Abstract
Early detection of falls is important for reducing fall injuries. However, existing fall detection strategies mostly focus on reducing impact injuries rather than avoiding falls. This study proposed the concept of identifying “Imbalance Point” to warn the body imbalance, allowing sufficient time to recover balance. And if falling cannot be avoided, an impact sign is released by detecting the “Fall Point” prior to the impact. To achieve this goal, motion prediction model and balance recovery model are integrated into a spatiotemporal framework to analyze dynamic and kinematic features of body motion. Eight healthy young volunteers participated in three sets of experiment: Normal trial, Recovery trial and Fall trial. The body motion in the trials was recorded using Microsoft Azure Kinect. The results show that the developed algorithm for Fall Point detection achieved 100% sensitivity and 98.6% specificity, along with an average lead time of 297 ms. Moreover, Imbalance Point was successfully detected in all Fall trials, and the average time interval between Imbalance Point and Fall Point was 315 ms, longer than reported step reaction time for elderly (approximately 270 ms). The experiment results demonstrate that the developed algorithm have great potential for fall warning and protection in the elderly.
Graphical Abstract
Similar content being viewed by others
References
Alghwiri AA, Whitney SL (2012) Geriatric physical therapy. Mosby, Saint Louis, p 353
Rubenstein LZ (2006) Falls in older people: epidemiology, risk factors and strategies for prevention, (in English). Age Ageing 35:37–41
Xu T, Zhou Y, Zhu J (2018) New advances and challenges of fall detection systems: a survey. Appl Sci Basel 8(3):418
Wang Y, Wu K, Ni LM (2016) Wifall: Device-free fall detection by wireless networks. IEEE Trans Mob Comput 16(2):581–594
Tzeng H-W, Chen M-Y, Chen J-Y (2010) Design of fall detection system with floor pressure and infrared image. In: 2010 International Conference on System Science and Engineering, pp 131–135: IEEE
Makhlouf A, Boudouane I, Saadia N, Cherif AR (2019) Ambient assistance service for fall and heart problem detection. J Ambient Intell Humanized Comput 10(4):1527–1546
Wang J, Zhang ZQ, Li B, Lee S, Sherratt RS (2014) An enhanced fall detection system for elderly person monitoring using consumer home networks. IEEE Trans Consum Electron 60(1):23–29
Dang TN, Le TK, Hong TP, Nguyen V (2021) Fast and accurate fall detection and warning system using image processing technology. In: International Conference on Advanced Technologies for Communications (ATC), Ho Chi Minh City, Vietnam, 2021, pp 207–210
Saadeh W, Butt SA, Bin Altaf MA (2019) A patient-specific single sensor IoT-based wearable fall prediction and detection system. IEEE Trans Neural Syst Rehabil Eng 27(5):995–1003
Al Nahian MJ et al (2021) Towards an accelerometer-based elderly fall detection system using cross-disciplinary time series features. IEEE Access 9:39413–39431
Panahi L, Ghods V (2018) Human fall detection using machine vision techniques on RGB-D images. Biomed Signal Process Control 44:146–153
Min WD, Yao LY, Lin ZR, Liu L (2018) Support vector machine approach to fall recognition based on simplified expression of human skeleton action and fast detection of start key frame using torso angle. IET Comput Vision 12(8):1133–1140
Hu XY, Qu XD (2016) Pre-impact fall detection. Biomed Eng Online 15:61
Yajai A, Rasmequan S (2017) Adaptive directional bounding box from RGB-D information for improving fall detection. J Vis Commun Image Represent 49:257–273
Chen W, Jiang Z, Guo H, Ni XJS (2020) Fall detection based on key points of human-skeleton using openpose. Symmetry 12(5):744
Tamura T, Yoshimura T, Sekine M, Uchida M, Tanaka O (2009) A wearable airbag to prevent fall injuries. IEEE Trans Inf Technol Biomed 13(6):910–914
Ahn S et al (2018) Optimization of a pre-impact fall detection algorithm and development of hip protection airbag system. Sensors Mater 30:1743–1752
Yao M et al (2015) A wearable pre-impact fall early warning and protection system based on MEMS inertial sensor and GPRS communication. In: 2015 IEEE 12th international conference on wearable and implantable body sensor networks (BSN), pp 1–6: IEEE
Choi W, Wakeling J, Robinovitch S (2015) Kinematic analysis of video-captured falls experienced by older adults in long-term care. J Biomech 48(6):911–920
Vallee P, Tisserand R, Robert T (2015) Possible recovery or unavoidable fall? A model to predict the one step balance recovery threshold and its stepping characteristics. J Biomech 48(14):3905–3911
