台灣人口結構正快速轉變準備進入超高齡社會，因此高齡者健康和安全會是重要的議題。其中跌倒是最常見，且會造成高齡者身心損傷以及後續長期醫療保健系統的需求。而現今台灣臨床上判定是否有跌倒風險通常是使用伯格平衡量表（Berg Balance Scale, BBS），但其主要依靠醫療人員的經驗與主觀判斷而缺乏客觀的數據。
本研究為提供客觀之步態數據，徵求了9位年齡介於65至85歲之間的受測者，依醫療人員進行BBS的評估後分依照分數高低分為三組：A組為55分以上；B組為50-54分；C組為49分以下。受測者穿戴黏貼F-scan與4個Xsens dot（分別至於左右腳尖與腳跟）來蒐集足底壓力中心軌跡（Center of Pressure, COP）與加速度和角速度的步態數據。接著分析數據後整理出57個特徵來輸入支持向量機（Support Vector Machine, SVM）分類器來進行三組平衡分數的分類。
結果顯示擷取COP軌跡以及加速度與角速度之57項特徵，透過徑向基函數核（Radial basis function, RBF）支持向量機（RBF-SVM）可以得到97%的分類準確率。若只透過年齡這項特徵也可以得到約78%的準確率，可以反映BBS分數越低，跌倒風險越大的組別有年齡較高的傾向。而若加上COP左右方向之位移更可以得到約95%的準確率，反映出BBS分數越低，跌倒風險越大的組別有較大的COP左右方向位移之傾向。除此之外若藉由本研究找出的9項關鍵特徵進行分類更可得到99%的準確率。
本研究為傳統跌倒風險評估上，提供了客觀的步態數據作為臨床醫療判斷的參考。雖然本研究之受測者在臨床上都還未達到跌倒風險的評估標準，但就量表之間的分數區間差異，利用穿戴式裝置所量測之步態特徵就具有顯著差異。代表本研究擷取之特徵是可以有效地為傳統評估方式提供客觀性的依據。更重要的，利用穿戴方式的量測方式，可於日常生活隨時進行評估，不需浪費醫療臨床人力使用傳統量表評估方式。未來，可進一步增加量表分數區間的步態數據，提供閥值警告參考，當被分類某特定類別時，當應進一步就醫做臨床評估。

Society is rapidly converted to the super-aged population in Taiwan, so the health and safety of the elderly are important issues. Falling is the most common risk factor and could cause physical and mental diseases to the elderly and the long-term impact on the medical care system. Nowadays, the risk of falling is still determined by clinical assessment and usually using the Berg Balance Scale（BBS）. However, it mainly depends on subjective judgment and the health professionals’ experience, but which was the lack of objective evaluation.
In the experiment, nine participants were recruited, and their ages ranged from 65 and 85 years old. After the evaluation of BBS, they were divided into three groups according to their BBS: Group A: above 55; Group B: 50-54 and Group C: below 49. The participants wore the F-scan and four Xsens dots attached on the left and right toes and heels, respectively. The plantar pressure trajectory（Center of Pressure, COP）and acceleration and angular velocity were measured in walking through the 50 meters corridor. Subsequently, the data were calculated and extracted 57 features as inputs of the support vector machine（SVM）classifier for the BBS three classifications.
The results revealed that 57 features extracted from COP trajectories, angular velocity, and acceleration as inputs of the radial basis function kernel SVM（RBF-SVM）, which could achieve 97% accuracy. When age was the only key feature, it could also reach an accuracy of about 78%. It indicated that the lower the BBS score was, the greater the risk of falling was observed. Furthermore, the accuracy achieved 95% when adding the medial-lateral displacement of the COP as the input. Similarly, it also reflected that the lower the BBS score was, the greater the risk of falling and the greater COP displacement in the medial-lateral directions was observed. Finally, the accuracy rate achieved 99% when the nine key features were used for classification.
The study provided objective information for the traditional fall risk assessment. Although the participants in the present study have not reached the fall risk according to the definition in clinical, their BBS scores interval reflected a significant difference in the gait characteristics measured by the wearable device. The gait features extracted in the present study can effectively provide an objective evaluation for traditional fall risk assessment. More importantly, the wearable device can measure gait performance in daily life instead of the traditional assessment by the medical staff. For future researches and applications, the BBS score interval can be further collected to provide a threshold warning basis. When users are classified into a specific category, they should seek medical attention for further evaluation.

參考文獻
國家發展委員會. （2020）. 中華民國人口推估（2020至2070年）.
