一种用于部分负重行走期间实时监测胫骨力的可穿戴监测系统

Wearable Monitoring System for Real-Time Tibial Force Monitoring During Partial Weight-Bearing Walking

摘要

胫骨骨折患者通常采用部分负重(PWB)的方式进行康复治疗．PWB的关键是控制负荷，负重不足会导致愈合延迟，过度负重会导致二次骨折．因此，康复行走期间对胫骨受力的实时监测和负重范围的精准控制至关重要．现有受力监测设备存在无法实时监测或监测不准确的问题，为此，通过惯性测量单元(IMU)和柔性足底压力鞋垫搭建可穿戴监测系统，能够实时采集下肢运动学与动力学数据，并通过振动/语音反馈提供实时负荷预警；在此基础上，基于可穿戴监测系统分别提出了逆动力学模型(物理驱动)、多层感知机预测模型(静态映射)及双层长短期记忆(LSTM)预测模型(时序建模)，实现对胫骨轴向力的动态预测．为验证系统预测性能，招募6名受试者模拟4种PWB行走实验，并对比逆动力学模型、多层感知机模型及双层LSTM 模型的预测结果．结果表明：双层LSTM 模型在不同负重任务下的预测表现优异，平均绝对误差在(0.04±0.02)～(0.08±0.05) BW间，均方根误差在(0.04±0.03)～(0.11±0.04) BW之间，预测结果与参考值拟合程度最高．可穿戴监测系统的多模态传感器方案与LSTM时序建模的协同作用，能够有效捕捉步态周期内的时序动态特征，提高预测的准确性．这套便携式可穿戴监测系统能够为下肢胫骨骨折患者提供有效的康复指导，具有重要临床意义．

Abstract

Patients with tibial fractures typically undergo partial weight-bearing(PWB)rehabilitation. The key of PWB lies in precise load control，as insufficient loading may delay healing，while excessive loading risks secondary fractures. Therefore，real-time monitoring of tibial forces and accurate regulation of weight-bearing ranges are critical during rehabilitation. The existing monitoring devices suffer from challenges in achieving real-time measurement and accuracy. To address this issue，a wearable monitoring system that integrates inertial measurement units and customized plantar pressure insoles had been developed to collect lower-limb kinematic and kinetic data in real time，with vibration/audio feedback providing instant load warnings. Building upon this system，three “tibial axial force” prediction models were proposed：an inverse dynamics model(physics-driven)，a multilayer perceptron model(static mapping)，and a two-layer long short-term memory(LSTM)model(temporal modeling). Six subjects were recruited to simulate four PWB walking conditions，validating the predictive performance of these models. Results demonstrate the superior performance of the two-layer LSTM model across tasks，achieving a mean absolute error ranging from (0.04±0.02) to (0.08±0.05) body weight(BW)and a root mean square error ranging from (0.04±0.03) to (0.11±0.04) BW，with the highest correlation to reference values. The synergistic integration of the multimodal sensor configuration and LSTM-based temporal modeling results in the effective capturing of temporal dynamic characteristics within gait cycles，significantly increasing the prediction accuracy. This portable，wearable monitoring system offers a reliable solution for guiding tibial fracture rehabilitation，showing a substantial clinical value.

关键词

胫骨力 / 可穿戴监测系统 / 部分负重 / 双层长短期记忆神经网路

Key words

tibial force / wearable monitoring system / partial weight-bearing / two-layer long short-term memory(LSTM) neural network

参考文献

[1] Shen F Y, Chen B L, Guo Q S, et al. Augmented reality patient—specific reconstruction plate design for pelvic and acetabular fracture surgery[J]. International Journal of Computer Assisted Radiology and Surgery, 2013, 8:169-179.

[2] Hurkmans H L, Bussmann J B, Benda E, et al. Effective—ness of audio feedback for partial weight—bearing in and outside the hospital：A randomized controlled trial[J]. Archives of Physical Medicine and Rehabilitation, 2012, 93(4):565-570.

