, Wang Y, Sha J, et al. Clinical reliability and validity of a video-based markerless gait evaluation method［J/OL］. Front Pediatr, 2023, 11: 1331176. DOI: 10.3389/fped.2023.1331176.
［57］	Sinha N, Chen KW. Novelty driven evolutionary neural architecture search［C］//Proceedings of the Genetic and Evolutionary Computation Conference Companion. Boston Massachusetts. ACM, 2022: 671-674.
［58］	Park G, Yi Y. CondNAS: neural architecture search for conditional CNNs［J/OL］. Electronics, 2022, 11（7）: 1101. DOI: 10.3390/electronics11071101.
［59］	Wang X, Wang X, Lv T, et al. HARNAS: human activity recognition based on automatic neural architecture search using evolutionary algorithms［J/OL］. Sensors, 2021, 21（20）: 6927. DOI: 10.3390/s21206927.
［60］	Ismail WN, Alsalamah HA, Hassan MM, et al. AUTO-HAR: an adaptive human activity recognition framework using an automated CNN architecture design［J］. Heliyon, 2023: e13636.

[1]	陈韵澄, 张仄骞, 江梓雯, 陈玉梅, 余淑萍, 梁晓瑜. 老年髋部骨折术后老老照护者社会支持与照顾负担分析[J/OL]. 中华关节外科杂志(电子版), 2026, 20(02): 185-193.
[2]	谢海波, 王志文, 李恒. 早期膝骨关节炎影像学检查与深度学习融合应用[J/OL]. 中华关节外科杂志(电子版), 2026, 20(01): 97-103.
[3]	李婷, 郭超, 李晨曦, 李柔演, 张一涵, 张雨晗, 陈琰, 斯琴高娃, 龚忠诚. 基于深度学习的上颌阻生尖牙自动分割与体积量化研究[J/OL]. 中华口腔医学研究杂志(电子版), 2026, 20(01): 9-16.
[4]	孟笑菲, 龚卓弘, 万耀文, 李沛达, 胡琪琪, 邱龙诗语, 刘恒毅, 谢伟丽. 深度学习与多模态大语言模型在笑线分类任务中的应用初探[J/OL]. 中华口腔医学研究杂志(电子版), 2026, 20(01): 17-24.
[5]	吕思怡, 王琰琪, 仇珺, 陈宇江, 高洁. 人工智能图像处理技术在口腔医学中的应用[J/OL]. 中华口腔医学研究杂志(电子版), 2026, 20(01): 40-46.
[6]	叶佳慧, 张建薇, 葛兆霞, 沈艳婷, 马慧珍, 房芳, 张寅. 完全植入式输液港在高龄非肿瘤患者中的长期安全性及风险因素分析[J/OL]. 中华普外科手术学杂志(电子版), 2026, 20(02): 191-194.
[7]	黄楚曦, 吴卓. 人工智能在膀胱癌影像中的应用进展[J/OL]. 中华腔镜泌尿外科杂志(电子版), 2026, 20(03): 241-247.
[8]	梅昊楠, 杨瑞, 刘修恒. 人工智能辅助病理学图像分析在前列腺癌诊断中的研究进展[J/OL]. 中华腔镜泌尿外科杂志(电子版), 2026, 20(01): 1-7.
[9]	张迪, 王素美, 陈阳, 熊霞鹂, 梁兵, 白雁飞, 郑雪云, 李华. 老年女性盆腔器官脱垂手术患者合并症分布及多学科诊疗围手术期管理效果分析[J/OL]. 中华疝和腹壁外科杂志(电子版), 2026, 20(02): 206-212.
[10]	唐玥, 陈家璐, 覃德龙, 李宗龙, 汤朝晖, 全志伟. 基于AI的多模态影像在肝癌诊治中应用及面临挑战[J/OL]. 中华肝脏外科手术学电子杂志, 2026, 15(01): 4-9.
[11]	张忠, 王喆, 黄其日麦拉图, 齐岩松. 人工智能在肘关节骨关节炎诊疗中的应用进展[J/OL]. 中华肩肘外科电子杂志, 2026, 14(02): 109-114.
[12]	陈思潮, 姬涛, 吴屹冰, 郭明君, 甘永雄, 苏炜, 魏鹏, 宋庆华. 组配式半骨盆假体治疗老年髋臼周围转移性肿瘤患者的临床疗效评估[J/OL]. 中华老年骨科与康复电子杂志, 2026, 12(02): 79-84.
[13]	王晓伟, 杨红梅, 孙天胜, 刘智, 刘川, 高杰. 髓内钉治疗老年股骨转子间骨折术后内固定相关并发症预测模型建立与验证[J/OL]. 中华老年骨科与康复电子杂志, 2026, 12(01): 38-46.
[14]	陈小坤, 杜顺达. 影像组学在肝细胞癌中的应用进展及挑战[J/OL]. 中华消化病与影像杂志(电子版), 2026, 16(02): 97-100.
[15]	肖骥峰, 廖兴志, 汤燕彬, 陈湉, 施冬冬, 乔燕. 环泊酚复合小剂量艾司氯胺酮在老年患者腹腔镜胆囊切除术中的麻醉效果及术后认知功能影响[J/OL]. 中华消化病与影像杂志(电子版), 2026, 16(02): 179-184.

阅读次数

全文

摘要

选择文件类型/文献管理软件名称

选择包含的内容

Translating gait biomechanics into bedside fall risk prediction for hospitalized older adults

模态框（Modal）标题

选择文件类型/文献管理软件名称

选择包含的内容

Translating gait biomechanics into bedside fall risk prediction for hospitalized older adults