陳玉枝, 林麗華, & 簡淑芬. （2002）. 住院病患傷害性跌倒的影響因素與其醫療資源耗用之相關性. 志為護理-慈濟護理雜誌, 1（3）, 66-77.
陳立軒. （2020）. 基於步態壓力中心軌跡進行性別與年齡辨識. NTHU.
蔡宗棋, & 曾櫻枝. （2019）. 步態訓練對社區重覆跌倒老人往前及往後步態之成效探討. 新臺北護理期刊, 21（2）, 53-63.
蔡素蘭, & 林茂榮. （2002）. 跌倒對台灣鄉村社區老人日常生活活動能力衰退與害怕跌倒心理的影響. 物理治療, 27（4）, 155-164.
Alexander, B. H., Rivara, F. P., & Wolf, M. E. （1992）. The cost and frequency of hospitalization for fall-related injuries in older adults. American journal of public health, 82（7）, 1020-1023.
Alexander, N. B. （1996）. Gait disorders in older adults.
Anwary, A. R., Yu, H., & Vassallo, M. （2018）. An automatic gait feature extraction method for identifying gait asymmetry using wearable sensors. Sensors, 18（2）, 676.
Anwary, A. R., Yu, H., & Vassallo, M. . （2018）. Optimal foot location for placing wearable IMU sensors and automatic feature extraction for gait analysis. . IEEE Sensors Journal, 18（6）, 2555-2567.
Bautmans, I., Jansen, B., Van Keymolen, B., & Mets, T. （2011）. Reliability and clinical correlates of 3D-accelerometry based gait analysis outcomes according to age and fall-risk. Gait & posture, 33（3）, 366-372.
Begg, R., & Kamruzzaman, J. （2005）. A machine learning approach for automated recognition of movement patterns using basic, kinetic and kinematic gait data. Journal of biomechanics, 38（3）, 401-408.
Blake, A., Morgan, K., Bendall, M., Dallosso, H., Ebrahim, S., Arie, T. a., . . . Bassey, E. （1988）. Falls by elderly people at home: prevalence and associated factors. Age and ageing, 17（6）, 365-372.
Bogle Thorbahn, L. D., & Newton, R. A. （1996）. Use of the Berg Balance Test to predict falls in elderly persons. Physical therapy, 76（6）, 576-583.
Borowicz, A., Zasadzka, E., Gaczkowska, A., Gawłowska, O., & Pawlaczyk, M. （2016）. Assessing gait and balance impairment in elderly residents of nursing homes. Journal of Physical Therapy Science, 28（9）, 2486-2490.
Burnfield, M. （2010）. Gait analysis: normal and pathological function. Journal of Sports Science and Medicine, 9（2）, 353.
Catalfamo, P., Moser, D., Ghoussayni, S., & Ewins, D. （2008）. Detection of gait events using an F-Scan in-shoe pressure measurement system. Gait & posture, 28（3）, 420-426.
CDC. （2017）. Timed Up & Go.
Chen, B., & Bates, B. T. （2000）. Comparison of F-Scan in-sole and AMTI forceplate system in measuring vertical ground reaction force during gait. Physiotherapy Theory and Practice, 16（1）, 43-53.
Chiu, M.-C., Wu, H.-C., Chang, L.-Y., & Wu, M.-H. （2013）. Center of pressure progression characteristics under the plantar region for elderly adults. Gait & posture, 37（3）, 408-412.
Choi, A., Jung, H., Lee, K. Y., Lee, S., & Mun, J. H. （2019）. Machine learning approach to predict center of pressure trajectories in a complete gait cycle: a feedforward neural network vs. LSTM network. Medical & Biological Engineering & Computing, 57（12）, 2693-2703.
Choi, A., Kang, T. G., & Mun, J. H. （2016）. Biomechanical evaluation of dynamic balance control ability during golf swing. Journal of Medical and Biological Engineering, 36（3）, 430-439.
Choi, A., Lee, J.-M., & Mun, J. H. （2013）. Ground reaction forces predicted by using artificial neural network during asymmetric movements. International Journal of Precision Engineering and Manufacturing, 14（3）, 475-483.
Cortes, C., & Vapnik, V. （1995）. Support-vector networks. Machine learning, 20（3）, 273-297.
Deandrea, S., Lucenteforte, E., Bravi, F., Foschi, R., La Vecchia, C., & Negri, E. （2010）. Risk factors for falls in community-dwelling older people:" a systematic review and meta-analysis". Epidemiology, 658-668.