Cyr M-A, Smeesters C (2009) Kinematics of the threshold of balance recovery are not affected by instructions limiting the number of steps in younger adults. Gait Posture 29(4):628–633
Smeesters C, Hayes WC, McMahon TA (2001) The threshold trip duration for which recovery is no longer possible is associated with strength and reaction time. J Biomech 34(5):589–595
Xu T, Zhou Y (2018) Elders’ fall detection based on biomechanical features using depth camera. Int J Wavelets Multiresolution Inf Process 16(02):1840005
Lugade V, Lin V, Chou L-S (2011) Center of mass and base of support interaction during gait. Gait Posture 33(3):406–411
Hof AL, Gazendam MGJ, Sinke WE (2005) The condition for dynamic stability, (in English). J Biomech 38(1):1–8
Bi QL et al (2017) An automatic camera calibration method based on checkerboard. Traitement Du Signal 34(3–4):209–226
Albert JA, Owolabi V, Gebel A, Brahms CM, Granacher U, Arnrich B (2020) Evaluation of the pose tracking performance of the Azure Kinect and Kinect v2 for gait analysis in comparison with a gold standard: a pilot study. Sensors 20(18):5104
Manghisi VM et al (2020) A body tracking-based low-cost solution for monitoring workers’ hygiene best practices during pandemics. Sensors 20(21):6149
Romeo L, Marani R, Malosio M, Perri AG, D'Orazio T, and Ieee (2021) Performance Analysis of Body Tracking with the Microsoft Azure Kinect. In: 29th Mediterranean Conference on Control and Automation (MED), Puglia, Italy, 2021, pp 572–577
Berg WP, Alessio HM, Mills EM, Tong C (1997) Circumstances and consequences of falls in independent community-dwelling older adults. Age Ageing 26(4):261–268
Roudsari BS, Ebel BE, Corso PS, Molinari NAM, Koepsell TD (2005) The acute medical care costs of fall-related injuries among the US older adults. Injury-Int J Care Injured 36(11):1316–1322
Roos PE, McGuigan MP, Trewartha G (2010) The role of strategy selection, limb force capacity and limb positioning in successful trip recovery. Clin Biomech 25(9):873–878
King ST, Eveld ME, Martinez A, Zelik KE, Goldfarb M (2019) A novel system for introducing precisely-controlled, unanticipated gait perturbations for the study of stumble recovery. J Neuroengineering Rehabil 16:69
Shirota C, Simon AM, Kuiken TA (2014) Trip recovery strategies following perturbations of variable duration. J Biomech 47(11):2679–2684
van Dieen JH, Pijnappels M, Bobbert MF (2005) Age-related intrinsic limitations in preventing a trip and regaining balance after a trip. Saf Sci 43(7):437–453
Pijnappels M, Reeves ND, Maganaris CN, Van Dieen JH (2008) Tripping without falling; lower limb strength, a limitation for balance recovery and a target for training in the elderly. J Electromyogr Kinesiol 18(2):188–196
Chen SH, Liu BA, Feng C, Vallespi-Gonzalez C, Wellington C (2021) 3D point cloud processing and learning for autonomous driving: impacting map creation, localization, and perception. IEEE Signal Process Mag 38(1):68–86
Soomro K, Idrees H, Shah M (2018) Online localization and prediction of actions and interactions. IEEE Trans Pattern Anal Mach Intell 41(2):459–472
Ghosh P, Song J, Aksan E, Hilliges O, and Ieee (2017) Learning human motion models for long-term predictions. In: International Conference on 3D Vision (3DV), Qingdao, Canada, 2017, pp 458–466
Jin X, Guo J, Li Z, Wang RH (2020) Motion prediction of human wearing powered exoskeleton. Math Probl Eng 2020:8899880
Xu Z, Wang Y, Long M, Wang J, KLiss M (2018) PredCNN: predictive learning with cascade convolutions. In: Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence, pp 2940–2947
Liu XL, Yin JQ, Liu J, Ding PX, Liu J, Liu HP (2021) TrajectoryCNN: a new spatio-temporal feature learning network for human motion prediction. IEEE Trans Circuits Syst Video Technol 31(6):2133–2146
Fragkiadaki K, Levine S, Felsen P, Malik J (2015) Recurrent network models for human dynamics. In: Proceedings of the IEEE International Conference on Computer Vision, pp 4346–4354
Cheng Y, Zhao W, Liu C, Tomizuka M (2019) Human motion prediction using semi-adaptable neural networks. In: 2019 American Control Conference (ACC), pp 4884–4890: IEEE
Liu RX, Liu CL (2021) Human motion prediction using adaptable recurrent neural networks and inverse kinematics. IEEE Control Syst Lett 5(5):1651–1656
Melzer I, Kurz I, Shahar D, Levi M, Oddsson L (2007) Application of the voluntary step execution test to identify elderly fallers. Age Ageing 36(5):532–537
Tisserand R, Robert T, Chabaud P, Bonnefoy M, Cheze L (2016) Elderly fallers enhance dynamic stability through anticipatory postural adjustments during a choice stepping reaction time. Front Human Neurosci 10:613
Paolo DL (1996) Adjustments to Zatsiorsky-Seluyanov’s segment inertia parameters, (in English). J Biomech 29(9):1223–1230