[3] Hustedt J W, Blizzard D J, Baumgaertner M R, et al. Current advances in training orthopaedic patients to comply with partial weight—bearing instructions[J]. The Yale Journal of Biology and Medicine, 2012, 85(1):119-125.

[4] Houben I B, Raaben M, van Basten Batenburg M, et al. Delay in weight bearing in surgically treated tibial shaft fractures is associated with impaired healing：A cohort analysis of 166 tibial fractures[J]. European Journal of Orthopaedic Surgery & Traumatology, 2018, 28:1429-1436.

[5] Mavčič B, Antolič V. Optimal mechanical environment of the healing bone fracture/osteotomy[J]. International Orthopaedics, 2012, 36:689-695.

[6] Komi P V. Relevance of in vivo force measurements to human biomechanics[J]. Journal of Biomechanics, 1990, 23：23-34.

[7] Tveit M, Kärrholm J. Low effectiveness of prescribed partial weight bearing：Continuous recording of vertical loads using a new pressure—sensitive insole[J]. Journal of Rehabilitation Medicine, 2001, 33(1):42-46.

[8] Raaben M, Holtslag H R, Leenen L P H, et al. Real—time visual biofeedback during weight bearing improves therapy compliance in patients following lower extremity fractures[J]. Gait & Posture, 2018, 59:206-210.

[9] Matijevich E S, Branscombe L M, Scott L R, et al. Ground reaction force metrics are not strongly correlated with tibial bone load when running across speeds and slopes：Implications for science，sport and wearable tech[J]. PLOS One, 2019, 14(1):e0210000.

[10] Elstub L J, Nurse C A, Grohowski L M, et al. Tibial bone forces can be monitored using shoe—worn wearable sensors during running[J]. Journal of Sports Sciences, 2022, 40(15):1741-1749.

[11] Dorschky E, Nitschke M, Martindale C F, et al. CNN—based estimation of sagittal plane walking and running biomechanics from measured and simulated inertial sensor data[J]. Frontiers in Bioengineering and Biotechnology, 2020, 8:604.

[12] Ma T, Xu X, Chai Z W, et al. A wearable biofeedback device for monitoring tibial load during partial weight—bearing walking[J]. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 2023, 31:3428-3436.

[13] Matijevich E S, Scott L R, Volgyesi P, et al. Combining wearable sensor signals，machine learning and biomechanics to estimate tibial bone force and damage during running[J]. Human Movement Science, 2020, 74:102690.

[14] Liu J X, Wang M F, Wang P, et al. Cost—efficient flexible supercapacitive tactile sensor with superior sensitivity and high spatial resolution for human—robot interaction[J]. IEEE Access, 2020, 8:64836-64845.

[15] Rüedi T P, Murphy W M, Colton C L, et al. AO Principles of Fracture Management[M]. Stuttgart： Thieme, 2000.

[16] Wehner T, Claes L, Simon U. Internal loads in the human tibia during gait[J]. Clinical Biomechanics, 2009, 24(3):299-302.

[17] Romani W A, Gieck J H, Perrin D H, et al. Mechanisms and management of stress fractures in physically active persons[J]. Journal of Athletic Training, 2002, 37(3):306-314.

[18] Farris D J, Sawicki G S. Human medial gastrocnemius force—velocity behavior shifts with locomotion speed and gait[J]. Proceedings of the National Academy of Sciences, 2012, 109(3):977-982.

[19] Hochreiter S, Schmidhuber J. Long short—term memory[J]. Neural Computation, 1997, 9(8):1735-1780.

[20] Zadpoor A A, Nikooyan A A. The relationship between lower—extremity stress fractures and the ground reaction force：A systematic review[J]. Clinical Biomechanics, 2011, 26(1):23-28.

[21] van der Worp H, Vrielink J W, Bredeweg S W. Do runners who suffer injuries have higher vertical ground reaction forces than those who remain injury—free? A systematic review and meta—analysis[J]. British Journal of Sports Medicine, 2016, 50(8):450-457

基金资助

国家自然科学基金资助项目(62027812)