Derlatka, M., & Bogdan, M. （2015）. Ensemble kNN classifiers for human gait recognition based on ground reaction forces. 2015 8th International Conference on Human System Interaction （HSI）,
Diaz, M. A., Gibbons, M. W., Song, J., Hillstrom, H. J., Choe, K. H., & Pasquale, M. R. （2018）. Concurrent validity of an automated algorithm for computing the center of pressure excursion index （CPEI）. Gait & posture, 59, 7-10.
Florence, C. S., Bergen, G., Atherly, A., Burns, E., Stevens, J., & Drake, C. （2018）. Medical costs of fatal and nonfatal falls in older adults. Journal of the American Geriatrics Society, 66（4）, 693-698.
Gadaleta, M., & Rossi, M. . （2018）. Idnet: Smartphone-based gait recognition with convolutional neural networks. . Pattern Recognition, 74, 25-37.
Gaebler, S. （1993）. Predicting which patient will fall again… and again. Journal of advanced nursing, 18（12）, 1895-1902.
Gillespie, L. D., Robertson, M. C., Gillespie, W. J., Sherrington, C., Gates, S., Clemson, L. M., & Lamb, S. E. （2012）. Interventions for preventing falls in older people living in the community. Cochrane database of systematic reviews（9）.
Gouwanda, D., Gopalai, A. A., & Khoo, B. H. （2016）. . （2016）. A low cost alternative to monitor human gait temporal parameters–wearable wireless gyroscope. . IEEE Sensors Journal, 16（24）, 9029-9035.
Höflinger, F., Müller, J., Zhang, R., Reindl, L. M., & Burgard, W. （2013）. A wireless micro inertial measurement unit （IMU）. IEEE Transactions on instrumentation and measurement, 62（9）, 2583-2595.
Hagedorn, T. J., Dufour, A. B., Golightly, Y. M., Riskowski, J. L., Hillstrom, H. J., Casey, V. A., & Hannan, M. T. （2013）. Factors affecting center of pressure in older adults: the Framingham Foot Study. Journal of foot and ankle research, 6（1）, 1-5.
Haim, A., Rozen, N., & Wolf, A. （2010）. The influence of sagittal center of pressure offset on gait kinematics and kinetics. Journal of biomechanics, 43（5）, 969-977.
Han, S., Qubo, C., & Meng, H. （2012）. Parameter selection in SVM with RBF kernel function. World Automation Congress 2012,
Harris, G. F., & Wertsch, J. J. （1994）. Procedures for gait analysis. Archives of Physical Medicine and Rehabilitation, 75（2）, 216-225.
Hausdorff, J. M., Rios, D. A., & Edelberg, H. K. （2001）. Gait variability and fall risk in community-living older adults: a 1-year prospective study. Archives of Physical Medicine and Rehabilitation, 82（8）, 1050-1056.
Hessert, M. J., Vyas, M., Leach, J., Hu, K., Lipsitz, L. A., & Novak, V. （2005）. Foot pressure distribution during walking in young and old adults. BMC geriatrics, 5（1）, 1-8.
Howcroft, J., Kofman, J., Lemaire, E. D., & McIlroy, W. E. （2016）. Analysis of dual-task elderly gait in fallers and non-fallers using wearable sensors. Journal of biomechanics, 49（7）, 992-1001.
Huang, S.-C., Lu, T.-W., Chen, H.-L., Wang, T.-M., & Chou, L.-S. （2008）. Age and height effects on the center of mass and center of pressure inclination angles during obstacle-crossing. Medical engineering & physics, 30（8）, 968-975.
Jančová, J. （2008）. Measuring the balance control system–review. Acta Medica （Hradec Kralove）, 51（3）, 129-137.
Jung, J.-W., Bien, Z., Lee, S.-W., & Sato, T. （2003）. Dynamic-footprint based person identification using mat-type pressure sensor. Proceedings of the 25th Annual International Conference of the IEEE Engineering in Medicine and Biology Society （IEEE Cat. No. 03CH37439）,
Lee, H.-S., & Choi, J.-H. （2008）. Correlation between BBS, FRT, STI, TUG, MBI, and falling in stroke patients. Journal of Korean Physical Therapy, 20（4）, 1-6.
Leveille, S. G., Guralnik, J. M., Ferrucci, L., Hirsch, R., Simonsick, E., & Hochberg, M. C. （1998）. Foot pain and disability in older women. American journal of epidemiology, 148（7）, 657-665.