Pavol MJ, Owings TM, Foley KT, Grabiner MD (2001) Mechanisms leading to a fall from an induced trip in healthy older adults. J Gerontol Ser a-Biol Sci Med Sci 56(7):M428–M437
Pijnappels M, Bobbert MF, van Dieen JH (2004) Contribution of the support limb in control of angular momentum after tripping. J Biomech 37(12):1811–1818
Deb K, Pratap A, Agarwal S, Meyarivan T (2002) A fast and elitist multiobjective genetic algorithm: NSGA-II. IEEE Trans Evol Comput 6(2):182–197
Charfi I, Miteran J, Dubois J, Atri M, Tourki R (2013) Optimized spatio-temporal descriptors for real-time fall detection: comparison of support vector machine and Adaboost-based classification. J Electron Imaging 22(4):041106
Hsiao ET, Robinovitch SN (1999) Biomechanical influences on balance recovery by stepping, (in English). J Biomech 32(10):1099–1106
Li M, Xu G, He B, Ma X, Xie J (2018) Pre-impact fall detection based on a modified zero moment point criterion using data from Kinect sensors. IEEE Sens J 18(13):5522–5531
Rescio G, Leone A, Siciliano P (2018) Supervised machine learning scheme for electromyography-based pre-fall detection system. Expert Syst Appl 100:95–105
Lee CM, Park J, Park S, Kim CH (2020) Fall-detection algorithm using plantar pressure and acceleration data. Int J Precis Eng Manuf 21(4):725–737
Jung H, Koo B, Kim J, Kim T, Nam Y, Kim Y (2020) Enhanced algorithm for the detection of preimpact fall for wearable airbags. Sensors 20(5):1277
Yu X, Jang J, Xiong S (2021) A large-scale open motion dataset (KFall) and benchmark algorithms for detecting pre-impact fall of the elderly using wearable inertial sensors. Front Aging Neurosci 13:692865
de Sousa F, Escriba C, Bravo EGA, Brossa V, Fourniols JY, Rossi C (2022) Wearable pre-impact fall detection system based on 3D accelerometer and subject’s height. IEEE Sens J 22(2):1738–1745
Neptune RR, Vistamehr A (2019) Dynamic balance during human movement: measurement and control mechanisms, (in English). J Biomech Eng Trans Asme 141(7):070801
Acknowledgements
This work was supported by National Key Research and Development Program of China (2022YFC2009500) and Key Research Project of Zhejiang Lab (2022ND0AC01).
Author information
Authors and Affiliations
Corresponding authors
Ethics declarations
Ethical approval
The study was conducted in accordance with the Declaration of Helsinki. The study was approved by the ethical review boards of Zhejiang Hospital (2021(108 K)).
Conflict of interest
The authors declare no competing interests.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendix
Appendix
The model motion governing equations and the detailed derivation during pre-contact phase and contact phase are given in the appendix.
1.1 Basic settings
pre-contact phase: \({F}_{leg}={\left[0\quad0\quad0\right]}^{T}\)
contact phase:
where \(\alpha ,\beta ,\gamma ,\theta\) are joint angles and \({k}_{\text{ankle}},{k}_{\text{knee}},{k}_{hip}\) are the spring coefficients of the torque springs acting on each joint, respectively; \({k}_{leg}\) is the spring coefficient of the linear spring Leg; \({F}_{leg}\) is the force generated by the Leg on the recovery side, \({f}_{leg}\) is the magnitude of force; \({l}_{leg\_o},{l}_{leg\_r}\) are the Leg original length and the length during the balance recovery process, respectively. \(\tau\) is the joint torque; F is the joint reaction force; I is body segment moment of inertia, m is body segment mass, l is body segment length. The first letter of the subscript indicates the body segment, and the second letter indicates the proximal or distal end of the body segment. s stands for Shank, t stands for Thigh, h stands for Hat, p stands for proximal, d stands for distal. \({P}_{s},{P}_{t},{P}_{h}\) are the coordinates of the center of mass of the Shank, Thigh and HAT, respectively.
1.2 Shank motion equations
1.3 Thigh motion equations
1.4 HAT motion equations
1.5 Standard motion equation form of balance recovery model
where \(D\left(q\right)\) is inertia matrix, \(C\left(q,\dot{q}\right)\) is centrifugal force and Coriolis force matrix, \(G\left(q\right)\) is gravity matrix, \(U\left(q\right)\) is Generalized force matrix.
\(D\left(q\right)\):inertia matrix
\(C\left(q,\dot{q}\right)\):centrifugal force and Coriolis force matrix
\(G\left(q\right)\):gravity matrix
\(U\left(q\right)\):generalized force matrix
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Xu, S., Yang, Z., Wang, D. et al. A dynamic spatiotemporal model for fall warning and protection. Med Biol Eng Comput 62, 1061–1076 (2024). https://doi.org/10.1007/s11517-023-02999-5
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11517-023-02999-5