Manap, H. H., Tahir, N. M., & Yassin, A. I. M. （2011）. Anomalous gait detection based on support vector machine. 2011 IEEE International Conference on Computer Applications and Industrial Electronics （ICCAIE）,
Marigold, D. S., & Patla, A. E. （2008）. Age-related changes in gait for multi-surface terrain. Gait & posture, 27（4）, 689-696.
Mason, J. E., Traoré, I., & Woungang, I. （2016）. Machine Learning Techniques for Gait Biometric Recognition. Springer.
Meyer, B. M., Tulipani, L. J., Gurchiek, R. D., Allen, D. A., Adamowicz, L., Larie, D., . . . McGinnis, R. （2020）. Wearables and deep learning classify fall risk from gait in multiple sclerosis. IEEE journal of biomedical and health informatics.
Muñoz-Organero, M., Parker, J., Powell, L., Davies, R., & Mawson, S. （2017）. Sensor optimization in smart insoles for post-stroke gait asymmetries using total variation and L 1 distances. IEEE Sensors Journal, 17（10）, 3142-3151.
Muro-De-La-Herran, A., Garcia-Zapirain, B., & Mendez-Zorrilla, A. （2014）. Gait analysis methods: An overview of wearable and non-wearable systems, highlighting clinical applications. Sensors, 14（2）, 3362-3394.
Nakano, T., Nukala, B. T., Tsay, J., Zupancic, S., Rodriguez, A., Lie, D. Y., . . . Nguyen, T. Q. （2017）. Gaits classification of normal vs. patients by wireless gait sensor and Support Vector Machine （SVM） classifier. International Journal of Software Innovation （IJSI）, 5（1）, 17-29.
Nigg, B., Fisher, V., & Ronsky, J. （1994）. Gait characteristics as a function of age and gender. Gait & posture, 2（4）, 213-220.
O'Loughlin, J. L., Robitaille, Y., Boivin, J.-F., & Suissa, S. （1993）. Incidence of and risk factors for falls and injurious falls among the community-dwelling elderly. American journal of epidemiology, 137（3）, 342-354.
Özdemir, A. T., & Barshan, B. . （2014）. Detecting falls with wearable sensors using machine learning techniques. Sensors, 14（6）, 10691-10708.
Park, J. W., Jung, M., & Kweon, M. （2014）. The mediolateral CoP parameters can differentiate the fallers among the community-dwelling elderly population. Journal of Physical Therapy Science, 26（3）, 381-384.
Park, S.-H. （2018）. Tools for assessing fall risk in the elderly: a systematic review and meta-analysis. Aging clinical and experimental research, 30（1）, 1-16.
Patterson, M. R., Johnston, W., O'Mahony, N., O'Mahony, S., Nolan, E., & Caulfield, B. . （2016）. Validation of temporal gait metrics from three IMU locations to the gold standard force plate. In 2016 38th Annual International Conference of the IEEE Engineering in Medicine and Biology Society （EMBC） （pp. 667-671）. IEEE.
Perry, J., & Davids, J. R. （1992）. Gait analysis: normal and pathological function. Journal of Pediatric Orthopaedics, 12（6）, 815.
Pfortmueller, C., Lindner, G., & Exadaktylos, A. （2014）. Reducing fall risk in the elderly: risk factors and fall prevention, a systematic review. Minerva Med, 105（4）, 275-281.
Podsiadlo, D., & Richardson, S. （1991）. The timed “Up & Go”: a test of basic functional mobility for frail elderly persons. Journal of the American Geriatrics Society, 39（2）, 142-148.
Pradhan, C., Wuehr, M., Akrami, F., Neuhaeusser, M., Huth, S., Brandt, T., ... & Schniepp, R. （2015）. Automated classification of neurological disorders of gait using spatio-temporal gait parameters. Journal of Electromyography and Kinesiology, 25（2）, 413-422.
Qian, G., Zhang, J., & Kidané, A. （2008）. People identification using gait via floor pressure sensing and analysis. European Conference on Smart Sensing and Context,
Rehman, R. Z. U., Del Din, S., Guan, Y., Yarnall, A. J., Shi, J. Q., & Rochester, L. （2019）. Selecting clinically relevant gait characteristics for classification of early parkinson’s disease: A comprehensive machine learning approach. Scientific reports, 9（1）, 1-12.
Reynaud, V., Muti, D., Pereira, B., Greil, A., Caillaud, D., Richard, R., . . . Costes, F. （2019）. A TUG value longer than 11 s predicts fall risk at 6-month in individuals with COPD. Journal of clinical medicine, 8（10）, 1752.
Rispens, S. M., van Schooten, K. S., Pijnappels, M., Daffertshofer, A., Beek, P. J., & van Dieën, J. H. （2015）. Do extreme values of daily-life gait characteristics provide more information about fall risk than median values? JMIR research protocols, 4（1）, e4.
Salzman, B. （2010）. Gait and balance disorders in older adults. American family physician, 82（1）, 61-68.
Schinkel-Ivy, A., Wong, J. S., & Mansfield, A. （2017）. Balance confidence is related to features of balance and gait in individuals with chronic stroke. Journal of Stroke and Cerebrovascular Diseases, 26（2）, 237-245.
Scott, G., Menz, H. B., & Newcombe, L. （2007）. Age-related differences in foot structure and function. Gait & posture, 26（1）, 68-75.
Segel, J. D., & Crawford, S. （2014）. Anatomy of the COP gait line and computer-aided gait analysis. Pm’s Tech Forum/Orthotics & Biomechanics,
Senden, R., Savelberg, H., Grimm, B., Heyligers, I., & Meijer, K. （2012）. Accelerometry-based gait analysis, an additional objective approach to screen subjects at risk for falling. Gait & posture, 36（2）, 296-300.
Shetty, S., & Rao, Y. （2016）. SVM based machine learning approach to identify Parkinson's disease using gait analysis. 2016 International Conference on Inventive Computation Technologies （ICICT）,
Song, J., Hillstrom, H., Secord, D., & Levitt, J. （1996）. Foot type biomechanics. comparison of planus and rectus foot types. Journal of the American Podiatric Medical Association, 86（1）, 16-23.
Spaepen, A., Vranken, M., & Willems, E. （1977）. COMPARISON OF MOVEMENTS OF CENTER OF GRAVITY AND OF CENTER OF PRESSURE IN STABILOMETRIC STUDIES. Agressologie, 18（2）, 109-113.
Steffen, T. M., Hacker, T. A., & Mollinger, L. （2002）. Age-and gender-related test performance in community-dwelling elderly people: Six-Minute Walk Test, Berg Balance Scale, Timed Up & Go Test, and gait speeds. Physical therapy, 82（2）, 128-137.
Sudheer, C., Maheswaran, R., Panigrahi, B. K., & Mathur, S. （2014）. A hybrid SVM-PSO model for forecasting monthly streamflow. Neural Computing and Applications, 24（6）, 1381-1389.
Syarif, I., Prugel-Bennett, A., & Wills, G. （2016）. SVM parameter optimization using grid search and genetic algorithm to improve classification performance. Telkomnika, 14（4）, 1502.
Taborri, J., Palermo, E., Rossi, S., & Cappa, P. （2016）. Gait partitioning methods: A systematic review. Sensors, 16（1）, 66.
TarniŢă, D. （2016）. Wearable sensors used for human gait analysis. Rom J Morphol Embryol, 57（2）, 373-382.
Tinetti, M. E. （1986）. Performance-oriented assessment of mobility problems in elderly patients. Journal of the American Geriatrics Society.
Tinetti, M. E., De Leon, C. F. M., Doucette, J. T., & Baker, D. I. （1994）. Fear of falling and fall-related efficacy in relationship to functioning among community-living elders. Journal of gerontology, 49（3）, M140-M147.
Tinetti, M. E., Speechley, M., & Ginter, S. F. （1988）. Risk factors for falls among elderly persons living in the community. New England journal of medicine, 319（26）, 1701-1707.
Ungar, A., Rafanelli, M., Iacomelli, I., Brunetti, M. A., Ceccofiglio, A., Tesi, F., & Marchionni, N. （2013）. Fall prevention in the elderly. Clinical Cases in mineral and bone metabolism, 10（2）, 91.
Weerdesteijn, V., Niet, M. d., Van Duijnhoven, H., & Geurts, A. C. （2008）. Falls in individuals with stroke.
Winter, D. A. （1995）. Human balance and posture control during standing and walking. Gait & posture, 3（4）, 193-214.
Wu, C.-C., Chen, Y.-J., Hsu, C.-S., Wen, Y.-T., & Lee, Y.-J. （2020）. Multiple Inertial Measurement Unit Combination and Location for Center of Pressure Prediction in Gait. Frontiers in Bioengineering and Biotechnology, 8, 1252.